(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Malacologia"

HARVARD UNIVERSITY 

Library of the 

Museum of 

Comparative Zoology 



14 i -f- 
VOL 36, NO. 1-2 1995 



MALACOLOGIA 



nternational Journal of Malacology 
Revista Internacional de Malacologia 
Journal International de Malacologie 
Международный Журнал Малакологии 
Internationale Malakologische Zeitschrift 



MALACOLOGIA 

Editor-in-Chief: 
GEORGE M. DAVIS 

Editorial and Subscription Offices: 

Department of Malacology 

The Academy of Natural Sciences of Philadelphia 

1900 Benjamin Franklin Parkway 

Philadelphia, Pennsylvania 19103-1195, U.S.A. 



EUGENE COAN 

California Academy of Sciences 

San Francisco, CA 



Co-Editors: 



Assistant Managing Editor: 

CARYL HESTERMAN 

Associate Editors: 



CAROL JONES 
Denver, CO 



JOHN B. BURCH 
University of Michigan 
Ann Arbor 



ANNE GISMANN 
Maadi 

Egypt 



MALACOLOGIA is published by the INSTITUTE OF MALACOLOGY, the Sponsor Members of 
which (also serving as editors) are: 



KENNETH J. BOSS 

Museum of Comparative Zoology 

Cambridge, Massachusetts 

JOHN BURCH, President 

MELBOURNE R. CARRIKER 
University of Delaware, Lewes 

GEORGE M. DAVIS 
Secretary and Treasurer 

CAROLE S. HICKMAN 
University of California, Berkeley 
President-Elect 



JAMES NYBAKKEN 

Moss Landing Marine Laboratory 

California 

CLYDE F. E. ROPER 
Smithsonian Institution 
Washington, D.C. 

W. D. RUSSELL-HUNTER 
Syracuse University, New York 

SHI-KUEI WU 

University of Colorado Museum, Boulder 



Participating Members 

EDMUND GITTENBERGER JACKIE L. VAN GOETHEM 

Secretary, UNITAS MAU\COLOGICA Treasurer, UNITAS MAUACOLOGICA 

Rijksmuseum van Natuurlijke Koninklijk Belgisch Instituut 

Historie voor Natuurwetenschappen 

Leiden, Netherlands Brüssel, Belgium 



J. FRANCIS ALLEN, Emérita 
Environmental Protection Agency 
Washington, D.C. 



Emeritus Members 

ROBERT ROBERTSON 

The Academy of Natural Sciences 

Philadelphia, Pennsylvania 



ELMER G. BERRY, 
Germantown, Maryland 



Copyright © 1995 by the Institute of Malacology 



Lib 



Ml Û 3 1У95 



J. A. ALLEN 

Marine Biological Station 

Millport, United Kingdom 

R. BIELER 
Field Museum 
Chicago, U.S.A. 

E. E. BINDER 

Museum d'Histoire Naturelle 

Genève, 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. CU\RKE, Jr. 
Portland, Texas, U.S.A. 

B. С CLARKE 
University of Nottingham 
United Kingdom 

R. DILLON 

College of Charleston 

SC, U.S.A. 

C. J. DUNCAN 
University of Liverpool 
United Kingdom 

D. J. EERNISSE 
University of Michigan 
Ann Arbor, U.S.A. 

E. GITTENBERGER 
Rijksmuseum van Natuurlijke Historie 
Leiden, Netherlands 



1995 
EDITORIAL BOARD 

HARVARD 

F. GIUSTI ' ■ ' ^ '■ — 
Université di Siena, Italy 

A. N. GOLIKOV 
Zoological Institute 
St. Petersburg, Russia 

S. 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 
Göteborg, Sweden 

S. HUNT 
Lancashire 
United Kingdom 

R. JANSSEN 

Forschungsinstitut Senckenberg, 
Frankfurt am Main, Germany 

R. N. KILBURN 
Natal Museum 
Pietermaritzburg, South Africa 

M. A. KLAPPENBACH 

Museo Nacional de Historia Natural 

Montevideo, Uruguay 

J. KNUDSEN 

Zoologisk Institut & Museum 

Kebenhavn, Denmark 



A. J. KOHN 

University of Washington 

Seattle. U.S.A. 

A. LUCAS 

Faculté des Sciences 

Brest, France 

C. MEIER-BROOK 
Tropenmedizinisches Institut 
Tübingen, Germany 

H. К. 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. 0KLAND 
University of Oslo 
Norway 

T. OKUTANI 
University of Fisheries 
Tokyo, Japan 

W. L. PARAENSE 

Instituto Oswalde 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. F. PONDER 
Australian Museum 
Sydney 

Ol Z. Y. 

Academia Sínica 

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. STAÑCZYKOWSKA 

Siedlce, Poland 

F. STARMÜHLNER 

Zoologisches Institut der Universität 

Wien, Austria 

Y. I. STAROBOGATOV 
Zoological Institute 
St. Petersburg, Russia 

W. STREIFF 
Université de Caen 
France 

J. STUARDO 
Universidad de Chile 
Valparaiso 

S. TILLIER 

Muséum National d'Histoire Naturelle 

Paris, France 

R. D. TURNER 
Harvard University 
Cambridge, Mass., U.S.A. 

J.A.M. VAN DEN BIGGELAAR 
University of Utrecht 
The Netherlands 

J. A. 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 



MAIJ\COLOGIA, 1995, 36(1-2): 1-14 

MORFOLOGÍA DEL ESTOMAGO Y PARTES BLANDAS EN MYTELLA STRIGATA 
(HANLEY, 1843) (BIVALVIA: MYTILIDAE) 

Maria 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 M. 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 III (Purchon, 
1957) with sorting mechanisms of type В (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, 1 955) 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 (Merhl & Turner, 1963), Xe- 
nostrobus (Wilson, 1967), Adula (Botula) 
(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), M. strigata (Hanley, 
1843), M. speciosa (Reeve, 1857), y M. tum- 
bezensis (Pilsbry & OIsson, 1 935) (Keen 1 971 ; 
García-Cubas & Reguero, 1987); aunque 
Bernard (1983) considera a las dos últimas 
como sinónimos. 

La distribución de M. 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, M. 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 & 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 & 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 M. 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 & 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 & 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 /W. charruana, rea- 
lizado por Narchi & Galváo-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'W) y Chautengo (99°02'N, 
99°09'W) (Guerrero; Nov. 1974-Mayo 1975) y 



^Esc. Biología, Lab. Invertebrados, Univ. Michoacana de San Nicolás de Hidalgo, Apdo. Postal 59-3, Morelia 58021, 

Michoacán, México. 

^Fac. Cieñe, Biol. Rec. Nat.. Depto. de Oceanografía Univ. Concepción, Casilla 2407, Concepción, Chile. 



1 



VILlJ\RROEL & STUARDO 



de la Laguna de Cuyutlán, (19°02'N, 
104^^1 9'W) (Colima; Ago. 1977-Jul. 1978). La 
colecta se realizó 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 ejemplares 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) y 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 M. strígata 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. la). 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- 
pante, 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 M. 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. la), a 
diferencia de otros mitílidos (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 & Galváo-Bueno, 1983; 
Paiva Avelar & 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 continuos, 
musculares y unidos en la región posterior 
para formar un sifón (Narchi & Galváo-Bueno, 
1983). Según White (1937), en Mytilus edulis 
este pliegue es ciliado y muy sensible al tacto 



MORFOLOGÍA DE MYTELLA STRIGATA 

RA Ri Ар 




(b) DD 



Pe V Ri 




FIG. 1. Mytella strigata. (a) Manto izquierdo, (b) Vista ventral del tracto digestivo, (с) Cavidad palea!, sin 
branquia izquierda ni glándula digestiva. 

A, ano; AA, aductor anterior; Aa, aorta anterior; AE, abertura exhalante; Ai, abertura inhalante; AM, arteria 
del manto; AR, aductor posterior; Ap, aorta posterior; C, ctenidio; CD, capuchón dorsal; CV, conectivo 
visceral; DD, ductos de los divertículos digestivos; E, esófago; Es, estómago; G, gónada; Gp, glándula 
pericárdica; I, 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, riñon; RP, retractor pedal 
posterior; SE, saco del estilo; V, ventrículo. 



VILU\RROEL & 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 débilmente. 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 & Galváo-Bueno, 1983). Yonge 
(1 948) 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 Mytella utilizando las carac- 
terísticas 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 strígata 
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. la, c, 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, c, RA, RP), como en Choro- 
mytilus, en cambio en Mytilus los haces están 
contiguos. Las diferencias funcionales no se 
han determinado pero, aparentemente están 
ligadas a una mayor actividad del animal. 



Pie. 



El pie de M. strígata (Fig. 1c, p), 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. strígata, 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. Ib, c, que muestra 
parte de las lámelas del lado derecho). Son 
branquias planas y homorrábdicas con una 
distribución de tractos ciliares descritos en 
detalle en M. charruana por Narchi & Galváo- 
Bueno (1983). 

Palpos Labiales. 

Los palpos de M. strígata, como los de M. 
charruana (Narchi & Galváo-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. 1c, PL, M); hay numerosos sur- 
cos que se aprecian por transparencia (Fig. 
1 с). 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. strígata (Fig. Ib, 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. strígata 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 



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 
intestine, 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 M. striatulus (Dinamani, 1967). 
Dos especies de Lithophaga: L. nasuta (Pur- 
chon, 1957) y L. gracilis (Dinamani, 1967); 
Limnoperna fortunei (Morton, 1973); Muscu- 
li sta 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- 
ruana (Narchi & Galváo-Bueno, 1983); y 
Brachidontes solisianus y В. 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. 1 b, 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 
embabiamiento 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 




FIG 2 Mytella strigata. Estómago, (a) Vista ventral, (b) Detalle del ciego seleccionador de alimento, (с), d) 
у (e) Vistas dorsal lateral derecha e izquierda respectivamente, (a), (с), (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; E , 
abertura esofágica; I, intestino; Ra, repliegue axial; RL, repliegue lateral; SE, saco del estilo; SI, surco 
intestinal; TY, tiflosol mayor; TY', término TY. 



morfología de mytella strigata 




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; I, 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- 
dor de alimento, sobre la pared izquierda la- 
teral a la abertura esofágica. En cambio, en 



Modiolus modiolus (Nelson, 1918) y en 
Mytilus edulis (Graham, 1949) termina dentro 
del ciego seleccionador de alimente. 



8 



VILLARROEL & 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 en 
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 partícu- 
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 В 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 atrás 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. strígata, 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 en M. 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 pronninente 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 ventrículo, 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 /И. 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 
o indiferenciado (Fig. la), 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 visceras. 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 1a. 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. 1c), 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 mitílidos (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 lumen 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 mitílidos, 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 & 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. 

Órganos Excretores. 

El riñon (Figs, la, c, Ri) en M. 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, 1 958: fig. 1 41 ) al igual que 
en Brachidontes darwinianus y S. solisianus 
(Avelar & Narchi, 1984 a, b). 

El riñon, 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 órgano 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 (FU). 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. la, G) 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 M. 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 & 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 & 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. 

IVIytella como Modiolus, no anida en el lodo 
con el biso como ocurre en Musculus, Mus- 
culista y Amyqdalum, a los que Morton (1 977) 
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 & Galväo-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 
definirían también una condición primitiva 
entre los Mytilidae (Yonge & Campbell, 1 968). 

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- 
idas 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 en M. strígata y M. 
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, o de arena fina en habitat rocosos con 
alta energía, como se ha observado en 
Mytilus edulis y en Perna perna (Narchi & Gal- 
vâo-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 partículas 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 III 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. Рею, tam- 
bién parece plausible considerar a este ciego 
como una estructura seleccionadora, basán- 
donos en que M. 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 partículas; 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 & 
Galváo-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 



VILL7\RR0EL & STUARDO 



ñeros Modiolus, Mytilus, Lithophaga, Perna, 
Arcuatula y Botula. AI igual que M. charruana 
posee sifones del tipo A (Yonge, 1948), 
ctenidios del tipo B(1) (Atkins, 1937) y el es- 
tómago es del tipo III (Purchon, 1957) con 
mecanismos de selección de tipo В (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 III: 
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 & С 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, L D. & N. R. 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. С PFEIF- 
FER & M. FISZMAN, 1983, Size and metal con- 
centration in the mangrove mussel Mytella guya- 
nensis (Mollusca: Bivalvia) from Baia de Sepetiva 
Brasil. Revista de Biología 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 sthgata (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, I. A., 1922, Biology and economic value of 
the sea mussel Mytilus edulis. United States Bu- 
reau of Fisheries, Bulletin, 38: 127-259. 

GARCÍA-CUBAS, A. & M. 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 Oceanógraficas 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: 737-778. 

HOLGUÍN Q., O. E. & A. С GONZÁLEZ, 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, T. С & M. 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, R. M. V. & I. N. SILVA, 1988, Estudo de 
sobrevivencia e de capacidade de ¡solamente de 
Mytella guyanensis (Moll, em diferentes salini- 
dades. Revista Nordestina Biología, 6: 34-41. 

MERRILL, A. S. & R. D. TURNER, 1963, Nest build- 
ing in the bivalve genera Musculus and Lima. 
Veliger, 6: 55-59. 

MORTON, В., 1973, Some aspects of the biology 
and functional morphology of the organs of 
feeding and digestion of Limnoperna fortune! 
(Dunkner) (Bivalvia: Mytilacea). Malacologia, 12: 
265-281 . 

MORTON, В., 1974, Some aspects of the biology, 
population dinamics, and functional morphology 
of Musculista senhausia Benson (Bivalvia, Myti- 
lidae). Pacific Science, 28: 19-33. 

MORTON, В., 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. & M. s. GALVAO-BUENO, 1983, 
Anatomía funcional de Mytella charruana (d'Or- 
bigny, 1846) (Bivalvia: Mytilidae). Boletim de 
Zoología da Universidade de Sào Paulo, 6: 1 1S- 
MO. 

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 cirh. Proceedings of the 
Royal Society of London, B, 187: 83-91. 

PAEZ-OSUNA, F.; G. IZAGUIRRE-FIERRO, R. I. 
GODOY-MEZA, F. GONZALEZ-FARIAS & J. I. 
OZUNA-LOPEZ, 1988, Metales pesados en 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 . 

PAIVA AVELAR, W. E. P. & W. NARCHI, 1984a, 
Functional anatomy of Brachydontes solisianus 
(Orbigny, 1846) (Bivalvia: Mytilidae). Boletim de 
Zoología da Universidade de Sào Paulo 8: 215- 
237. 

PAIVA AVELAR, W. E. P. & W. NARCHI, 1984b, 
Anatomía funcional de Brachidontes darwin- 
ianus darwinianus (Orbigny, 1846) (Mollusca: Bi- 
valvia). Papéis Avulsos de Zoología, Sâo Paulo, 
35: 331-359. 

PIERCE, S. K., 1973, The rectum of "Modiolus" 
demissus (Dillwyn) (Bivalvia: Mytilidae); a due 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, 1 31 : 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. В., 1965, The structure and function of 
the stomach in bivalve molluscs. Journal of Zo- 
ology, 147: 156-184. 

REPRIETO, J. F. & 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 Lepante, Puntarenas, 
Costa Rica. Revista de Biología Tropical, 33: 59- 
60. 

SIBAJA, W. G., 1 986, Madurez sexual en el mejillón 



chora Mytella quayanensis Lamarck, 1819 (Bi- 
valvia: Mytilidae) del manglar en jicaral, Punta 
Arenas, Costa Rica. Revista de Biología 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., 1 970, 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 Zoología, 1: 
187-229. 

STUARDO, J. & R. 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 3era. 
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. & M. 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, К. 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, С M., 1948, Formation of siphons in 
Lamellibranchia. Nature, 1961: 198-199. 

YONGE, С M., 1953, The monomyahan condition 
in the Lamellibranchia. Transactions of the Royal 
Society of Edinburg, 62: 443-478. 

YONGE, С M., 1955, Adaptation to rock boring in 
Botula and Littioptiaga (Lamellibranchia, Myti- 
lidae) with a discussion on the evolution of this 



14 VILU\RROEL & STUARDO 

habit. Quarterly Journal of Microscopical Sei- tain Mytilacea. Transactions of the Royal Society 

ence, 96: 383-410. of Edinburg, 68: 21-43. 

YONGE, С M. & J. I. CAMPBELL, 1968, On the 
heteromyarian condition in the Bivalvia with spe- 
cial reference to Dreisena polymorphe and cer- Revised Ms. accepted 25 Aug. 1993 



MAIJ\COLOGIA, 1995, 36(1-2): 15-27 

IJ\BORATORY EXPERIMENTS ON THE INFLUENCE OF FOOD AVAIU\BILITY, 

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 Köln, Germany 

ABSTRACT 

(1) Two groups of the mussel Dreissena polymorphs were collected in March (El) and Sep- 
tember (E2). Each group consisted of individuals at different stages of the annual gonadal cycle 
(e.g. females: in El 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 El), 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 El -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 is common in marine species, is char- 



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 1 2°C at three dif- 
iferent locations in two central European lakes 
(Borcherding, 1991). In the upper hypolim- 



15 



16 



BORCHERDING 



TABLE 1. Mean daily level of energy for each mussel (Joule/day, see text) under the various conditions 
employed during the experiments. E1 - experiment with mussels before the spawning season, E2 = 
experiment with mussels after the spawning season, HFL = high food level, LFL = low food level. 



El 



E2 



food 


photo- 




temperature (°C) 




level 


period 


5 


12 


19 


HFL 


LD 16:8 


3.4 


13.3 


19.3 


LFL 


LD 16:8 


0.2 


0.5 


0.9 


LFL 


LD 8:16 


0.2 


0.5 


0.9 


HFL 


LD 8:16 


6.2 


13.6 


20.0 


LFL 


LD 8:16 


0.2 


0.6 


0.9 


LFL 


LD 16:8 


0.2 


0.6 


0.9 



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 Fijh- 
linger See at a depth of 2-3 m (FS-2m) before 
(30 March 1987-experiment El) 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 I) filled 
with deionized water containing 1 % sea wa- 
ter and calcium chloride (final concentration 
= 160 mg Ca/I). 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. 



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 |.il 02/day in winter to 
about 500 |il Og/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 



17 



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 цт) 
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, 
Münster). 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 (W) 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 El . 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 
|am) identified within the gonads, a mean 
oocyte diameter was computed for all sizes 
above 40 |im. 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, 1 983). 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, W, Gl and the corresponding SL 
were fitted to the allometric equation у = 
aSL"^; у is the predicted value for GV, W 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^ ""^; 
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% confi- 
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 
F^a^-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, 
1 981 : 587-588). The final ANOVA of each ex- 
periment was followed by multiple classifica- 



11 



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 
|am/d (range 0-9.4 |im/d) over three months. 
This rate was low compared to that of pop- 
ulations in natural environments (e.g. up to 75 
|im/d in the Ijsselmeer, Bji de Vaate; 1991 ; up 
to 60 |im/d in the River Rhine, Jantz & Neu- 
mann, 1992). Further, there was no tendency 
towards lower rates of shell growth in larger 
mussels (e.g. Bji de Vaate, 1991; Jantz & 
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, HFL, LD 
16:8, mortality = 38%), the reason for which 
was not apparent. 

Experiments with Mussels Before the 
Spawning Season (El) 

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°C were run. After 90 days, gonad volume 
(GV) and the gonad index (Gl) of mussels 
kept at 19°C were significantly lower than 
those of mussels at lower temperatures (Li- 
test, p < 0.05). Under low food conditions 
(LFL), both GV and the visceral sack volume 
(W) decreased as temperature was in- 
creased (Fig. 1). Under conditions of high 
food availability (HFL), GV and Gl were sig- 
nificantly higher at 12°C 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°C (U-test, p 
< 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- 




Ж 




40- 


♦ +1 


^\ 


^ 


20- 




LFL 




start 


after 90 days 






1 


1 


1 




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. 



(El ) 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 El this was also the case for the sex 
with p = 0.714 [GV] and p = 0.646 [Gl], re- 
spectively). Thus, the results of El 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 polymorphe: summary statistics for the two-way-ANOVA and MCA, for all data (n = 
96 for each group of values) in El, 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 




of F 


accounted for 


of F 


accounted for 


covariate (SL) 


p < 0.001 


36.4 


p < 0.001 


11.2 


temperature 


p < 0.001 


19.4 


p < 0.001 


27.0 


food level 


p = 0.397 


— 


p = 0.038 


2.9 


2-way interactions 


p = 0.172 


— 


p = 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°C 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 (27%), fol- 
lowed by SL (1 1 .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 1 9° С 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 photoperlod (not shown In Fig. 3). 

Since a significant negative correlation was 
found between OD and CON (Spearman 



20 



BORCHERDING 



X 

'S 



"^S^ 



00 ° 



e 



60 
40 
20 



e 




R 






^0 


и 




в 




3 




о 


?0 


> 




•a 




re 




С 

о 


1U 


&0 





-20- 



-40- 



-60 - 



-80- 






12 



'■»¡"M»*») 



...J.j 



i.j.*?. 



12 







5 




12 


40- 








• - 


- - - 


- - 




30- 








20- 










10 - 








..4*4.4 .-*,« 



t 



19 



f^ 



19 



I 



19 




1 

5 12 19 

temperature [°C] 

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 El (dotted lines). All 
values were adjusted for the independent factors 
and covahates given in Tables 2 and 3. 



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 ail the factors as well 
as the covahates on both these variables (p < 
0.001). In addition to the distinct influence of 
the covahates (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 expehment 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 1 8.6°C. Duhng the 90 
days of this experiment, the mussels were 
either maintained at this temperature or at 
reduced temperatures of 12°C and 5°C (Ta- 
ble 1). 

For GV and Gl the total variance of all mea- 
surements (n = 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 vah- 
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 (El). The 
maximum value at 12'C occurred for Gl un- 



GONAD DEVELOPMENT OF DREISSENA 



21 



<D 

a 

a 

• T-H 

ел 

CD 
N 

CD 
^— > 

о 
о 

О 



62 - 


n=ll 


54 - 


d=31.7 - 


^46- 


"l g=0.37 - 


0,38- 
¿^30 - 


h 


22- 


P^ : 


14 - 


^ 




) 20 40 




n=9 

Ld=31.0 
c=-15 

i_ 






' ГЦ î£rCSEî22.1i 








Ü3;:: 


;Шл 



f^íí f 



^1Л 

i 
г I 

щ 
I 

i 



^iitiiiii 




20 40 



ft t «»» Ï 
ft ft.« 
t 1 1 1 t-s 



Í J» > Í f « <- ? 
' > if ? >, 

if > t«i t i M 



й1л»*ты -л 



i rim 

tu ifUri 



t:;::;:.:::;:::;;;6 2-t 
i;:;:;:;:;::;:::;:;; 5 4 1 
■•:'!":--''-:"'';^: 4 6 
:;':•::, ■:";;;: 3 8 
I-;:,,:, •,::::; ЗОН 
:.'.^,.'^^..^'\í: 2 2 

тишш 1 4 









ГЗ 



d'=54.3 
c=-93 



-•"■ 1 




-. "i 


-; ::. ::':::!5 


- y///ij 




I n= 


1 


: d« 


=52.0 


c= 


-97 




— I 




iff t 
( f f t'f f 



n=l 
d=21.2 

H c=-37 



n=2 

d=25.9 

c=-49 



^ 



20 40 



n=3 



O 20 400 20 40 20 40 



Start 



HFL 



LFL 



HFL 



LFL 



HFL 



LFL 



5°C 



12°C 



19°C 



FIG. 3. Dreissena polymorpha (20 mm shell length): Relative frequency of oocytes in the oocyte size classes 
(y-axes show 4-|am 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, d = 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; с = 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 < 
0.001) and GV under HFL conditions (Fig. 4; 
three-way-ANOVA: p < 0.001, Table 4). The 
decrease in GV in mussels under LFL condi- 



tions at increasing temperatures also corre- 
sponds with the results of experiment El. 
Overall, there were distinct similarities be- 
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 El, 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 


p < 0.001 


21.8 


temperature 


p = 0.016 


11.6 


p = 0.001 


20.3 


food level 


p = 0.003 


9.6 


p = 0.040 


4.8 


2-way interactions 


p = 0.374 


— 


p = 0.420 


— 


multiple r^ 




0.464 




0.469 



and E2. Only food availability had a stronger 
effect in E2, especially at 5°C. 



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 in- 
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 in- 
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°C 
(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°C and 
12°C 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 1 9°C, 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°C than at 18^0, 
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 



^ 30 



s 20 



10 - 




after 90 days 



•? 



-- 15 



45 - 




1 1 




30 - 


+ 


*-^\ 


\jHFL 


15 - 




^"^^7 к ^ 


Л 




start 


after 90 days 


LFL 






1 1 


1 



10 - 



5 - 



+ 




after 90 days 



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) of the 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. ¡rradians, re- 
ported slightly higher gonad indices (the con- 
tribution of gonad to the body weight) at 
15°C and 20' С 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°C 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 (El: 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 is 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 ripe 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 El . Oocyte 
diameter, which is 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) 


gonad index (Gl) 






percentage 




percentage 






of total 




of total 




significance 


variation 


significance 


variation 




of F 


accounted for 


of F 


accounted for 


covahate (SL) 


p < 0.001 


48.3 


p < 0.001 


23.9 


temperature 


p < 0.001 


9.6 


p = 0.001 


13.0 


food level 


p < 0.001 


9.0 


p = 0.049 


2.9 


sex 


p = 0.051 


— 


p = 0.185 


— 


2-way interactions (t-f) 


p = 0.051 


— 


p = 0.428 


— 


2-way interactions (t-s) 


p = 0.893 


— 


p = 0.831 


— 


2-way interactions (f-s) 


p = 0.518 


— 


p = 0.669 


— 


3-way interactions 


p = 0.875 


— 


p = 0.867 


— 


multiple r^ 




0.672 




0.399 



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 & 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- 



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. C. virginica, Loosanoff & 
Davis, 1951; A. irradians, Sastry, 1966; /W. 
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''C, 
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 El (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 El, the oocytes did not mature at 
12° С 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 1 2°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''C 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, 1 9° C) 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. El), 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 El revealed that oocytes did not mature at 
low temperatures (5 0), 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 El). 
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 El 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. I 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. & 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., 
Estuahne processes, Vol. 1. Academic Press, 
New York. 

BAYNE, B. L., A. BUBEL, P. A. GABBOTT, D. R. 
LIVINGSTONE, D. M. LOWE & M. N. MOORE, 
1982, Glycogen utilisation and gametogenesis in 
My til us edulis L. Marine Biology Letters, 3: 89- 
105. 

BAYNE, B. L., D. L. HOLLAND, M. N. MOORE, D. 
M. LOWE & J. WIDDOWS, 1978, Further studies 
on the effects of stress in the adult on the eggs 
of Mytiius eduiis. Journal of the Marine Biological 
Association of the United Kingdom, 58: 825-841 . 

BAYNE, B. L & R. 0. NEWELL, 1983, Physiological 
energetics of marine mussels. Pp. 407-515, in: 



26 



BORCHERDING 



к. 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. Malacologie, 33: 31-42. 

BU DE VAATE, A., 1991, Distribution and aspects 
of population dynamics of the zebra mussel, 
Dreissena polymorpha (Pallas, 1771), in the lake 
Ijsselmeer 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. Inter- 
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 zeiora mussels. Pp. 87- 
99, in: D, NEUMANN & H. 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. & G. UHLIG, 1981, Caloric and car- 
bon values of marine and freshwater protozoa. 
Helgoländer Meeresuntersuchungen, 34: 401- 
412 

GABBOTT, P. A., 1975, Storage cycles in marine 
bivalve molluscs: an hypothesis concerning the 
relationship between glycogen and gametogen- 
esis. Pp. 191-211, in: н. barnes, ed., Proceed- 
ings, 9th European Symposium on Marine Biol- 
ogy University Press, Aberdeen, Scotland. 

GABBOTT, P. A. & B. 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- 
pérí mentale du cycle sexuel d'un mollusque bi- 
valve Cardium edule L. Analyse des populations, 
évolution de la gonade et action de quelques 
facteurs: nutrition, température, photopériode. 
Ph.D. thesis. University of Caen. 

HEBERT, P. D. N., B. W. MUNCASTER & 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, B. & D. NEUMANN, 1992, Shell growth and 
aspects of the population dynamics of Dreissena 
polymorpha in the River Rhine. Pp. 49-66, in: 
D. NEUMANN & H. 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. L. & H. 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. & 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. 1 1 3-292, in: A. с GiESE & j. s. pearse, 
eds.. Reproduction of marine invertebrates. Vol. 
V. Academic Press, New York. 

SASTRY, A. N. & N. J. BU\KE, 1971 , Regulation of 
gonad development in the bay scallop, Ae- 
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, 1 1 5: 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. 

STÁNCZYKOWSKA, 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- 
handbuch, Vol. 1. Gustav Fischer Verlag, Stutt- 
gart — New York. 

UNDERWOOD, A. J., 1 981 , Techniques of analysis 
of variance in experimental marine biology and 
ecology. Oceanography and Marine Biology, an 
Annual Revue, 19: 513-605 

WALZ, N., 1978, The energy balance of the fresh- 
water mussel Dreissena polymorpha Pallas in 



laboratory experiments and in lake Constance. 
II. Reproduction. Archiv für Hydrobiologie/Sup- 
plement, 55: 106-119. 
WIESER, W., 1986, Bioenergetik— Energietransfor- 
mationen bei Organismen. Thieme, Stuttgart, 
New York. 



Revised Ms accepted 25 January 1994 



MALACOLOGIA, 1995, 36(1-2): 29-41 

RELAXING TECHNIQUES FOR FRESHWATER MOLLUSCS: TRIALS FOR 
EVALUATION OF DIFFERENT METHODS 

R. Araujo, J. M. Remón, D. Moreno & M. A. Ramos 

Museo Nacional de Ciencias Naturales (CSIC), José Gutiérrez Abascal, 
2, 28006 Madrid, 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- 
phometries (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 ie\N 
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- 



stone et al. (1989) used different concentra- 
tions of three volatile products to reversibly 
anaesthetize specimens of Lymnaea stagna- 
lls (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 of the 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 



29 



30 



ARAUJO, REMON, MORENO & 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, 1 774), 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 (^%), magnesium chloride (7.5%i), 
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°C). Menthol was also 
tested at room temperature (20-28° C). The 
effect of "drowning" over the specimens was 
proved in 1 5 ml of water (1 00 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 '0 below zero (± 2°C), liquid nitrogen after 
ten seconds and hot water (60° C) 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 Bittiynia tentaculata 
and Pisidium amnicum (Müller, 1774) and 
one of Corbicula fluminea (Müller, 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 (Müller, 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 Horatia sturmi 
were sampled at Pilar del Mono fountain 
(Granada, Spain) in February 1992. Speci- 
mens of Melanopsis sp. were collected at the 
Moll 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'C 
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 is 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 
Haën. 

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 1 0-1 3°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 of the 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 6-1 0°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°C). 

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 of the 
foot visible outside the shell), and 1 = bad 



32 



ARAUJO, REMON, MORENO & 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% 


2.0% 


0.200% 1.000% 


0.10% 


0.50% 


1 .0% 


0.100% 0.500% 


0.05% 




0.5% 


0.050% 0.250% 








0.025% 0.125% 









1 .0% 1 .0% 



(*) 



n 

15 drops 
10 drops 
5 drops 



''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. In 
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 I) and after 24 hours (fix- 
ation II) 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 of the 
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 I and a 
fair fixation II. 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 


of. luisi 


Horatia sturmi 


PENTOBARBITAL 














0.400% 


3.3.4.1 


2.1.2.2 


3.2.4.4 


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 


4.1.4.3 


4.3.4.4 


4.3.4.4 


4.3.4.4 


0.050% 


3.2.4.4 


4.1.4.4 


3.3.4.4 


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.-.-.- 


3.1.4.4 


2.2.4.3 


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.-.-.- 


4.1.1.1 


1.-.-.- 


2.1.3.3 


2.1.3.3 


PHENOXYETHANOL 














1.00% 


1.-.-.- 


1.-.-.- 


1.-.-.- 


1.-.-.- 


2.2.3.3 


2.3.3.3 


0.50% 


1.-.-.- 


1.-.-.- 


1.-.-.- 


1.-.-.- 


3.2.3.3 


3.3.3.3 


URETHANE 














2.0% 


1.-.-.- 


1.-.-.- 


4.1.4.4 


3.4.3.2 


2.1.4.4 


3.3.4.3 


1 .0% 


1.-.-.- 


1.-.-.- 


4.1.4.2 


1.-.-.- 


2.1.3.3 


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 














D 


3.4.2.2 


2.1.2.2 


2.2.4.3 


1.-.-.- 


3.4.4.3 


4.3.4.4 


CLOVE OIL 














15 drops 


1.-.-.- 


2.1.2.2 


1.-.-.- 


1.-.-.- 


1.-.-.- 


1.-.-.- 


10 drops 


1.-.-.- 


1.-.-.- 


1.-.-.- 


1.-.-.- 


1.-.-.- 


1.-.-.- 


5 drops 


1.-.-.- 


1.-.-.- 


1.-.-.- 


1.-.-.- 


1.-.-.- 


1.-.-.- 


DIETHYLETHER 














(") 


4.3.2.1 












CHLOROFORM 


1.-.-.- 
1.-.-.- 












( ) 
DROWNING 1 












DROWNING 2 














2.1.3.3 













Results are the best obtained including repetitions. When narcotization is bad (value 1) no results of tinne nor fixations I and 

II 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 

I and fixation II 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 
I and fixation II, 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 II) and 
with menthol in 24 h (fair fixation II). 

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 




fluviátil is 


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 


2.1.3.1 


0.250% 


3.4.3.3 


4.1.4.3 


0.125% 


4.4.4.3 


4.1.4.3 


MS 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 


и R ETHANE 






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 


2.1.3.2 


METHOL CRIST 






(*) 


4.4.4.4 


3.4.4.1 


CLOVE OIL 






15 drops 


2.4.4.2 


2.4.3.1 


1 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 
h. 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° 0) 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 I and fixation II 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 I, 
although animals withdrew slightly after 24 h 
(fixation II). 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 II was not 
good with the former dilution. A subsequent 
repetition of tests did not yield good results. 

Very good narcotization and fair fixation II 
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 I and 
II were very poor. 

With 0.4% and 0.2 %o 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%o pentobarbital yielded good 
narcotization in only one specimen in 32 h. 
Fixations I and II 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. 



Pisidium 


Corbicula 


amnicum 


fluminea 


2.1.3.1 


3.1.3.3 


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 


2.1.4.2 


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 


2.1.3.3 


3.2.4.2 


1.-.-.- 


3.1.4.1 


3.1.4.3 


2.1.4.2 


1.-.-.- 



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 


MS 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 




(*) 




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 



1.-.-.- 

4.2.4.4 

3.4.3.3 

3.4.4.4 
3.3.4.4 



1.-.-.- 

1.-.-.- 

3.2.4.4 

1.-.-.- 
1.-.-.- 
1.-.-.- 



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. 



0.5% sodium pentobarbital yielded very 
good narcotization and good fixations I and II 
after 24 h. 

Urethane (2%) worked well after 24 h, with 
good fixation I and fair fixation II; 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, REMÓN, MORENO & RAMOS 



Pseudamnicola cf. luisi (Table 3) 

Optimal results for narcotization and fixa- 
tions I and II 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 I and II 
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 I and 
good fixation II. 

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 I and II with 
0.4%, 0.2%, 0.1% and 0.05% pentobarbital 
and menthol. 

Results were similar after 51 h with 0.1 25% 
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, 1 good, 5 fair and 6 with- 
drawn). 

Ancylus fluviatilis (Table 4) 

Best results for narcotization and fixations 
I and II 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%o 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 1 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 II. 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 I and good fixation II, 
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 1 00 
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 II 
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 I in 5 h, but very 
poor for fixation II. Repetitions of these tests 
yielded poor results in 72 h. 

Unio sp. (Table 5) 

Very good narcotization and fixations I and 
II were obtained after 24 h using 0.2% MS 
222, although after repeating the test, fixation 
II 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 I 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 II was very good). 

In 72 h the test with 0.5% urethane gave 
good relaxation and fixations I 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'X (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 I 
and II 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 II was only fair. With 



0.1% and 0.05% MS 222, narcotization was 
very good, time ranging between 48 and 56 
h, while fixation II was unsatisfactory in both 
tests. Initial results of narcotization and fixa- 
tions I 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 I and a poor fixation II. 

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 I and II. 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 1 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°C and 28°C. 

Corbicula fluminea (Table 5) 

The most satisfactory narcotic was 0.1% 
MS 222 in 130 and 100 h, fixations I and II 
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, REMON, MORENO & RAMOS 



100 h but fixation II 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 I, 
and good fixation II. 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 I but poor fixation II. For 72 h, fixation II 
was just fair. 

Menthol and 0.5% sodium pentobarbital 
were good narcotics after 48 h, being very 
good fixation I and good fixation II. When in- 
creasing action time of narcotics to 72 h, fix- 
ation II with menthol was very good, but re- 
sults with 0.5% sodium pentobarbital were 
unsatisfactory. 

Although repeated tests were not made, 
narcotization and fixation I 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 I, respec- 
tively, and good fixation II. 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 Bitliynia 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 (1 958) 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 С 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 optimal concentration needed varied. The 
minimum effective concentration of pento- 
barbital was 0.025% for Unio sp. and Pisid- 
ium 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 (Müller, 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, 1 981 , and 0.2% in 1 5 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 (M. A. R. and D. M.) with 
species of different genera {Lymnaea, Physa, 
Ancylus, Ferhssia, Gyraulus, Planorbarius, 
Potamopyrgus, Pseudamnicola, Mercuria, 
Horatia, Neohoratia, Belgrandia, Belgran- 
diella, and Theodoxus) successful results 



40 



ARAUJO, REMON, MORENO & 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 
Schalle, 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 it 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 1 958) 
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 Schalle, 
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. I. Camacho for their construc- 
tive criticism of the manuscript. B. Arano, B. 
Kelly and a anonymous reviewer improved 
the English. Thanks also to R. Sánchez and L. 
G. Alaejos for their help in the samples trips. 

This work received financtial support from 
the Project "Fauna Ibérica II" (SEUL 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- 
173. 

BERRY, E. G., 1943, The Amnicoiidae of Michigan: 
distribution, ecology and taxonomy. Miscella- 
neous Publications Museum of Zoology. Univer- 
sity of Michigan, 57: 68 pp., 9 pis. 

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 Pliysiology (C-Compara- 
tive Pliarmacology and Toxicology), 92: 39-44. 

GIUSTI, F. & 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. 0. JAGER, & A. WESTERVELD, 1964, 
A new anaesthetization technique for fresh water 
snails, tested on Lymnaea stagnalis. Malacolo- 
gia, 1: 331-338. 

LINCOLN, R. J. & J. G. SHEALS, 1985, Invertebrate 
animals: collection & preservation. British Mu- 
seum (Natural History), London: 150 pp. 

MCCRAW, B. 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: 1 1 5-1 1 7. 

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, N. W., K. ISARANKURA & B. J. SMITH, 
1965, Methods for narcotizing and anaesthetiz- 
ing gastropods. Malacologia, 2: 231-238. 

SMALDON, G. & 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 

U\ND-SNAIL COMMUNITY MORPHOLOGIES OF THE HIGHEST-DIVERSITY SITES 

OF MADAGASCAR, NORTH AMERICA, AND NEW ZEALAND, WITH 

RECOMMENDED ALTERNATIVES TO HEIGHT-DIAMETER PLOTS 

Kenneth С 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 outward displacement (Raupian parameters modified for 
mathematical 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 Resen/e [= 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- 



nombo, south of Farafangana, Flanarantsoa 
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 глт. 
While collecting and sorting, the author was 
impressed by the difference in shell-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 compahng land-snail community mor- 
phologies (Cain, 1977, 1978a, 1978b, 1981a, 
1981b; Cameron, 1988; Cameron & Cook, 
1989; Heller, 1987) seemed inadequate for 



43 



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 ¡mages 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 
1 990, 1 0:00 am-1 2:00 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 is 
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 



L7\ND-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- 
ity" (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 it 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 in- 
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; liiert, 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 



U\ND-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-1 1 9) 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 mm) are roughly equivalent 
throughout: NZL 15%, MDG 1^3%, 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 sfielter), 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 



EMBERTON 






MZL 



—J- 
20 



Shell Diameter (mm) 



—r- 

40 



70 



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 (Gyclophorldae), 



all of the presumably more ancient families of 
the Achatlnlda (sensu Nordsieck, 1986), and 
most of the presumably more ancient fami- 
lies of the most recent Heliclda. NZL is next, 
with Its complement of prosobranchs. Its 
dominance by Achatlnlda, and Its absence of 
Heliclda; and USA follows, with all of the 
most recent Heliclda (but also with the "prim- 
itive" Vertiglnidae). As in other geographically 
Isolated land-snail faunas (Gain, 1977, 
1978a, 1980; Peake, 1978; Gameron & Cook, 
1992), phylogenetic overlap among these 



U\ND-SNAIL COMMUNITY MORPHOLOGIES 

TABLE 1. A model to predict shell size distributions assuming pure natural selection 



49 





Physical Niches 




Niche-Filling Aid 




Site 


Minute Medium 


Giant 


Frost-free 


Drought-free 


Storm-free 


USA 
MDG 
NZL 


2 2 

2 2 

3 2 

Niche-Filling 
Aid 


2 
3 

1 


1 1 
3 2 

2 3 

Importance to Snails 


2 
1 
3 




Minutes 


Mediums 


Giants 






Frost-free 

Drought-free 

Storm-free 

Site 


1 
3 
3 


2 
2 
2 

Predicted Snails 


3 

1 
1 






Minutes 


Mediums 


Giants 






USA 
MDG 
NZL 


20 

24 
60 


16 
24 
32 


12 
36 
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. 





l_ot_8pfolee_ 
HBO нгЬ üSft 


Thaaa Morld- 
BUes Wide 


These 
Bites 


ВЬаве 

eoEld-Hide. 


Cyclophor-Liarei 


29 


4 





min-med 


min-med 


glob-tall 


Vf lat-xtall 


Vertiginidaa 








14 


mln 


min 


tall 


glob-tall 


Bubulinidae 


12 








min 


min-med 


tall-xtall 


tall-xtall*» 


Btreptaxldaa 


10 








min 


min-med 


tall-xtall 


vflat-xtall 


Aoavidaa 


10 








gnt 


med-gnt 


glob 


Vf lat-tall 


Punct-Charop-Diso 


4 


92 


10 


min-med 


min-med 


Vf lat-glob 


Vf lat-tall 


Hellxarlon-Eucon 


25 





5 


min-med 


min-med 


glob 


glob** 


Zonitidae 








29 


min-med 


min-med* 


vflat-glob 


vflat-glob 


Polygyrldae 


_й 


_o 


_M 


med 


med* 


(flat) 


vflat-glob 


Total 


90 


9в 


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 follov\/s Nordsiecl< (1986). Abbreviated family names: Cyclophoridae and Liareidae; Punctidae, Charop- 
idae, and Discidae; Helixahonidae and Euconulidae. Shell sizes: min = minute = diameter 0-5 mm; med = 
medium = 1 0-20 mm; 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 

.4 
.3 
.2 

.1 

.0 



EMBERTON 









08Л actual 














1 1 



.4 
.3 

.2 
.1 

.0 



OSA pradlotad by 
natural salaction 




DBA predicted by 
phylogenetic 
constraints 



minute 



nediuB 
Shell eise 



giant 



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 giant; and for 
USA 10.5 minute, 9.5 medium, and 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 









MDQ actual 





















MOO pradiotad by 
natural salaction 



HOG pradlctad by 
phyloganatlo 
oonstralnta 



minute 



madium 
Shall Slza 



giant 



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) is 
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 



NZL actual 



NZL predictad by 
natural Balacticn 



NZL pradiotad by 
phyloganatic 
constraints 



minute 



medium 
Shell Size 



-| 1 

giant 



FIG. 5. Shell-size distributions at Waipipi Reserve (= Jones Bush), North Island, New Zealand, as repre- 
sented by Manai<au-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 internnediate 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- 



U\ND-SNAIL COMMUNITY MORPHOLOGIES 



53 




-1 1 1 1 




MDG 




MZL 



0.0 



0.4 



0.8 



1.6 2.0 2.4 2.8 

Shell Hclght/Dlaffletar 



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- 
nnon 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 







Physical 


Niches 






VFIat 


Globose 


Tall 


XTall 


Site 










USA 


3 


2 


2 


2 


MDG 


1 


2 


3 


3 


NZL 


1 


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 
Predicted Snails 


3 


3 


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, xtall = 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 1 2.8 very flat, 1 7.8 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 of the 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 1 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 








OSA actual 


OS 










f 






1 




00 

10 
05 


1 


1 


1 






1 1 1 

USA predicted by 
natural Beleotlon 












00 
10 








1 


1 


1 




1 


t 

DBA predicted by 
phylogenetic constraints 


05 
00 












very 
flat 




glo- 
bose 


Shell 


1 

tall 

Shape 


1 1 1 
extremely 
tall 



FIG. 7. Shell-shape distributions at Hubricht's (unpublished) site on Pine Mountain, Kentucl<y, 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. 1 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 
Г2 = radii of the smaller and larger circles (Fig. 
lOd), 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 = (jT/w) (Гз^ - r/) 

D = d/w 

T = t/w 

Table 3 provides all these measurements and 
calculated variables for Figure 10's four 
shells. 



Figure lOe 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 с 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 с and d. 

Figure lOf 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 с 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 p 



MDO aotual 



.15 
.10 

.05 
.00 



MDG pradiotad by 
natural aalaotlon 



,20 
.15 
.10 

.05 
.00 



very 
flat 



glo- 
bose 



MDG predicted by 
pbylogenatio constralnta 



tall 
Shell Shape 



extremely 
tall 



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. Ramirez, pers. 
comm.), Costa Rica (C. 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, 
1 983; Solem, 1 984, 1 990; Murray et al., 1 988; 
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 



U\ND-SNAIL COMMUNITY MORPHOLOGIES 



57 



10 
,05 
. 00 



NZL aotual 



10 
,05 
, 00 



NZL predlated by 
natural selection 



10 



NZL predicted by 
phylogenetic constraints 



,00 



very 
flat 



glo- 
bose 



tall 
Shell Shape 



n 1 

extremely 
tall 



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 a phylogenetically 
depauperate fauna (Peake, 1978; Cameron & 
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 




ht* 




dm 



3 units 
I I I I 






O'O 



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 Г2 = 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 = [л /w] [r2^ — r^^]), outward displacement (D = d/w), and downward translation 
(T = t/w). 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 & Cook, 
1992), speciation within clades with short- 
term phylogenetic constraints (Cameron & 
Cook, 1992), interspecific competition (Cain, 
1977, ff.), direct environmental induction 
(Gould, 1970), and constructional con- 
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 



U\ND-SNAIL COMMUNITY MORPHOLOGIES 



59 



TABLE 3. Measurements and calculated variables for the four hypothetical shell 
x-rays shown in Fig. 10 







Shell 






a 


b 


с 


d 


Cainian 










measurement: ht. 


7.5 


7.5 


2.3 


2.3 


measurement: diam. 


3.0 


3.0 


7.0 


7.0 


calculation: ht./diam. 


2.5 


2.5 


0.3 


0.3 


Raupian 










measurement: w 


2 


7 


4 


2 


measurement: r^ 


0.3 


0.15 


0.2 


0.4 


measurement: Г2 


1.0 


0.45 


0.4 


0.9 


measurement: d 


0.5 


0.5 


1.8 


1.4 


measurement: t 


4.2 


4.1 


1.0 


1.0 


calculation: W 


1.4 


0.1 


0.1 


1.0 


calculation: D 


0.3 


0.1 


0.5 


0.7 


calculation: T 


2.1 


0.6 


0.3 


0.5 



among USA, MDG, 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- 



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. A 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 is 
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. 1 0b- 
d). Mapping community morphologies by 
Raup's W, D, and T (Fig. lOf) 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 

I 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. 



U\ND-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. 

CAIN, A. J., 1977, Variation in the spire index of 
some coiled gastropod shells, and its evolu- 
tionary significance. Philosophical Transactions 
of the Royal Society of London, (B) 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. 

CAIN, A. J., 1981a, Possible ecological signifi- 
cance of variation in shape of Cerion shells with 
age. Journal of Conchology, 30: 305-315. 

CAIN, A. J., 1981b, Variation in shell shape and 
size of helicid snails in relation to other pulmo- 
nales of the Palearctic region. Malacologia, 21: 
149-176. 

CAIN, A. J. & R. H. COWIE, 1978, Activity of dif- 
ferent species of land-snail on surfaces of differ- 
ent inclinations. Journal of Conchology, 29: 267- 
272. 

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, 1 6: 1 57- 
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. & 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, K. C, 1986, The evolution of multiple 
sympatric homeomorphy among three genera of 
land snails. Ph.D. Thesis, University of Chicago. 

EMBERTON, K. 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, K. 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, K. C, 1 991 a. The genitalic, allozymic, 
and conchological evolution of the Mesodontini, 
trib. nov. (Gastropoda: Pulmonata: Polygyridae). 
Malacologia, 33: 71-178. 

EMBERTON, K. C, 1991b, Polygyhd 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, Polygynd 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: A.c. 
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, K. С & S. TILLER, 1 994, 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., 1 968, Ontogeny and the explanation 
of form: an allomethc 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 Paleontológica! Con- 
vention, Septennber 1969, Part E: 486-521. 

GOULD, S. J-, 1971, D'Arcy Thompson and the 
science of form. New Literary History, 2: 229- 
258. 

GOULD, S. 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, S. J. & D. S. 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, 1 981 : 
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. M. C, 1989, Control of gastro- 
pod shell shape: the role of the preceding whorl. 
Journal of Theoretical Biology, 140: 431-444. 

HUTCHINSON, J. M. 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, & P. MORAT, 
1974, Flore at Végétation de Madagascar J. 
Cramer in der Gantner Verlag Kommanditgesell- 
schaft, FL-9490, Vaduz, Germany, 687 pp. 

KOHN, A. J. & A. С. RIGGS, 1 975, Morphometry of 
the Conus shell. Systematic Zoology, 24: 346- 
359. 

MURRAY, J., E. MURRAY, M. S. JOHNSON & 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 für 
Molluskenkunde, 1 1 7: 93-1 1 6. 

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. В., 1979, Nevi/ Zealand Mollusca. 
Marine, land and freshwater shells. Collins, 
Auckland, 500 pp; 82 pis. 

RAUP, D. M., 1 961 , 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. schöpf, ed.. Mod- 
els in Paleobiology. Freeman, Cooper & Com- 
pany, San Francisco. 

RAUP, D. M. & R. R. GRAUS, 1972, General equa- 
tions for volume and surface area of a 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 & 
A. 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. & F. M. CLIMO, 1985, Structure and 
habitat correlations of sympatric New Zealand 
land snail species. Malacologia, 26: 1-30. 

SOLEM, A., F. M. CLIMO & 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, 
Tübingen: 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. & B. С CLARKE, 1983, Lutte bi- 
ologique et destruction du patrimoine génétique: 
le cas des mollusques gastéropodes pulmones 
dans les territoires français 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- 
brüder Bornträger, Berlin. 



Revised Ms accepted 9 February 1 994 



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 & Boss (1989) for subclasses Proso- 
branchia and Gymnomorpha and Nordsieck (1986) for subclass Pulmonata: order Stylommatophora. 



Genus & Higher Classification 



Species 
Number 



Measurements (mm) 



Height 



Diam. 



Ht./Diam. 



Subclass PROSOBRANCHIA 
Order MESOGASTROPODA 
Superfamily CYCLOPHOROIDEA 
Cyclophoridae 
Boucardicus 



Cyathopoma 
Hainesia 
Superfamily LITTORINOIDEA 
Pomatiasidae 
Tropidophora 



Superfamily RISSOIDEA 
Assimineidae 
Omphalotropis 

Subclass PULMONATA: Order STYLOMMATOPHORA 
Superfamily BULIMINOIDEA 
Buliminidae (= Enidae) 
Rachis 
Suborder SIGMURETHRA 
Infraorder ACHATINI DA 
Superfamily ACHATINOIDEA 
Subulinidae 
'"Subulina" 



Superfamily STREPTAXOIDEA 
Streptaxidae: Streptaxinae 
"Edentulina" 



Gulella 
Streptaxidae: Enneinae 
Streptostele 



1 


9.1 


8.8 


1.03 


2 


5.3 


5.0 


1.06 


3 


2.9 


3.2 


0.91 


4 


3.2 


2.9 


1.10 


5 


3.8 


2.3 


1.65 


6 


2.8 


2.5 


1.12 


7 


3.0 


2.9 


1.03 


8 


2.5 


1.6 


1.56 


9 


2.3 


2.0 


1.15 


10 


3.5 


2.9 


1.21 


11 


1.6 


1.2 


1.33 


12 


1.9 


1.2 


1.58 


13 


3.1 


2.7 


1.15 


1 


2.0 


2.2 


0.91 


1 


31.0 


13.5 


2.30 


1 


16.0 


16.0 


1.00 


2 


30.0 


29.3 


1.02 


3 


24.0 


29.0 


0.83 


1 


3.9 


2.9 


1.34 


2 


5.6 


3.7 


1.51 



11.4 



7.5 



1.52 



1 


3.5 


1.8 


1.94 


2 


2.7 


1.3 


2.08 


3 


4.9 


1.8 


2.72 


4 


2.9 


1.2 


2.42 


5 


5.1 


2.0 


2.55 


6 


13.6 


3.7 


3.68 


1 


7.2 


3.4 


2.12 


2 


5.2 


2.3 


2.26 


3 


4.8 


2.4 


2.00 


1 


2.9 


1.5 


1.93 



25.2 



7.1 



3.55 



(continued) 



64 



EMBERTON 



APPENDIX 1. {Continued) 








Genus & Higher Classification 


Species 
Number 


Measurements (mm) 
Height Diam. 


Ht./Dlam. 



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 



1 


19.0 


46.0 


0.41 


2 


17.3 


34.5 


0.50 


1 


32.5 


57.0 


0.57 


2 


46.7 


70.4 


0.66 


3 


80.0 


65.0 


1.23 


1 


5.7 


9.0 


0.63 


2 


4.3 


8.5 


0.51 



3.6 



3.2 



1.12 



1 


3.6 


6.7 


0.54 


2 


1.7 


2.3 


0.74 


3 


1.6 


2.3 


0.70 


4 


2.3 


3.6 


0.64 


5 


2.4 


3.9 


0.62 


1 


9.3 


17.4 


0.53 


2 


13.0 


26.2 


0.50 


3 


11.0 


18.0 


0.61 


1 


5.6 


8.3 


0.68 


2 


2.3 


3.4 


0.68 


1 


4.1 


3.7 


1.11 


2 


3.3 


4.8 


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 venustas and Pallifera secreta. 
Classification as in Appendix 1, except that Riedel (1980) was followed for the Zonitidae. Polygyhd 
measurements were from Pine Mountain specimens, all others were from the literature for geo- 
graphically and ecologically relevant specimens. 







Measurements (mm) 




Genus & Higher Classification 


Species 


Height 


Diam. 


Ht./Diam. 


Subclass PROSOBRANCHIA 










Order ARCHAEOGASTROPODA 










Superfamily HELICINOIDEA 
Helicinidae 










Hendersonia 


occulta 


4.0 


6.6 


0.61 


Order MESOGASTROPODA 










Superfamily RISSOIDEA 
Hydrobiidae 
Pomatiopsis 
Subclass PULMONATA: Order ARCHAEOPULMONATA 


lapidaria 


6.3 


3.2 


1.97 


Suborder ELLOBIOIDEA 










Ellobiidae 










Carychium 
С 


clappi 
nannodes 


1.9 

1.4 


0.8 
0.6 


2.38 
2.33 


Subclass PULMONATA: Order STYLOMMATOPHORA 










Suborder ORTHURETHRA 










Superfamily COCHLICOPOIDEA 
Cochlicopidae 
Cionella 


morseana 


7.1 


2.3 


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 


1.0 


1.60 


V. 


clappi 


1.5 


0.8 


1.88 


Columella 


simplex 


2.2 


1.4 


1.57 


Gastrocopta 


pentodon 


1.8 


1.1 


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 


oval is 


25.0 


13.5 


1.85 


Infraorder HELICIDA 










Superfamily HELIXARIONOIDEA 










Euconulidae 










Euconulus 


fulvus 


2.4 


3.1 


0.77 


Guppya 


sterkii 


0.8 


1.2 


0.67 


Superfamily VITRINOIDEA 










Zonitidae: Gastrodontinae 










Gastrodonta 


interna 


5.0 


7.4 


0.68 


Ven tri den s 


collisella 


7.2 


8.7 


0.83 


Striatura 


meridionalis 


1.0 


1.7 


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 


5.6 


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 


7.1 


11.9 


0.60 


Glyphyalinia 


cumberlandiana 


1.2 


3.0 


0.40 


G. 


rimula 


3.4 


7.7 


0.44 


Superfamily POLYGYROIDEA 










Polygyridae: Triodopsinae: Triodopsini 










Neohelix 


albolabris 


21.5 


31.7 


0.68 


Xolotrema 


denotata 


8.8 


16.3 


0.54 


Triodopsis 


tri den tata 


9.6 


15.4 


0.62 


T. 


vu 1 gata 


8.3 


14.3 


0.58 


Polygyridae: Polygyrinae 










Allogona 


profunda 


15.0 


28.6 


0.52 


Stenotrema 


edvardsi 


5.6 


7.8 


0.72 


S. 


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 


zaietus 


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 hochsteteh 

Cytora cytora 

С torquilla 

Subclass PULMONATA: Order 
STYLOMMATOPHORA 
Suborder ORTHURETHRA 
Superfamily ACHATINELLOIDEA 
Achatinellidae 
Lamellidea novoseelandica 

Suborder SIGMURETHRA 
Infraorder ACHATINIDA 
Superfamily RHYTIDOIDEA 
Rhytididae 
Delos coresia 

D. jeffreysiana 

Rhytida greenwood! 

Superfamily PUNCTOIDEA 
Charopidae 



Caviella 

С 

Mocella 

M. 

"Charapa" 

"С." 

"С." 

"С." 

"С." 

Fectola 

F. 



buccinella 

roseveari* 

eta 

äff. maculata 

pseudanguicula 

chrysaugeia 

äff. pseudanguicula 1 

fuscosa 

pilsbryi 

mira 

unidentata 



Genus & Higher Classification Species 




F. 


infecta 




Huanodon 


pseudoleiodon 




H. 


hectori 




Allodiscus 


urqutiarti 




A. 


äff. granum 




Geminoropa 


cookiana 




Serpho 


l<ivi 




Flammulina 


perdita 




F. 


Chiron 




"Thalassohelix" 


ziczac 




Suteria 


ide 




Phenacohelix 


giveni 




P. 


pilula 




P. 


n. sp. 1 




Therasiella 


neozelandica 




T. 


serrata 




T. 


äff. neozelandica 




T. 


celinda 




Punctidae 






"Laoma" 


mariae 




"L" 


marina* 




"L" 


äff. marina 1 




"L" 


leimonias 




"Phrixgnathus" 


erigone 




"P." 


ariel 




"P." 


elaiodes 




"P." 


moellendorffi 




"P." 


conella 




"P." 


n. sp. 59 




"P." 


poecilosticta 




^^Paralaoma" 


n. sp. 38 




"P." 


n. sp. 29 




"P." 


lateumbilicata 




"P." 


n. sp. 1 




"P." 


n. sp. 8 




"P." 


äff. n. sp. 33 




"P." 


n. sp. 40* 




"P." 


serratocostata 





* collected by Roscoe and Haziewood in 1977 but not 
included in Appendix ЗА of Solem et al., (1981: 462). 



MAU\COLOGIA, 1995, 36(1-2): 67-77 

DISTRIBUTIONAL DIFFERENCES AMONG ACAVID LAND SNAILS 

AROUND ANTAIJ\HA, MADAGASCAR: INFERRED CAUSES AND 

DANGERS OF EXTINCTION 

Kenneth С 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. lamarei nor Helicoplianta 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 
moreleti southward along the coast and Ampelita julii northward), by climatic exclusion and 
coastal land clearing {Ampelita xystera), and by local speciation events {A. 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, 
1 990, 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 Hellcophanta, Am- 
pelita, and Clavator based on a new data set 
of 71 informative allozyme characters, and 



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- 



67 



68 



EMBERTON 




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 I 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 & 
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). Ampelltajulii's 
presumed sister species, A. 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, 
northeastem Madagascar (Fig. 1). All are to the same size scale; Cm is 70.0 mm in height. AI = Лтре// a 
(Ampelita) lamarei (two shells showing variation, station 205, Academy of Natural Sciences of Philade phia 
ANéP] catalog number 391391), Aj = A. (Eurystyla) julii (206, ANSP 391399), As = A fjA°i^'^f;4<248^ 
ANSP 391 41 1 ), Ax = A. (Xysters) xysters (206, ANSP 391 429), Af = A. (X.) fulguráis (205. ANSP 391 434), Cm 
= CIsvator moreleti (206, ANSP 391452), Ha = Helicophsnts smphibulima (204, ANSP 391482). The phy- 
logeny is based on allozymes for Ha, Cm, AI, Aj, and Ax (Emberton, in review b); and on shell similarity for 
As and Af (Emberton, 1990). 



глоге of collections 3-1 2 km from the coast, 
but virtually absent from both coastal and 
more inland stations; latitudinal effects were 



not statistically significant, despite the fact 
that no A. xystera were collected from the 
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 





Lat 


Inl 


Elv 










Ampelita 










Clavator 


Helicophanta 




Sta 


lamarei 


J 


ulii 


soulaiana 


xystera 


fulgurata 


moreleti 


am 


ohibulima 




# 


L 


D 


L 


D 


L 


D 


L 


D 


L 


D 


L 


D 


L 


D 


Total 


216 




1 


2 





29 





























13 





3 


45 


250 




1 


3 
































11 


26 





22 


59 


244 




2 


2 









































3 


3 


245 




2 


2 





8 











5 





28 

















2 


43 


248 




2 


2 





7 











5 





8 

















5 


25 


249 




2 


2 





4 











6 





9 














1 


72 


92 


246 




2 


3 





4 











7 





12 

















6 


29 


247 




2 


3 





2 






































2 


223 




3 


2 









































21 


21 


225 




3 


2 





4 





1 
































5 


226 




3 


2 





10 






































10 


229 




3 


2 





2 



































4 


6 


232 




3 


2 





1 





1 





























8 


10 


233 




3 


2 





2 





2 





























25 


29 


234 




3 


2 





12 





6 


























1 


97 


116 


235 




3 


2 





1 



































8 


9 


236 




3 


2 





7 





2 





























1 


10 


237 




3 


2 


1 


16 





5 





























43 


65 


238 




3 


2 





16 



































1 


17 


239 




3 


2 





6 

















12 




















18 


240 




3 


2 





89 



































41 


130 


241 




3 


2 





41 






































41 


243 




3 


2 





2 






































2 


218 




3 


3 





6 



































1 


7 


224 




3 


3 





8 



































100 


108 


200 


2 


1 


































7 


2 








9 


252 


2 


1 




2 









































2 


201 


2 


2 

























1 








1 


1 








3 


206 


2 


2 







224 





202 








42 


614 








97 


1118 








2297 


207 


2 


2 













2 











60 








17 


330 








409 


253 


2 


2 

























46 








40 


23 








109 


204 


2 


3 


3 


6 


60 
































1 


299 


366 


205 


2 


3 


3 


3 


92 


2 


53 

















65 





1 





50 


266 


208 


3 


1 


3 





5 





2 
































7 


210 


3 


2 


1 





1 



































1 


2 


211 


3 


2 


1 





1 





14 





























21 


36 


214 


3 


2 


1 









































8 


8 


215 


3 


2 


2 





2 





4 





























1 


7 


212 


3 


3 


1 





7 





6 





























10 


23 


Total 








12 


669 


2 


300 





23 


42 


790 





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 C. moreleti in 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*lnl 


Unexpl. 


A. lamarei 


0.01 


0.09 


0.04 


0.85 


A. julii 


0.35"* 


0.02 


0.01 


0.63 


A. soulaiana 


0.08 


0.08 


0.23* 


0.61 


A. xystera 


0.03 


0.16* 


0.07 


0.74 


A. fulgurata 


0.13* 


0.11* 


0.29** 


0.48 


C. moreleti 


0.23*** 


0.15** 


0.30*** 


0.32 


H. amphibulima 


0.03 


0.07 


0.07 


0.82 


A. (Eurystyla) 


0.26** 


0.01 


0.05 


0.68 


A. (Xystera) 


0.05 


0.13 


0.06 


0.76 



* p < 0.05, ** p < 0.01 , '** p < 0.001 (ANOVAs, df 2, 30 or 
[interactions] 4, 30). 



TABLE 3. 
ble 2 



Cell means from the ANOVAs of Ta- 



Cell Sizes 



Inland 



1 
1 2 

Lat 2 2 

3 1 

Ampelita (A.) lamarei 



Lat 



0.32 
0.50 
0.71 



A. (Eurystyla) julii 

1 0.00 

Lat 2 0.00 

3 0.29 
A. (E.) soulaiana 

1 0.00 

Lat 2 0.00 

3 0.00 
A. (Xystera) xystera 

1 0.00 

Lat 2 0.00 

3 0.00 
A. (X.) fulgurata 

1 0.00 

Lat 2 0.00 

3 0.00 
Clavator moreleti 

1 0.46 

Lat 2 0.50 

3 0.00 
Helicophanta amphibulima 

1 0.22 

Lat 2 0.00 

3 0.00 



2 

6 

4 
4 

0.28 
0.02 
0.20 

0.00 
0.02 
0.24 

0.10 
0.00 
0.00 

0.25 
0.30 
0.00 

0.00 
0.00 
0.00 

0.00 
0.66 
0.00 

0.37 
0.00 
0.56 



3 

17 
2 

1 

0.49 
0.27 
0.30 

0.04 
0.10 
0.26 

0.00 
0.00 
0.00 

0.04 
0.00 
0.00 

0.00 
0.12 
0.00 

0.00 
0.00 
0.00 

0.43 
0.50 
0.44 



coast) along the lower Ankavanana River val- 
ley (< 12 km frorn 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: 



Elevation 



Inland Distance 

2 

1 

5 

15 



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 






FIG 3 Distribution maps based on this study and Fischer-Piette et al. (1973). AI = Ampeita (Ampelita) 
lamarei Ai = A. (Eurystyla) julii (dots), As = A. (E.) soulaiana (stars), Ax = A. (Xystera) xystera (dots), At = A. 
(X.) fulgurata (stars), Cm = Clavator moreleti (open circle = dubious record), Ha = Helicophanta amphibu- 
lima. 



74 



EMBERTON 



preted here as spurious (Fig. 3, open cir- 
cle). 

Ampelita julii's 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 localités along a short stretch 
of the Andempona River valley, where it is 
closely bounded both upstream and down- 
stream by A. 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 amphlbulima 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. amphlbulima 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 С 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. julii 
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. julii, 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 is 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 xysíera'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 julii), by climatic exclusion and coastal 
land clearing {Ampelita xystera), and by local 
speciation events {A. soulaiana and possibly 
A. fulgurata). This area is unprotected and is 
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 Andhamamonjy 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. 
I 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. с 
VAN BRUGGAN & 8. 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., M. 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. & N. CARREAU DE LOU- 



76 



EMBERTON 



BRESSE, 1965, Mollusques terrestres de Mada- 
gascar Famille Acavidae. Journal de Conchylio- 
logie. Paris, 104: 129-160. 

FISCHER-PIETTE, E. & F. SALVAT, 1963, Mol- 
lusques terrestres de Madagascar, genre Clava- 
tor. Journal de Conchyliologie, 1 03: 53-74. 

GREEN, G. M. & 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. & 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, R. R. & F. J. ROHLF, 1969, Biometry. W. 
H. Freeman, San Francisco, 776 pp. 

VIETTE, P., 1 991 , 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 I 

200. 14.52.05S 50.14.40E, elev. approx. 
1 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. 14.51.40S 50.13.40E, elev. approx. 
1 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 49.59.1 OE, 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 m, 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. 1 4.59.1 5S 50.1 2. 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. OOE, 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 50.1 5.1 OE, 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. 14.37.35S 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 m, vicinity of Mangatsahatsa (1 .5 km S 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 



77 



jarivo (approx. 40 km NNW of Antalaha), par- 
tially cleared forest, 19 Oct. 1990, approx 4 
person-hours. 

232. 14.36.05S 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. 14.36.05S 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. 14.36.05S 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. 14.36.05S 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. 14.36.05S 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. 14.36.30S 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. 14.36.30S 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. 14.36.30S 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. 14.36.40S 50.07.1 5E (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. 14.36.40S 50.07.1 5E (Ambinanifaho), 
elev. approx. 30 m, region of Ambinanifaho 
(approx. 36 km NNW of Antalaha), virgin for- 
est, 18 Oct. 1990, >20 person-hours. 

246. 14.38.00S 50.07.25E, elev. approx. 
70 m, approx. 3 km S of Ambinanifaho (ap- 
prox. 36 km NNW of Antalaha), partially 
cleared forest, 19 Oct. 1990, approx. 4 per- 
son-hours. 

247. 14.36.30S 50.09. OOE, 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. 14.37.20S 50.08.1 OE, 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. 14.37.20S 50.08.1 OE, 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. 14.37.20S 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. IOS 50.13.20E, elev. approx. 5 
m, 1 km E of Tampolo, near coast (approx. 20 
km N of Antalaha), on trees near shore. 18 
Oct. 1990, approx. 8 person-hours. 

253. 14.51.25S 50.13. OOE, 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. 



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 & Jacques Daguzan 

Laboratoire de Zoologie et Ecoptiysiologie (L. A. INRA), Université de Rennes I, Campus de 
Beaulieu, Av. du Général Ledere, 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°C. All reproduction parameters were affected by temper- 
ature. Planorbis planorbis began to lay eggs from 10°C, whereas P. corneus reproduced from 
15°C. 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 nnany papers 
(Cole, 1925; Chemin & Michelson, 1957; van 
der Schalle & Berry, 1973; Aboul-Ela & Bed- 
dlny, 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 & 
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 is 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; 



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, 1 0, 1 5, 20 and 25°C. In the case 
of P. corneus, 374 individuals were reared as 
five sets at 5°C and 1 0°C, and four sets at 1 5, 
20 and 25° С 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 1 50 
ml of pond water per individual, and they 
were fed with fresh lettuce ad libitum in a 
12/12 h light/dark photoperiod. 

Sexual maturity was considered acquired 
when the first egg mass was observed. Every 



79 



80 



COSTIL & 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. corneas reared at 
20^0 and 25 C, but only at fixed dates in the 
other cases. 

At each temperature, the following param- 
eters were computed: 

mean age (AJ and, mean (0^), minimum (0^^,^) 
and maximum (D^ax) shell diameters at onset 
of sexual maturity; 

reproduction period duration, T^ep (in weeks); 

reproduction period duration in relation to max- 
imum life span, Trep (in %); 

mean numbers of egg capsules (NJ and eggs 
(Ng) per snail alive per two weeks; 

mean number of eggs per egg capsule (N^/c); 

fecundity of snails (Fee): cumulative number of 
eggs laid per snail throughout the life span; the 
fecundity per reproduction week is also calcu- 
lated (РесЯ.ер); 

proportion of capsules without eggs (in relation 
to NJ (in %); 

proportion of capsules containing one or more 
eggs without egg cells (in relation to NJ (CJ 
(in %); 

proportion of eggs without egg cells (in relation 
to NJ (EJ (in %); 

proportion of eggs including two or more egg 
cells (in relation to NJ (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, 1 945). 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 N to age N + 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°C, whereas P. 
planorbis reproduced from Ю'С (Table 1). 
The higher the temperature, the faster the 
maturation was: 15th week at 25°C and 49th 
week at IS'' 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: N = 7, i = -0.617, 
p = 0.05), and positively correlated with max- 
imum longevity (i = 0.905, p = 0.04). At 15°C, 
P. corneus reproduced for 1 75 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° C, 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°C, 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: т = 
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^X) 
(Fig. 2). Nevertheless, rhythms of 56 weeks 
(10°C) and 58 weeks (15°C) were found by 



REPRODUCTION IN PU\NORBIDS 



81 



Ш 


^ 


ю 


о 


■о 


с 




•Э 


го 




"D 


о. 


С 


00 


го 


-g 


w 


о 



о 




с 




та 








Q. 




.00 


3 


-Q 


Oí 


Í-. 


с 


О 


с 


С 


о 


CD 


о 






5: 


со 



о 



о 



о 



о 



Ü 



Ü 



о 



Е .-^ 



Ф ^ 



^ О 



со N- Ц-) 

СО N-' -ñt СП Т-' 

(М со CD 



CD о -^ 

"* CJ5 1Л СО N-' 

C\J ■* СО 



^- C\J т- СО CJ) 
^ "* о CD 

т- C\J 



<J) CD 00 

О) CD Ln -1- cri 

•■- t^ со r^ 

■r- C\J 



^eb .E 



о '-г 



p -Ь о ^ о 
~ го -^ го 






г^о 



ф ф 



О) '— . 

с Ф с _ 
го § го Q 



goo ^ 



^^'го 

(D Ф 

О-'ГО 
Ф — 

>- ^ 

■t; W 
> О. 
Ф ГО 
О) о 

о ^ 

-У О) 

ф 



13 



^ Е 



3 о 



E.i 



о Е о 

^^ ш ф го ф 

^ 2 СП 2 ОС 



s -Q 



с — W 

го с/3 1- 

о. ф Ф 

ел 3 Q- 

ф (/3 -i£ -¡-J 

-t CL Ф - 

— ГО ф i5 

^ " 5 ^ 

±3 оз с оз 

4^ ОЗ о Ö3 

3 Ф -^ Ф 

о ч- о > 

лоз 
оз s_ тз 
3 О) о 

9 ^ Q.- 



<^ -1- 



3 -^ Э "- Z! 



ф 
о. 

_ГО 03 

-гт- со 5 
ф оз с 

¡- ч- о 

го о 3 

(Л Ф Р 

ф -g Q- 
^ Е Ф 

— D ■- 



^ +1 
со о 



CD +1 

■^ со 



CD +1 

" оз 



■1 — hl 

CJ3 



г- +1 

C\J C\J 



ф 


о.^^ 


1_ 


ел ел 


ф 


оз 


ел 


ф ^ 


оз 




ся 


о ф 


ф 




о 


Ф го 




Е р 


ф 


3 s 


JD (Л 


с л 


Е +1 


1" 


3 _ 


с Ü 


If 


с Ф 


го 2 


X го 


ш ^ 


го ф 



82 



COSTIL & DAGUZAN 



Ne 

lu- 

9- 
8" 
7- 
6- 
5- 
4- 

3- 

2 - 
1 - 
O 



Nc 
Ne/c 



Ne/c 



Nc 

10- 



' ! ' ' ! ,' ' J Vi'' ' '',' 

г" ' ' í I L 1\Л 'il ''H 






Ne 


TT 


Ют 


20 


9- 


18 


8- 


16 


7- 


14 


6- 


|Т 




10 


5- 


8 


4- 


6 


3- 


4 


т - 



20 40 60 80 100 120 140 160 180 200 220 240 

Age (weeks) 



20°С 



Ne/c 

г 24 




Nc 
Ne/c 



00 120 140 

Age (weeks 

25°C 



Ne/c 

4 
"22 
20 



16 
14 
12 

tío 

8 
6 
4 




50 60 70 

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°C and 25°C, the number of eggs per cap- 
sule, lovy/ 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°C 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). 




20 40 60 



100 120 140 160 

Age (weeks) 




00 120 140 

Age (weeks) 




60 80 

Age (weeks) 




25 _'0 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 
(Ne/c). 



REPRODUCTION IN PL^NORBIDS 



83 




Ew (%) 




20 140 160 

Age (weeks) 

FIG. 3. Reproductive "abnormalities" of Planorbar- 
ius corneus reared at 20"C and 25°C: variation in 
the number of capsules including one or more 
eggs v^/ithout egg cells (C J; variation in the number 
of eggs without egg cells (EJ. 



"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, i.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°C and 25°C 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°C and 20°C 
were respectively 7.1% (standard deviation, 
s = 7.5) and 5.5% (s = 6.8). For P. corneus 
reared at 15°C (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 (%) 



40: 



30 



;o 



Cw 
Ew 



Ew (%) 
:40 

30 



oO 



Age 



80 85 
(weeks) 



FIG. 4. Reproductive "abnormalities" of Planorbar- 
ius corneus reared at 15°C (from 49 to 81 weeks 
old): the number of capsules containing one or 
more eggs without egg cells (CJ; the number of 
eggs without egg cells (E J. 



particularly numerous but rarely developed 
further. Whatever the species, no eclosión 
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° С (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 & DAGUZAN 



TABLE 2. "Reproductive abnormalities" in Planorbarius corneus reared at 20 С and 25-C: number of 
egg cells per egg. 





20^ 


С 




25 


С 




Number of eggs 


Percentage of 


Number of 


eggs 


Percentage of 




containing such 


eggs containing 


containing 


such 


eggs containing 


Number of egg 


a number of 


such a number 


a number of 


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 










9 


2 


0.002 










6, 10, 13 & 15 


1 


0.001 










Total 


253 


0.280 


11 




0.130 



The dynamics of pluhannual 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°C (1,600 in- 
dividuals/snail), whereas P. planorbis pro- 
duced the maximum number of newly 
hatched snails (3,357/adult) at 15°C. The 
eggs of P. corneus developed at 10°C, 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°C. How- 
ever, for the embryonic development study, 
the eggs placed in incubation were laid at 
20°C. 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° С 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-Llaud, 
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 Hellsoma 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 (Müller), 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 PIJ\NORBIDS 



85 



TABLE 3. Reproductive "abnormalities" ot Planorbis planorbis in relation to age and tennperature. 









Proportion of 










Proportion of 


capsules including 




Proportion of 






capsules 


1 or more eggs 


Proportion of 


eggs including 




Age 


without eggs 


without egg cells 


eggs without 


2 or more egg 


Temperature 


(weel<s) 


(%) 


(%) 


egg cells (%) 


cells (%) 




29 





20.5 


8.1 


0.43 




31 





5.0 


2.1 


0.70 




33 





4.6 


2.1 







35 











0.75 




37 











0.30 




51 





1.3 


0.5 


0.25 


10°C 


59 


0.9 


1.0 


0.6 







77 


3.4 


3.6 


0.5 


2.30 




105 


8.9 








0.24 




113 


17.2 








0.26 




129 


19.1 


3.9 


1.7 







141 


20.7 













153 


30.8 













31 





0.7 


0.3 







49 





0.9 


0.5 







57 


0.5 


0.5 


0.2 


0.04 




73 


0.3 


0.6 


0.3 







85 


0.6 


1.1 


0.6 


0.05 




93 


1.7 


0.4 


0.06 


0.03 


15"^ С 


117 


1.2 


0.4 


0.02 


0.02 




123 


1.5 


1.2 


0.3 


0.04 




127 


5.2 


4.7 


1.4 







131 


33.3 


20.0 


3.9 







133 


27.3 


30.0 


8.3 







135 


33.3 


33.0 


8.5 







21 


0.2 


1.1 


0.3 


0.02 




35 


0.4 


0.4 


0.1 


0.02 




51 


4.6 


5.0 


1.3 





20 С 


53 


2.9 


6.5 


1.4 







59 


8.5 


11.1 


2.9 


0.04 




61 


11.9 


8.6 


3.0 







63 


11.1 


19.3 


5.6 


0.11 




65 


8.8 


25.8 


9.4 







15 





11.2 


4.5 







25 


1.5 


27 


12.1 


0.10 


25°C 


27 





39.4 


21.4 







29 





90.6 


65.2 


0.60 




31 





95.7 


93.0 







33 





100 


100 






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° С 
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 & 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°C. 










3 


94 


243 


153 








0.75 


















15°C 





0.84 











555 












0.63 























0.11 







Planorbarius 







636 


890 


74 






corneus 


20° С 


0.76 













1600 









0.71 






















0.27 
















128 


22 










25°C 


0.72 




0.06 











150 









0(107) 


0(153) 


0(140) 








10°C 


0.44 













(404) 


Planorbis 







0.63 












planorbis 










1244 


0.21 
1059 











15°C 


0.84 




0.69 











3357 




20° С 



0.55 


1760 









2389 



abnormalities of eggs or egg masses (Bloch, 
1938, for P. corneus and L stagnalis; Bond- 
esen, 1950, ior Ancylus fluviatilis (Müller); 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-LJaud 
(1982) showed that at 33°C 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 8. glabrata low temperatures 
reduced fecundity without obviously damag- 
ing the reproductive system, whereas high 
temperatures reduced the female sexual or- 



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 Ptiysa fontinalis (L.) (per- 
centage of twins. 0.1-0.4%, De Witt, 1955), 
and 47 for P. acuta (Bigus, 1 981 ). 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 PU\NORBIDS 



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: 1 8.2 (Cole, 1 925); 27.8 (N = 63) 
and 33.3 (N = 39) (Oldham, 1930); 71 (N = 28) 
(Alyakhnskaya, 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 (Chemin & 
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 B. 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 & 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. Ptiysa 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 1 0^'C and 1 5°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: 8. 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 I; 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 tern- 



COSTIL & DAGUZAN 



perature for the minimum reproduction 
threshold (10 С for P. planorbis, 15"C for P. 
corneas) 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 
Schalle & Berry (1973) deduced that the lym- 
naeids reproduced and thrived best in cool 
(19' С 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, 
1 993), 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, I. A. & E. A. M. BEDDINY, 1980, Fac- 
tors affecting the reproductive capacity and 
population dynamics of Helisoma duryi (Weth- 
erby). I: effect of temperature. Journal of the 
Egyptian Society of Parasitology, 10: 135-150. 

ALYAKRINSKAYA, I. O., 1981, Egg nutrient con- 
tent in gastropods. Akademiia Nauk SSSR, Dok- 
lady, Biological Sciences, 260: 472-474. 

BIGUS, L., 1981, Polyvitellinität und Keimver- 
schimelzungen bei Pfiysa acuta (Pulmonata, 
Basommatophora). Zoologisctie Jafirbücher 
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: 1 57. 

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- 



turn 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, R. &W. RUSSELL-HUNTER, 1978, Re- 
productive effort in Molluscs. Oecologia, 37: 23- 
27. 

BUCKLEY, D., 1986, Bioenergetics of age-related 
versus size-related reproductive tactics in fe- 
male Viviparus georgianus. Biological Journal of 
ttie Linnean Society, 27: 293-309. 

CALOW, P., 1981, Adaptational aspects of growth 
and reproduction in Lymnaea peragra 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. Ill 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, К., 1993, Contribution à l'écologie des 
mollusques gastéropodes dulcicoles armoric- 
ains; recherches bio-écologiques et écophysi- 
ologiques sur deux espèces de pulmones Plan- 
orbidae: Planorbarius corneus (L.) et Planorbis 
planorbis (L.). Ph. D. Thesis, University of 
Rennes I, 421 pp. 

COSTIL, К., 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: 97-1 1 7. 

DUNCAN, С J., 1 975, Reproduction in pulmonates. 
pp. 309-365, in: V. FRETTER & J. PEAKE, eds., Pul- 
monates, Vol. 1, Academic Press, New York. 

EL EMAN, M. A. & H. MADSEN, 1 982, 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. & 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. K. M., 1976, The effects of low tem- 
perature on Lymnaea truncatula. ZeitschriU für 
Parasitenkunde, 48: 281-286. 



REPRODUCTION IN PU\NORBIDS 



89 



JOOSSE, J. & С J. VELD, 1972, Endocrinology of 
reproduction in the Inermaphrodite 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. 

IJ\RAMBERGUE, de M., 1939, Etude de l'autofé- 
condation chez les gastéropodes pulmones; re- 
cherches sur l'aphallie et la fécondation chez 
Bulinus (Isadora) 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. F. THIONGO & J. H. OUMA, 
1983, Egg-laying and growth in Helisoma duryi 
(Wetherby) (Pulmonata: Planordidae): effect of 
population density and mode of fertilization. Ну- 
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 oí Australorbis 
glabratus in the laboratory. American Journal of 
Hygiene, 73: 66-74. 

MOOIJ-VOGELAAR, J. W., J. 0. 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, С 1930, Fecundity of Planorbis cor- 
neus. Naturalist, 801: 177-178. 

PERRIN, N., 1 986, 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. & R. BLUZAT, 1983, Effets des condi- 
tions d'éclairement sur le potentiel reproducteur 
de Lymnaea stagnalis (gastéropode, pulmoné). 
Malacologia, 23: 321-331. 

VAN DER SCHAUE, H. & 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 Néerlandaises de Zoologie, 17: 403- 
468. 

VAUGHN, С. M., 1953, Effects of temperature on 
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, Gastéropode, Planorbidae). Thèse d'é- 
tat. University of Montpellier II, 185 pp. 

VINCENT, B. & M. GAUCHER, 1983, Variations de 
la fécondité 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 



MAU\COLOGIA, 1995, 36(1-2): 91-95 



COLOUR POLYMORPHISM IN THE MANGROVE SNAIL 
LITTORARIA INTERMEDIA IN SINAI 

L M. Cook & J. Bridle 

School of Biological Sciences, University of Manchester, Manchester M13 9PL 

United Kingdom 

ABSTRACT 

Samples of Littoraha 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 on the lower trunks and pneumatophores of Avicennia marina. Littoraria intermedia is 
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 prédation by crabs is inferred from 
the observed frequency of shell breakage. Polymorphism may therefore occur in species of 
Littoraria in populations subject to prédation. It is not, as has been suggested, restricted to 
situations where prédation may be assumed to be negligible. 

Key words: Littorinidae, prédation, 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 & Garbett, 1992). On the other hand. 
Professor J. Heller (personal communication) 
suggested that the snails living on mangrove 



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 (Forskal) 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 Aqaba 
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 



91 



92 



COOK & BRIDLE 



TABLE 1. Frequencies (%) of different colours in samples of shells of Llttoraha intermedia from Wadi 
Kid, Sinai Peninsula. The standard error is for arcsin transformed frequencies. 



Sample and 


Sample 






Grey/brown 




location 


size 


Orange 


dark 


mid 


light 


1. North, trunk 


232 


7.5 


30.1 


57.3 


5.0 


2. North, trunk 


297 


7.8 


24.5 


61.6 


6.1 


3. North, pneumatophore 


318 


6.4 


16.8 


74.2 


2.7 


4. Central, pneumatophore 


87 


7.9 


7.0 


58.4 


25.8 


5. South, trunk 


117 


7.2 


49.5 


37.6 


5.5 


6. South, trunk 


83 


9.3 


38.4 


33.7 


18.6 


7. South, pneumatophore 


74 


8.2 


14.9 


56.8 


20.3 


Standard error 




0.36 


3.69 


3.12 


3.14 



Kid. No snails were found on the leaves of the 
trees, and none of this species on rocks or 
stones. 



COLOUR VARIATION AND FREOUENCY 

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- 



dards in a colour handbook (Kornerup & 
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, x^ = 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 is 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 is 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 vahation in expression. 



EVIDENCE OF PREDATION 

Damage arising from crab prédation 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 pré- 
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 is 
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 


7.1 


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 pré- 
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, d.f. 6 & 133, P < 0.01), the strongest 
being in the two northerly sites (samples 1 & 
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 prédation, 
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 prédation, selects for polymorphism on 
the vivid and heterogeneous background of 
foliage, whereas prédation favours crypsis on 
bark (the selective hypothesis); or living on 
leaves removes species from the attention of 
predators to such an extent that variant 
iforms 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. It 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 be- 
tween these alternatives. There appears to 
be good evidence for prédation by crabs 
(Reid, 1992), which is sometimes heavy but is 
not necessarily selective. Reid (1987, 1992) 
has shown that apostatic selection, presum- 
ably through the agency of prédation, may 
operate on leaf-living species. Some at- 
tempts to demonstrate selection have failed, 
however (Reid, 1987; Cook & Garbett, 1992), 
and a certain amount of selection does not in 
itself show that {he 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 is 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 0-1 0%o (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 is, 

n = ln{[q„(u+v)-v]/[qn(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 -ln(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 prédation 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 prédation 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 prédation, 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, L. M., 1986b, Polymorphic snails on varied 
backgrounds. Biological Journal of thie 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. & S. D. GARBETT, 1992, Selection in 
the polymorphic mangrove snail Littoraria paii- 
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. M. & G. KENYON, 1993, Shell strength of 
colour morphs of the mangrove snail Littoraria 
pallescens. Journal of Molluscan Studies, 59: 
29-34. 

HAMILTON, P. v., 1976, Prédation on Littorina ir- 
rorata (Mollusca: Gastropoda) by Callinectes 
sapidus (Crustacea: Portunidae). Bulletin of Ma- 
rine Science, 26: 403-409. 



HUGHES, J. M. & M. P. JONES, 1985, Shell colour 
polymorphism in a mangrove snail Littorina sp. 
(Prosobranchia: Littorinidae). Biological Journal 
of the Linnean Society, 25: 365-378. 

KORNERUP, A. & J. H. WANSCHER, 1967, Meth- 
uen handbook of colour, 2nd ed. Methuen, Lon- 
don. 

POR, F. D., I. DOR, & A. AMIR, 1977, The mangal 
of Sinai: limits of an ecosystem. Helgoländer 
Wissenschaftliche Meersuntersuchungen, 30: 
295-314. 

REÍD, 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, Prédation by crabs on Littoraria 
species (Littorinidae) in a Queensland mangrove 

forest, pp. 141-151, in J. GRAHAME, p. J. MILL & 

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, G. J., 1978, Biogeography and adapta- 
tion: patterns of marine life. Harvard University 
Press, Cambridge, Massachusetts, xiii+332 pp. 

VERMEIJ, G. J., 1982, Gastropod shell form, 
breakage and repair in relation to prédation by 
the crab Calappa. Malacologia, 23: 1-12. 

VERMEIJ, G. 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 



MAU\COLOGIA, 1995, 36(1-2): 97-109 

THE RELATIONSHIP BETWEEN SHELL-PATTERN FREQUENCY 

AND MICROHABITAT VARIATION IN THE INTERTIDAL PROSOBRANCH, 

CLITHON OUALANIENSIS (LESSON) 

Michael G. Gardner\ Peter B. Mather^*, Ian Williamson^ & Jane M. 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 prédation. 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 & Emien, 
1993, Cepaea sp.). Differential prédation has 
been suggested as a factor maintaining poly- 
morphisms in many such species (Cook, 
1983, Littohna sp.; review in Cook, 1986; 
Hughes & Mather, 1986, Littohna sp.; Reid, 
1987, Littoraha 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 Cli- 
thon oualaniensis (Gruneberg, 1976, 1978, 
1979; Goodhart, 1987). The variation in С 
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, GPO 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. 



97 



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 
prédation). 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 С 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). If 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 
(0 m rep) 


207 


3 


July 


1992 


Dingo Beach 


shelly sand 
(15 m rep) 


250 


4 


July 


1992 


Dingo Beach 


shelly sand 
(30 m rep) 


221 


5 


July 


1992 


Dingo Beach 


shelly sand 
(45 m rep) 


223 


6 


July 


1992 


Nellie Bay 


shelly sand 


248 


7 


July 


1992 


Champagne 
Bay 


shelly sand 


287 


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 
1 50m^, 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 is 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°C 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- 
tern ("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 



1 



Ф 



Ф 



Ш 



Ш 



T 



1 o 



л л 



# ф ф 

12 13 14 



Ь , I 

10mm 



FIG. 1. Shell-form categories for Clithon oualaniensis. 



pie. 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, 
1 976). 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 HCl 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 III 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 CLITHON 



101 



TABLE 2. Description of Clithon oualaniensis forms at Dingo Beach. 



Morph No. 



Description 



form 1 



form 2 



form 


3 


form 


4 


form 


5 


form 


6 


form 


7 


form 


8 


form 


9 


form 10 


form 1 1 


form 12 


form 13 


form 


14 



characterised by closely spaced transverse lines with or without small tongues on a white or 

green background, 
has 3 spirals (the pattern colour of which is black) superimposed on the basic pattern with a 

green background and usually tongues in whorls, 
similar to Form 2 but with a red or purple pattern colour, 
relate to what Gruneberg (1976) called 'tigers' and 'zebras'. They have widely spaced 

transverse lines on a green background with little or no tongues, 
may be any pattern on a green background but must have a purple spiral, 
similar to Form 4 but with large tongues covering the shell. 
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, 
purple spiral on an orange or yellow background, 
plain white or green snails with little or no pattern, 
plain yellow or orange snails with little or no patterning. 

yellow or orange snails with closely spaced transverse lines with or without small tongues, 
tiger or zebra with or without tongues on a yellow or orange background, 
spiralled orange snails usually found without tongues in whorls, 
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.G. 2. 6. 1.1; Aat-1 
and Aat-2 loci); esterase D (E.C.3.1 .1.1; EstD- 
1); isocitrate dehydrogenase (E.G. 1.1. 1.42; 
ldh-1 and ldh-2) and 6-phosphogluconate de- 
hydrogenase (E.G.I. 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 (z). 
Alleles d and e at the EstD-1 locus were 
grouped as the d allele as they had very similar 
mobilities, and they were difficult to resolve on 
some plates. 

Analysis of Genetic Data 



genotype frequencies from expected values 
was estimated by the F13 statistic (Wright, 
1965), in which F^3 = 1 - (H/2Np,q,) where H 
is the observed frequency of heterozygotes 
in the sample and N is the sample size. 

Shell Form and Allelic Frequency 
Gomparisons (Using a x^ 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. 



Except where indicated, the computer pro- 
gram BIOSYS-1 (Swofford & 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 x^ fest. Deviation of observed 



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 



GARDNER ET AL. 



Shelly sand replicates 
Dingo Beach 




Om 
n = 207 



I I 15m 
I I n = 250 



30m ^Ш 45m 
n = 221 ^1 n = 223 



FIG. 2. Comparison of shell-form frequencies at four replicate sites in a shelly sand site. 



50 T 



February 




2 3 4 5 6 7 8 9 10 11 12 13 14 

form 

1^ seagrass |B shelly H| rock/coral 



FIG. 3. Shell-form frequencies in three microhabitats in February. 



SHELL PATTERN AND HABITAT ASSOCIATION IN CLITHON 



103 



sand microhabitat at Dingo Beach during 
July (Fig 2: X392 = 44.2, p > 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 Хго^ = 98.7, p < 0.01; Fig 4: 
July X26' = 50.0, p < 0.01). (Due to low ex- 
pected values in some cells, forms 8 and 13, 
1 and 1 1 , 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 1 4 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, 1 1 and 1 3 
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 Xi2= = 
35.3, p < 0.01 ; seagrass Хтг- = 29.9, p < 0.01 : 
forms 8 and 13 were grouped), but not in the 
rock/coral microhabitat (/132 = 19.2, p > 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 1 1 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 Xgg^ = 
26.8, p > 0.05; Fig 6: coral/rock Х^^г = 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,3 
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:x42 = 19.458, p < 0.01), but no 
significant difference was found in the July 
collections at this locus (6pgd-2:x42 = 3.965, 
p > 0.05). No significant differences were 
found at any other loci in either the February 
or July collections (February: \äh-^■.X2^ = 
0.339, p > 0.05; \óh-2:x2^ = 0.55, p > 0.05; 
EstD-1 -.Xe^ = 3.331 , p > 0.05) (July: ldh-1 ■.X2' = 
1.226, p > 0.05; Idh-2:x22 = 1.278, p > 0.05; 
EstD-1:x62 = 7.247, p> 0.05). 

There were no significant differences in al- 
lelic frequencies for any locus (ldh-1 :хз2 = 
3.601, p > 0.05; ldh-2:%32 = 2.6, p > 0.05; 
EstD-1 :X92 = 6.953, p > 0.05; 6pdg-2:x62 = 
3.57, p > 0.05), indicating that allelic frequen- 
cies are consistent within and between mi- 
crohabitats and that the Dingo Beach shelly 



104 



GARDNER ET AL. 



50 т 



July 




; seagrass ■■ shelly ^Ш rock/coral 
n = 244 ^1 n = 250 ^1 n = 263 

FIG. 4. Shell-form frequencies in three microhabitats in July. 



Shelly sand 
July 




Dingo Beach 



n = 901 I I n = 248 



Nellie Bay 



10 11 12 13 14 



Champagne Bay 
n = 287 



FIG. 5. Relative shell-form frequencies in shelly sand microhabitats at three localities. 



SHELL PATTERN AND HABITAT ASSOCIATION IN CLITHON 



105 



Rock/coral 
July 




Dingo Beach 
n = 263 



Nellie Bay 
n = 200 



FIG. 6. Relative shell-form frequencies in coral/rocl< microlnabitats 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 0.01 1 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 I 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 



GARDNER ET AL. 



TABLE 3. Allelic frequencies for all populations. 















PopL 


ilation 












Locus 


1 


2 


3 


4 


5 


6 


7 


8 


9 


10 


11 


12 


ldh-1 


























(N) 


22 


47 


45 


48 


48 


48 


47 


72 


59 


48 


58 


45 


A 


0.00 


0.01 


0.00 


0.01 


0.01 


0.03 


0.00 


0.01 


0.01 


0.00 


0.01 


0.01 


В 


0.86 


0.85 


0.87 


0.82 


0.71 


0.79 


0.77 


0.85 


0.80 


0.85 


0.82 


0.84 


С 


0.00 


0.01 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.02 


0.00 


0.00 


D* 


0.14 


0.13 


0.13 


0.17 


0.27 


0.17 


0.23 


0.14 


0.20 


0.13 


0.17 


0.14 


E 


0.00 


0.00 


0.00 


0.00 


0.01 


0.01 


0.00 


0.01 


0.00 


0.00 


0.00 


0.00 


ldh-2 


























(N) 


22 


45 


45 


48 


48 


48 


48 


72 


59 


48 


57 


45 


A 


0.91 


0.91 


0.89 


0.96 


0.96 


0.93 


0.89 


0.87 


0.94 


0.93 


0.91 


0.92 


В 


0.09 


0.09 


0.11 


0.04 


0.02 


0.07 


0.12 


0.10 


0.06 


0.07 


0.09 


0.08 


С 


Ü.00 


0.00 


0.00 


0.00 


0.02 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


D 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.03 


0.00 


0.00 


0.00 


0.00 


Aat-1 


























(N) 


32 


48 


48 


48 


48 


48 


48 


72 


60 


48 


60 


46 


A 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.01 


0.00 


0.00 


В 


0.00 


0.02 


0.00 


0.00 


0.02 


0.00 


0.00 


0.00 


0.00 


0.01 


0.00 


0.00 


С 


0.98 


0.98 


0.99 


1.00 


0.98 


1.00 


1.00 


1.00 


1.00 


0.98 


1.00 


0.98 


D 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


E 


0.02 


0.00 


0.01 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.02 


Aat-2 


























(N) 


32 


48 


48 


48 


48 


48 


48 


72 


60 


48 


60 


46 


A 


0.00 


0.01 


0.00 


0.00 


0.00 


0.02 


0.00 


0.00 


0.00 


0.01 


0.00 


0.01 


В 


0.00 


0.00 


0.03 


0.00 


0.02 


0.01 


0.00 


0.01 


0.02 


0.02 


0.02 


0.01 


С 


0.00 


0.01 


0.00 


0.00 


0.01 


0.00 


0.00 


0.00 


0.00 


0.01 


0.00 


0.01 


D 


0.98 


0.95 


0.96 


1.00 


0.97 


0.95 


1.00 


0.97 


0.97 


0.96 


0.96 


0.97 


E 


0.00 


0.03 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


0.00 


F 


0.02 


0.00 


0.01 


0.00 


0.00 


0.02 


0.00 


0.02 


0.02 


0.00 


0.03 


0.00 


6pgd-2 


























(N) 


33 


46 


46 


48 


48 


48 


48 


72 


58 


46 


59 


45 


A 


0.00 


0.00 


0.00 


0.00 


0.00 


0.01 


0.00 


0.00 


0.00 


0.01 


0.00 


0.00 


В 


0.20 


0.02 


0.01 


0.02 


0.03 


0.01 


0.00 


0.01 


0.01 


0.07 


0.00 


0.01 


С 


0.64 


0.51 


0.49 


0.48 


0.64 


0.57 


0.63 


0.62 


0.50 


0.46 


0.62 


0.66 


D 


0.17 


0.44 


0.45 


0.45 


0.30 


0.37 


0.35 


0.35 


0.48 


0.41 


0.38 


0.30 


E 


0.00 


0.03 


0.05 


0.05 


0.03 


0.04 


0.02 


0.02 


0.01 


0.05 


0.00 


0.03 


EstD-1 


























(N) 


12 


47 


44 


47 


47 


48 


46 


69 


59 


46 


36 


45 


A 


0.08 


0.12 


0.26 


0.12 


0.14 


0.16 


0.11 


0.12 


0.14 


0.13 


0.18 


0.08 


В 


0.71 


0.59 


0.46 


0.66 


0.56 


0.58 


0.55 


0.49 


0.49 


0.69 


0.49 


0.62 


С 


0.13 


0.20 


0.16 


0.15 


0.18 


0.09 


0.23 


0.23 


0.23 


0.13 


0.17 


0.19 


D 


0.08 


0.10 


0.13 


0.08 


0.12 


0.17 


0.11 


0.17 


0.14 


0.05 


0.17 


0.11 



'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 С 
oualaniensis as being "pseudo-polymor- 
phic" has also been criticised by some au- 



SHELL PATTERN AND HABITAT ASSOCIATION IN CLITHON 



107 



TABLE 4. Departure from Hardy-Weinberg as estimated by the fixation index. 







Observed 


Expected 


Fixation index 


Pop. number 


Locus 


heterozygotes 


heterozygotes 


(F,) 


1 


ldh-1 


2 


5.182 


0.614 


2 


ldh-2 


4 


7.289 


0.451 


3 


Aat-2 


2 


3.865 


0.482 




ldh-2 


4 


8.889 


0.550 


4 


EstD-1 


20 


24.670 


0.189 


5 


ldh-2 


2 


3.875 


0.484 


12 


idh-1 


9 


11.967 


0.248 



thors (e.g., Goodhart, 1 987). 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 
prédation (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 CUthon. 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. Malacologie, 28: 1-15. 

CAIN, A. J. & 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 M. & 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 pré- 
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. & 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, С В., 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: Prosobranch ia). 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. M. & M. P. JONES, 1985, Shell colour 
polymorphism in a mangrove snail Littorina sp. 
(Prosobranchia: Littohnidae). Biological Journal 
of the Linnean Society, 25: 365-378. 

HUGHES, J. M. & P. B. MATHER, 1986, Evidence 
for prédation as a factor in determining shell co- 
lour frequencies in a mangrove snail Littorina sp. 
(Prosobranchia: Littorinidae). Evolution, 40: 68- 
77. 

KETTLEWELL, В., 1973, The evolution of melan- 
ism. Oxford University Press, London. 

MASTRO, E., V. CHOW & D. 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. е., 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, B. J., P. R. BAVERSTOCK & M. 
ADAMS, 1986, Allozyme electrophoresis. Aca- 
demic Press, Sydney. 

SWOFFORD, D. L. & R. B. 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). 
Hydrobiologia, 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 



MAIJ\COLOGIA, 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 R. Alonso 

Departamento de Biología Animal, Universidad de La Laguna, 
E-38206 La Laguna, Tenerife, España 

ABSTRACT 

Canariella is a poorly known genus of tine Hygromiidae, endemic to tine Canary Islands, with 
18 nominal taxa of specific and subspecific ranl<. Until now, no information on tine internal 
anatomy of its genital ducts was l<nown, and the external morphology of the genital system, 
which lacl<s 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, 1 878, 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 
deschbed, 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, Tyrrhenielllna 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). 

Ill 



112 



IBAÑEZ, PONTE-LIRA & ALONSO 



meras: Helix (Helicigona) afficta y H. (Helici- 
gona) lens, siendo ambos nombres "no dis- 
ponibles" (ICZN, Artículo 12). Al año 
siguiente, Lamarck (1822) describió ambas 
especies y las denominó Carocolla planaña 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, 1 822 

2. Carocolla planaña Lamarck, 1822 

3. Helix (Helicigona) afficta Férussac, 
1832 

4. Helix bertheloti Férussac, 1835 

5. /-/e//x /eprosa Shuttlewort h , 1852a 

6. Helix fortúnala Shuttieworth, 1852a 

7. Helix discobolus Shuttieworth, 1852b 

8. Helix eutropis Shuttieworth, in Pfeiffer, 
1860 

9. Helix (Macularia) plutonla 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) parryi Ponsonby y 
Sykes, 1894 

18. Helicodonta salten Gude, 1911 





FIG. 1. Medidas tomadas en la concha (explica- 
ción, en el texto). 

de Canariella y se discute su posición entre 
los Hygromiidae. 



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 1 0, 1 6 y 1 7, junto con una 
especie nueva para la ciencia: Groh et al., en 
prensa), incluyendo por prímera vez datos de 
la anatomía interna de sus aparatos repro- 
ductores. Finalmente, está en vías de publi- 
cación la descrípción 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- 
crítas), en este artículo se corrige la diagnosis 



METODOLOGÍA 

Para hacerlas más objetivas, las descríp- 
ciones se han basado, en parte, en datos 
biométrícos. 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, о 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: 
mediana: 
grande: 
muy grande: 



< 11.50 
11.50 - 15.76 
15.76 - 19.02 
> 19.02 



— Indice de la forma de la concha (A/B). Re- 
laciona la altura de la concha con su diá- 
metro. Según su valor, la concha es: 



aplanado-lenticular: < 0.485 

deprimida: 0.485 - 

cónico-ovalada: > 0.60 



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: 



aplanada: 
cónica: 



< 0.225 
> 0.225 



— 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: 



aplanada: 
ovalada: 



< 0.375 
> 0.375 



— Indice 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 - 

redondeada: > 0.895 



0.895 



— Indice 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: 
pequeño: 
mediano: 
grande: 



> 29.095 
20.57 - 29.095 
11.50 - 20.57 
< 11.50 



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 gónada, y "distal" para la más cer- 
cana al orificio genital. 

Abreviaturas: ANSP — Academy of Natural 
Sciences, Philadelphia; CGH — K. Groh prí- 
vate collection, Hackenheim; DMNH — Dela- 
ware Museum of Natural History, Greenville, 
FMNH — Field Museum of Natural History, 
Chicago, MHNG — Muséum d'Histoire Na- 
turelle, Genève; 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 Universität, Zürich. 



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 hakenförmig umgebo- 
gener Spitze. Der Retractor ist an der Grenze 
von Penis und Epiphallus angeheftet; an der 



114 



IBÁÑEZ, PONTE-LIRA & 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 continuación del pene, grueso, hay un del- 
gado epifalo que, después 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 cilindricas; 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 incluidas 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 
"D" 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 





FIGS. 2-3. Canariella hispidula var. hispidula 
(Tabaiba Alta, Tenerife). (2) Collar del manto. (3) 
Complejo palea!. 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 С 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 riñon 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. 28-35c), o in- 
cluso puede faltar completamente, como en 
Canañella 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. 
Éste 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 Canañella 

Collar del manto típicamente con cuatro 
lóbulos poco desarrollados (lateral derecho, 
dorsal derecho, dorsal izquierdo y subpneu- 
mostomal). Riñon 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 С eutropis), 
omitiéndose en ellas su descripción, así 
como la de otros caracteres comunes, para 
evitar repeticiones. 

Canañella hispidula (Lamarck, 1 822) 

Helix (Helicigona) lens Férussac, 1821: 41 
(Folio) о 37 (Quarto), no. 153 [nomen nu- 
dum; non Helix lens, — Deshayes, in Fé- 
russac & 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 Férussac, in Férus- 
sac & Deshayes, 1832: lam. "66*," fig. 4 
[no. fig. 3; non H. barbata Férussac, 
1821]. 

Helix hispidula,— Webb & Berthelot, 1833: 



116 



IBAÑEZ, PONTE-LIRA & ALONSO 




FIGS. 4-1 1 . Concha. (4-9) Canariella hispidula. (4) Lectotipo de Carocolla hispidula (MHNG, foto G. Dajoz; 
diámetro de la concha: 13 mm). (5) Holotipo de Helicodonta salten (NHM). (6) Lectotipo de Helix bertheloti 
(MNHN). (7) Lectotipo de Helix everia (MNHN). (8) Lectotipo de Helix fortunata (NMB). (9) Holotipo de Helix 
(Gonostoma) beata (NHM). (10-11) Canariella planaria. (10) Lectotipo de Helix (Helicigona) afficta (MNHN). 
(11) Lectotipo de Carocolla planaria (MHNG, foto G. Dajoz). Escala: (5-11) 5 mm. 



CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 



117 



TAB1_A 1. Datos biométricos (dimensiones en mm, e indices) de la concha en las diferentes variedades 
de Canariella hispidula. A: altura de la concha; В: diámetro de la concha; С: 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); n: número de 
ejemplares medidos. VALORES: M, valor máximo; m, valor mínimo; X, valor medio; CV: coeficiente de 
variación de Pearson (en %). 



A 



В 



D 



G 



А/В Н/В D/B E/F B/G n 



С hispidula 


var. hispidula 










M 


7.11 


15.33 


5.66 


4.65 


6.26 


7.19 


3.36 


m 


5.14 


11.74 


4.57 


3.33 


4.92 


5.24 


2.01 


X 


6.31 


13.15 


5.27 


4.01 


5.69 


6.26 


2.51 


CV 


5.00 


4.49 


3.78 


6.63 


4.19 


5.84 


10.61 


var. 


bertiieloti 












M 


7.08 


12.71 


5.68 


4.53 


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 


CV 


5.84 


6.46 


5.60 


7.71 


6.39 


7.76 


12.33 


var. 


fortunata 












M 


6.34 


14.30 


5.19 


4.20 


5.97 


7.55 


2.95 


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 


CV 


3.83 


4.02 


3.01 


3.52 


5.09 


4.20 


8.37 


var. 


beata 














M 


6.06 


12.82 


4.86 


3.60 


5.49 


5.70 


2.47 


m 


4.71 


10.07 


3.92 


2.87 


4.28 


4.73 


1.33 


X 


5.11 


11.16 


4.21 


3.17 


4.65 


5.22 


1.79 


CV 


4.96 


5.49 


4.80 


5.53 


5.32 


4.92 


15.46 


var. 


subliispidula 












M 


7.07 


11.99 


5.62 


4.18 


5.35 


6.43 


2.19 


m 


5.41 


9.75 


4.38 


3.21 


4.44 


4.87 


1.11 


X 


6.32 


11.09 


5.13 


3.80 


4.95 


5.62 


1.65 


CV 


5.48 


3.98 


4.56 


4.89 


4.41 


5.47 


11.81 


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 


3.78 


4.30 


0.21 


CV 


4.61 


4.73 


4.09 


5.25 


5.58 


5.89 


40.03 


Conjunto de todas las variedades 








M 


7.11 


15.33 


5.68 


4.65 


6.26 


7.55 


3.36 


m 


4.61 


7.56 


3.70 


2.74 


3.24 


3.81 


0.05 


X 


5.93 


11.47 


4.87 


3.69 


5.02 


5.67 


1.71 


CV 


8.76 


12.06 


8.54 


9.69 


10.56 


11.03 


36.33 



0.48 0.17 0.31 0.91 5.31 24 
5.96 15.94 5.63 2.85 9.23 



0.56 0.22 0.34 0.90 8.38 28 
3.69 13.75 5.52 5.20 10.82 



0.45 0.15 0.30 0.84 5.48 15 
3.45 12.52 5.16 3.55 7.92 



0.46 0.17 0.28 0.89 6.42 12 
5.92 13.01 4.62 5.14 12.47 



0.57 0.23 0.34 0.88 6.87 17 
3.10 8.62 3.75 3.35 10.29 



0.61 0.23 0.38 0.88 53.73 13 
4.38 12.91 3.10 4.88 49.81 



0.52 0.20 0.33 0.89 12.26 109 
10.60 18.82 8.66 4.50 80.68 



314; Pfeiffer, 1 848: 209; 1 868: 260; 1 876: 
294-295 [partim]; Deshayes, in Férussac 
& Deshayes, 1851: 372-373; Mabiíle, 
1884: 69-70; Krause, 1895: 25, fig. 5; 
Kraepelin, 1895: 9 [patiim (loe. = 
Güímar)]; Shuttieworth, 1975: lám. 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 Shuttieworth, 1852a: 141; 
1975: lám. 2, fig. 4; Pfeiffer, 1853: 162; 
1868: 260; 1876: 296; González Hidalgo, 
1869: 37-38; Smith, 1884: 276; Mabille, 
1884: 81-82 ¡partim]; Krause, 1895: 25. 

Helix berthelotii,— Gray, 1854: 1 1 . 



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- 

390 [partim]. 
Helix (Hispidella) lanosa, — Wollaston, 1878: 

384-385. 



118 



IBÁÑEZ, PONTE-LIRA & ALONSO 



Helix (Gonostoma) beata Wollaston, 1878: 

390-391; Mabille, 1884: 85. 
Helix averia Mabille, 1882: 147; 1884: 71, 

lám. 17, fig. 13. 
Helix (Anchistoma) hispidula subhispidula, — 

Tryon, 1887: 122. 
Helix (Anchistoma) evería, — Tryon, 1887: 

123, lám. 38, figs. 4-6. 
Helix (Anchistoma) fortunata, — Tyron, 1887: 

123, ím. 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, — Boettger, 1908: 

246-247. 
Gonostoma fortunata, — Boettger, 1908: 247. 
Helicodonta salten Gude, 1911: 268. 
Canariella hispidula, — Hesse, 1918: 107; 

Odhner, 1931: 86, figs. 370, 380, 40A, 

41 A; Richardson, 1980: 422. 
Canariella fortunata, — Hesse, 1931: 54-55, 

lám. 8, fig. 67a-e; Odhner, 1931 : 87, figs. 

370, 38D; Richardson, 1980: 422; Git- 

tenberger & Groh, 1986: 222-223. 
ICanariella leprosa, — Hesse, 1931: 55, lám. 

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. 

Es una especie polimorfa conquiológica- 
mente, con seis poblaciones diferenciadas, 
siendo ésto un indicio de que se están 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; 1 2 de la var. subhispidula; 1 2 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. 

Oomo 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- 
rife).— 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 
salten (NHM 1922.8.29.33). Lectotipo (selec: 
K. Groh) y 2 paralectotipos de Helix bertheloti 
(MNHN, com. Webb, coll. Férussac). Lec- 
totipo (selec: К. 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 
foríunafa(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 y M. Ibáñez) y 9 paralectotipos de 
Helix (Gonostoma) hispidula subhispidula 
(ZMZ 508663, leg. Fritsch), de Paso Alto, y 
otros 6 (ZMZ 508665/5, 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, С 
L. Richardson y 128654/8, R. Jackson), de 
Candelaria; 3 (MHNG 984/201); 2 (MNHN, 
Maugé); 5 (CGH) y 36 (TFMC), de Candelaria; 
6 (TFMC, J. F. Guerra, 1953), de la calle En- 
rique Wolf son, Sta. Cruz. Var. bertheloti. — 4 
conchas (NHM 1854.9.28.39, Webb y Ber- 
thelot); 4 (FMNH 94096, С D. Nelson), de La 
Orotava; 2 (FMNH 37784, G. K. Gude, ex. G. 
S. Parry); 5 (FMNH 37783, G. K. Gude, ex. H. 
B. Preston); 1 (ZMZ 508669, Tarnier, 1864), 
de Güímar; 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 Güímar; 6 (DMNH 128692, R. Jackson), 
de Güímar; 5 (NMB, Blauner, 1851), de 
Güímar; 5 (MNHN, Jousseaume, Letellier, 
1949 y Bourgeau, 1856); 7 (TFMC, J. M. 
Fernández, 1965), de la Ladera de Güímar. 
Var. fortunata.—6 conchas (FMNH 158.206), 
de Candelaria; 6 (FMNH 1 58.21 3), 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 Всо. Tahodio; 4 (TFMC), del Всо. de San- 
tos; 4 (TFMC), de Las Mesas; 1 1 (TFMC), de 
la Ladera de Pino de Oro. Var. beata. — 5 con- 
chas (NHM 1 854. 9.28.35, Webb y Berthelot), 
de Santa Cruz; 8 (MNHN/2, Locard, 
MNHN/2, M. 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 "He//x berthe- 
loti'' (NHM 1854.9.28.34, d'Orbigny). Var. 
lanosa. — 1 concha (TFMC, J. M. Fernández, 
1955), del Monte Las Mercedes; 23 (TFMC), 
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, 1 822) 
Helicodonta salten Gude, 1911 

Descripción (Tabla 1 , Figs. 2-5, 28) 

El animal tiene el cuerpo de color gris con 
manchas más oscuras, alargadas, que se dis- 
ponen en filas longitudinales en el dorso de la 
cabeza. La concha es aplanado-lenticular, 
con 4 a 5У2 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 
iam, 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 
¡am. En el lado ventral son menores y su lon- 
gitud disminuye hacia el ombligo, en el que 
no sobrepasan los 160 цт. 

Mandíbula con más de 15 costillas. La rá- 
dula tiene la siguiente fórmula: (C + 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 



IBAÑEZ, PONTE-LIRA & 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 dista! 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 copulathz. 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, 1 835) (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 цт de lon- 
gitud en el lado dorsal y no pasan de 1 10 цт 
en el ombligo. Con 3 a 6 glándulas vaginales. 

var. fortunata (Shuttieworth, 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 |im de longitud, mientras que en el 
ombligo no sobrepasan los 100 цт. Con 6 a 
7 glándulas vaginales. 

var. beata (Wollaston, 1878) (Tabla 1 , Figs. 9, 
31). 

La concha es más aplanada y algo más 
pequeña, 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 цт de 
longitud); los del ombligo son todavía más 
cortos, alcanzando hasta 100 |am. 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 цт de longi- 
tud; los del ombligo, en cambio, son 
muchísimo más cortos, no pasando de 60 
цт. 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 



blerto 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 
|.im, 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 цт de longitud. En el 
interior del ombligo, no sobrepasan los 100 
Jim. 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 (Shuttieworth, 1852) 

Helix discobolus Shuttieworth, 1852b: 290 

[loe. typ.: Gomera]; 1975: lám. 6, fig. 6; 

Pfeiffer, 1853: 643; 1868: 260; 1870-76: 

84, lám. 123, figs. 1-2; 1876: 296. 
Helix aflicta Férussac, — 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. 

Habitat 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 6У2 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 цт de longitud, y al ombligo, 
donde son mucho más numerosos y cortos, 
no pasando de 45 цт. 

Mandíbula con 10 a 13 costillas. La rádula 
tine la siguiente fórmula: (C + 18-30L) x 
98-130. Los dientes cercanos al margen 
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 dista! 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 



IBAÑEZ, PONTE-LIRA & ALONSO 




FIGS. 12-20. Concha y SEM detalles. (12) Canariella discobolus (Barranco de la Bajita, La Gomera). (13) 
Canariens 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 jam; 
(19-20) 20 цт. 



CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 



123 



TABLA 2. Datos biométricos (dimensiones en mm, e indices) de la concha de las especies de 
Canariella. Símbolos, como en la Tabla 1 . 



В 



А/В Н/В D/B E/F B/G n 



Canariella discobolus 
M 7.81 20.21 
m 6.62 16.08 
X 7.16 18.14 

CV 4.68 4.78 
Canariella gomerae 
M 6.55 13.79 
m 5.22 11.51 
X 5.89 12.54 

CV 4.07 4.29 
Canariella planaria 
M 5.79 14.95 
m 4.11 12.70 
X 5.13 13.96 

CV 5.07 2.98 
Canariella leprosa 
M 4.68 9.18 
m 4.25 7.97 
X 4.56 8.58 

CV 3.35 5.16 
Canariella eutropis 
M 8.84 16.52 
m 7.32 13.82 
X 8.00 15.00 

CV 4.33 2.81 



5.97 


5.56 


7.11 


9.73 


4.94 


5.14 


4.21 


5.67 


3.56 


2.95 


5.61 


5.03 


6.41 


7.93 


3.95 


3.45 


7.03 


5.43 


9.52 


9.95 


5.16 


4.53 


5.69 


6.67 


2.73 


4.33 


2.93 


4.59 


5.47 


2.07 


4.74 


3.80 


5.01 


5.91 


2.36 


5.28 


7.30 


5.83 


4.58 


9.36 


4.53 


4.37 


6.23 


7.05 


3.29 


3.24 


2.76 


4.53 


5.69 


2.17 


4.03 


3.44 


5.48 


6.31 


2.78 


5.50 


8.84 


4.96 


4.74 


8.29 


3.94 


3.48 


3.91 


5.00 


0.30 


3.49 


2.71 


3.50 


3.90 


0.20 


3.70 


3.07 


3.80 


4.59 


0.25 


5.60 


9.19 


3.95 


7.54 


20.00 


7.47 


6.47 


7.27 


8.06 


3.00 


5.88 


4.83 


5.08 


6.47 


1.62 


6.74 


5.65 


6.43 


6.99 


2.22 


4.36 


5.70 


3.95 


4.06 


12.03 



0.40 0.12 0.28 0.85 4.65 17 
5.65 26.42 6.36 14.52 7.88 



0.47 0.17 0.30 0.85 5.35 11 
2.24 16.21 7.96 2.30 8.93 



0.37 0.12 0.25 0.87 5.07 32 
4.52 18.17 8.31 4.78 8.94 



0.53 0.17 0.36 0.83 35.88 4 
2.49 15.89 4.05 4.09 21.81 



0.53 0.16 0.38 0.92 6.89 29 
4.26 16.96 4.68 4.80 11.29 



Canariella gomerae (Wollaston, 1878) 

Helix (Gonostoma) gomerae Wollaston, 1 878: 
392-393 [loc. typ.: Hermigua, La Go- 
mera]; Mabille, 1884: 84-85. 

Helicodonta (Caracollina) gomerae, — Pilsbry, 
1895: 289; Gude, 1896: 19. 

Caracollina gomerae, — Richardson, 1980: 
424. 

Material Examinado 

Material tipo (conchas vacío). — Lectotipo 
(selec: E. Ponte-Lira y M. Ibáñez) y 3 para- 
lectotipos de Helix (Gonostoma) gomerae 
(NHM 95.2.216-19), de Hermigua (La Go- 
mera); y otro paralectotipo (FMNH 37781, 
colección de Gerard К. 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. 

Habitat 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. 



Descripción (Tabla 2, Figs. 13, 21, 35) 

El animal fijado tiene el cuerpo de color 
blanquecino. La concha es aplanado-lenti- 
cular, con 5V2 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 granulos 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 
|am de longitud, y al ombligo, donde son mu- 
cho más numerosos y cortos (menores de 35 
lam). 

Mandíbula con más de 13 costillas. La rá- 
dula tiene la siguiente fórmula radular: (C + 
25 - 31 L) X 97 - 125. Los dientes cercanos 



124 



IBANEZ, 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. Éste, es largo y muy 
esbelto, con un diámetro ligeramente supe- 
rior al del conducto deferente. El pene es casi 
cilindrico 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, 1 822) 

Carocolla planaria Lamarck, 1822: 99 [loe. 
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; IGZN, 
Art. 12]; Férussac, in Férussac & Des- 
hayes, 1832 (Atlas), lám. 66*, fig. 5; Pfei- 
ffer, 1848: 211; 1853: 162; 1868: 260; 
Deshayes, in Férussac & 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,— WoWaston, 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 (FMNH 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 AI 7996 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. 

Habitat 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 
5У2 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- 



CANARIELLA Y SU POSICIÓN EN HYGROMIIDAE 



125 




22 





Ш 




1 


r 




m 




1 


1 


11 




"^yn 


Çfi 


1^ 


- 




..*í^ 








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) С 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 |am; (22, 23) 1 mm; (24, 25) 125 um; 
(26) 250 jam. 



126 



IBÁÑEZ, PONTE-LIRA & ALONSO 



dula tiene la siguiente fórmula; (0 + 27 31 L) 
X 1 00 - 1 30. Los dientes cercanos al margen 
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 2У2 
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 (Shuttieworth, 1852) 

Helix leprosa Shuttieworth, 1852a: 142 [loe. 
typ.: Tenerife, hic restr.: zona norte de 
Anaga]; non Canariella leprosa, — Hesse, 
1931: 55, lám. 8, fig. 68a-b]; 1975: pl. 3, 
fig. 10; Pfeiffer, 1853: 130; 1870-76: 82- 
83, lám. 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 



los días 3-03-1 981 y 1 5-01 -1 993, en diversas 
localidades. 

Habitat y Distribución (Fig. 41) 

Endémica de Tenerife. Se distribuye por la 
zona norte de la isla, entre 400 y 800 m 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- 
diéndose 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 granulos alargados que 
dan a la concha un aspecto muy caracterís- 
tico, más acusado en el lado ventral. En las 
proximidades del ombligo, los granulos 
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 granulos dorsales. La pi- 
losidad está restringida a la zona de la pe- 
riferia, donde son finos y escasos, alcanzando 
hasta 360 цт 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. Él 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. piñonera 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 C/7/e//a 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, 
1 991) 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 représentâtes en el Archi- 
piélago. 

Canariella eutropis (Shuttieworth, 1860) 

Helix eutropis Shuttieworth, in Pfeiffer, 1860: 
237 [loe. typ.: montes de Jandía, Fuer- 
teventura (figura en la etiqueta del mate- 
rial tipo)]; Albers, 1860: 139; Reiffer, 
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 y M. Ibáñez) y 1 para- 
lectotipo de Helix eutropis (NMB 139, "Dr. 
Bolle detexit"), de los montes de Jandía. 
Otro material.— 2 conchas (FMNH 37788); 3 
(ANSP 397000/2 y Al 7995) 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. 

Habitat 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. 

Descripción (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 4У2 a SVa 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: (C + 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 

30 




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) 
Canahella hispidula var. hispidula (Tabaiba Alta, Tenerife). (29) С hispidula var. bertheloti (Arafo, Tenerife). 
(30) С hispidula var. fortúnala (Cabezo de las Mesas, Tenerife). (31) С liispiduia var. beata (Lomo Bermejo, 
Tenerife). (32) C. Iiispidula var. subhispidula (Montaña del Cascajo, Tenerife). (33) C. Iiispidula var. lanosa 
(Agua García, Tenerife). (34) Canahella discobolus (Barranco de La Rajita, La Gomera). (35) Canahella 
gomerae (Aguajilva, La Gomera). (36) Canahella planaha (Playa Fabián, Tenerife). (37) Canahella leprosa 
(Monte Tenejías, Tenerife). (38) Canahella 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 




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- 



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 

38 




FIG. 38. 



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; Ziich, 
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), in- 
cluyédolos en la superfamilia Helicoidea 
(Damianov & 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 & Manganelli, 1987; Puente & 
Prieto, 1992). 



CANARIELLA Y SU POSICIÓN EN HYGROMUDAE 
CanarJella hispidula 



131 




Canariella planaria 



Canariella leprosa 



40 




^r^ 




1 






■ 


■ 


> 




/ 


y 







41 




— ^ ■ 


) 




■ ■ 


■ 


l 


/ 


y 







FIGS. 39-44. Distribución geográfica. Mapas UTM con cuadrículas de 10 x 10 km, realizados según el 
procedimiento informático de La Roche & 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 




43 








i 


/ ■ ■ 

/ ■ ■ 

■ 

A 


m \ 

■ V 

■■ 

■■ 
■■ ■ ■ 


-} 


BS 


\a 


úy 


y 



44 






) 






y 




f 

ES 


■ ■ 


■ 


^ 



Canariella discobolus 
Canariella gomerae 



Canariella eutropis 



FIGS. 42-44. 



Los trabajos de Nordsieck y Schlleyko han 
tenido la virtud de reactivar el interés de los 
malacólogos en la sistemática de los Heli- 
coidea, pero están 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 C/7/e//a 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 (¡unto con C/7/e//a, Hapiohelix 



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 boisa copulatriz. 
Pero difieren esencialmente por la concep- 
ción plesiomórfica que cada una de ellas 
atribuye a estos caracteres: 

• 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]. 

• 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 (Soiem, 1985). 

Los géneros aparentemente más próximos 
a Canariella son Montserratina Ortiz de 
Zarate López, 1946, Ciliella, Schileykiella, 
Tyrrheniellina Giusti & Manganelli, 1992 
(sinonimia: Tyrrheniella Giusti & Manganelli, 
1989) y Ciliellopsis Giusti & Manganelli, 1990. 
Estos seis géneros comparten las siguientes 
características: 

(A) Microescultura espiral de la concha for- 
mada por crestas muy finas y numero- 
sas. 

(B) Ausencia de estilóforos, sacos y 
apédices accesorios. 

(C) El nervio peneano aparentemente se ori- 
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. 

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 



IBAÑEZ, 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- 
núan en su interior (Giusti & Manganelli, 
1990: 273, fig. 30). 

Canariella también podría estar rela- 
cionada con los géneros Gasulliella Git- 
tenberger, 1980, Caseolus Lowe, 1852, y 
Hapiohelix Pilsbry, 1919, que carecen com- 
pletamente de estructuras vaginales, según 
los dibujos de Gittenberger (1 980: 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 
Pinter, 1977, están incluidos 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 Ihering, 1892, Creti- 
gena Schileyko, 1972, y Caucasocressa 
Hesse, 1921, han sido incluidos 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, 1 929, 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 genero 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 "A" 
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 Férussac, 
1821, y Helicodonta (Férussac) 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), Rüdiger Bieler 
(DMNH; actualmente en FMNH), Alan Soiem 
(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 С 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. С, 1 860, 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 (Cañaren). Zeitschrift der Deutschen Geolo- 
gischen Gesellschaft. 60: 246-249. 

CHEVALLIER, H., 1965, Catalogue des collections 
du Muséum correspondant a l'Histoire Naturelle 
des Mollusques de Férussac. Bulletin du Mu- 
séum National d'Histoire Naturelle, (2)37: 476- 
489. 

DAMIANOV, S. G. & I. M. LIKHAREV, 1975, Fauna 
Bulgarica, 4. Gastropoda terrestha, 526 pp., So- 
fia. 

DESHAYES, G. P., 1850. Ver FÉRUSSAC & DE- 
SHAYES, 1832. 

FÉRUSAC, J. B. DE, 1821, Tableaux systématiques 
des animaux mollusques, classés en familles na- 
turelles, dans lesquels on a établi la concordance 
de tous les systèmes; suivis d'un prodrome gé- 
néral pour tous les mollusques terrestres ou flu- 
viátiles, vivants ou fossiles. II. Tableau de la 
famille des limaçons, 1-90 (Folio) о 1-95 
(Quarto), Paris. 

FÉRUSSAC, J. B. DE, 1832, Ver FÉRUSSAC & DE- 
SHAYES, 1832. 

FÉRUSSAC, J. B. DE, 1835, Commentaires à le 
travail: synopsis Molluscarum terrestrium et flu- 
viatilium quas, in itineribus per Ínsulas Canarias, 
observarunt P. В. Webb et S. Berthelot. Bulletin 
de Zoologie, Sec. 2(68): 90-91. 

FÉRUSSAC, J. B. DE & G. P. DESHAYES, 1830, 
1851 [1819-1851], Histoire naturelle générale et 
particulière des mollusques terrestres et fluviá- 
tiles tant des espèces que l'on trouve aujourd'hui 
vivantes, que des dépouilles fossiles de celles 
qui n'existent plus; classés d'après les ca- 
ractères essentiels que présentent ces animaux 
et leurs coquilles, vol. I: 402 pp. 

FÉRUSSAC, J. B. DE & G. P. DESHAYES, 1832, 
Histoire naturelle générale et particulière des mol- 
lusques terrestres et fluviátiles, Atlas, 22-27: pi. 
66* u. a. & Explications des planches des livrai- 
sons XXII, XXIII, 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- 
átiles. Faune de France, 21: 1-477, 13 pis. 

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. & K. GROH, 1986, Zum Status 
der Férussac'schen Taxa Helix lens und Helix 



barbata (Pulmonata: Helicidae). Archiv für Mol- 
luskenkunde, 1 1 6:21 9-223. 

GITTENBERGER, E. & A. J. DE WINTER, 1985, A 
pyrenean Monacha species, Monacha (Meta- 
theba) atacis spec. nov. (Mollusca: Gastropoda: 
Pulmonata). Zoologische Mededelingen, 59: 
197-207. 

GIUSTI, F. & G. MANGANELLl, 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. MANGANELLl, 1988, Notulae Ma- 
lacologicae XL. Montserratina Ortiz de Zarate 
López, 1946 (Pulmonata, Hygromiidae): a rede- 
scription. Iberus, 8:89-100. 

GIUSTI, F. & G. MANGANELLl, 1989, Notulae Ma- 
lacologicae XLIV. A new Hygromiidae from the 
Tyrrhenian islands of Capraia and Sardinia with 
notes on the genera Xeromicra and Xerothcha 
(Pulmonata: Helicoidea) (Studies on the Sardin- 
ian and Corsican malacofauna, VIII). Bollettino 
Malacologico, 25:23-62. 

GIUSTI, F. & G. MANGANELLl, 1990, Ciliellopsis 
oglasae, a new hygromiid from Montecristo Is- 
land (Tuscan Archipelago, Italy) (Pulmonata: He- 
licoidea). Journal of Conchology, 33:269-277. 

GIUSTI, F. & G. MANGANELLl, 1992, Tyrrheniellina 
nomen novum pro Tyrrheniella Giusti & Manga- 
nelli, 1989 (Pulmonata: Hygromiidae). Bollettino 
Malacologico, 28:231. 

GONZALEZ HIDALGO, J., 1 869, Moluscos del viaje 
al Pacífico verificado de 1862 a 1865 por una 
comisión de naturalistas enviada por el Gobierno 
Español, Parte Primera. Univalvos terrestres. — II 
+ 152 pp., 8 pis, Madrid. 

GRAY, J. е., 1 854, 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 Iles Canaries." 32 pp., London. 

GROK K., С E. PONTE-LIRA, M. R. ALONSO & M. 
IBAÑEZ (en prensa), Revision of the genus Ca- 
nariella P. Hesse 1918. I. Alvaradoa n. subgen., 
with description of a new species from El Hierro 
(Gastropoda: Pulmonata: Hygromiidae). Archiv 
für Molluskenkunde, 123. 

GUDE, G. К., 1 896, Description of a new species of 
Vithna 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., 1 91 1 , 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. Zoológica, 
31: 1-118, pis. 1-16. 

HESSE, P., 1934, Zur Anatomie und Systematik 



136 



IBANEZ, PONTE-LIRA & ALONSO 



palaearktischer Stylommatophoren. Zoológica, 
33: 1-59, pis. 1-9. 

HUTTERER, R. (en prensa), A new, possibly extinct 
species of Canahella from Tenerife, Canary Is- 
lands (Gastropoda: Hygromiidae). Vieraea. 

KERNEY, M. P., R. A. D. CAMERON & J. H. JUNG- 
BLUTH, 1983, Die Landschnecl<en Nord-und 
Mitteleuropas. 384 pp., 24 pis., Hamburg. 

KRAEPELIN, K., 1895, Zoologische Ergebnisse 
einer Frülnjahrs. Exkursion nach Madeira und 
den Kanarischen Inseln. Verhandlungen des 
Naturwissenschaftlichen Vereins in Hamburg, 
(3)2: 6-1 7. 

KRAUSE, A., 1895, Landschnecken von Tenerifa. 
Nachrichtsblatt der Deutschen Malakozoologi- 
schen Gesellschaft, 27: 20-29. 

U\-ROCHE, F. & J. BARQUÍN (en prensa), A com- 
puting method for distribution mapping. Boletim 
do Museu Municipal do Funchal, Suppl. 2. 

LAMARCK, J. B. DE, 1822, Histoire naturelle des 
animaux sans vertèbres, 6(2): 232 pp., Paris. 

LAMARCK, J. B. DE, 1838 (Rev: G. P. DESHAYES & 
H. MILNE EDWARDS), Histoire naturelle des ani- 
maux sans vertèbres. Deuxième édition, 8: 660 
pp., Paris. 

LOWE, R. T., 1861, Diagnoses of new Cañarían 
land-Mollusca. Annals and Magazine of Natural 
History, (3)7: 104-112. 

MABILLE, J., 1882, Molluscornm [sic] novorum di- 
agnoses succintae. Bulletin de la Société 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. II. Nouvelles Ar- 
chives du Muséum d'Histoire Naturelle, (2)8: 1 7- 
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, 1 856 and description of 
Schileykiella n. gen. (Pulmonata: Hygromiidae). 
Journal of Conchology, 33: 141-156. 

MERMOD, G., 1 951 , Les types de la collection La- 
marck au Muséum de Genève. Mollusques vi- 
vants. II. Revue Suisse de Zoologie, 58: 693- 
753. 

MOUSSON, A., 1872, Révision 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, I. Im- 
portance of the excretory and genital systems. 
Archiv für Molluskenkunde, 116: 1-24. 

NORDSIECK, H., 1987, Revision des Systems der 
Helicoidea (Gastropoda: Stylommatophora). Ar- 
chiv für Molluskenkunde, 118: 9-50. 

NORDSIECK, H., 1988. Das System der tertiären 
Helicoidea Mittelund Westeuropas (Gastropoda: 
Stylommatophora). Heidia, 1: 109-120, pis. 15- 
17. 

NORDSIECK, H., 1993a, Phylogeny and system of 
the Pulmonata (Gastropoda). Archiv für 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. Lameliibranchien, 
Cephalopoden, Gastropoden. Arkiv för Zoologi, 
23A: 1-116. 

ORBIGNY, A. D', 1839, Mollusques, echinoder- 
mes, foraminiféres et polypiers recueillis aux Iles 
Canaries par MM. Webb et Berthelot, In: p. в. 
WEBB & s. BERTHELOT, Histoire naturelle des Iles 
Canaries, 2(2)(Zoologie): 152 pp., 14 pis., Paris. 

PFEIFFER, L., 1848-1876, Monographie 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, pis. 
109-137. 

PILSBRY, H. A., 1895, Helicidae, vol. 7. Manual of 
conchology, structural and systematic, (2)9(36): 
161-366, pis. 41-71. 

PONSONBY, J. H. & E. R. SYKES, 1894, Two new 
land-shells from Tenerife. Proceedings of the 
Malacological Society of London, 1 : 55. 

PRIETO, С. е., a. i. PUENTE, К. ALTONAGA & В. J. 
GÓMEZ, 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, A. !. & 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, 1 980, 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 Institute, 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 pis.; Krefeld. 

SMITH, E. A., 1884, An account of the land and 
freshwater Mollusca collected during the voyage 
of the "Challenger" from December 1 872 to May 
1876. Proceedings of the Zoological Society of 
London, for 1884: 258-281, pis. 22, 23. 

SOLEM, A., 1985, Origin and diversification of pul- 
monate land snails, pp. 269-293, in: к. м. 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 I. Manual of con- 



chology, structural and systematic, (2)3: 313 pp., 
63 pis. 

TRYON, G. W., 1888, Helicidae II. Manual of con - 
chology, structural and systematic, (2)4: 296 pp., 
69 pis. 

VAU G HT, С, 1989, А classification of the living 
Mollusca. American Malacologists, Inc., R. т. Ab- 
bott & K. J. BOSS, eds., 1 89 pp., Melbourne, Flor- 
ida. 

VERDCOURT, В., 1975, On the systematic posi- 
tion of some east African Helicidae — 4. Revue de 
Zoologie Africaine, 89: 934-939, pi. X. 

VISSER, M. H. С, 1973, The ontogeny of the re- 
productive system of Gonaxis gw/andaensis, with 
special reference to the phylogeny of the sper- 
matic conduits of the Pulmonata. Annale Univer- 
siteit van Stellenbosch, 48: 1-79. 

VISSER, M. H. С, 1977, The morphology and sig- 
nificance of the spermoviduct and prostate in the 
evolution of the reproductive system of the Pul- 
monata. Zoológica Scripta, 6: 43-54. 

VISSER, M. H. С, 1988, The significance of termi- 
nal duct structures and the role of neoteny in the 
evolution of the reproductive system of Pulmo- 
nata. Zoológica Scripta, 17: 239-252. 

WEBB, P. B. &T. S. BERTHELOT, 1833, Synopsis 
molluscorum terrestrium et fluviatilium quas in 
itineribus per Ínsulas Canarias, observarunt. An- 
nales des Sciences Naturelles, 28: 307-326. 

WOLU\STON, T. v., 1878, Testacea atlántica 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 Paläozoologie, 6: 
601-835 Berlin. 



Revised Ms. accepted 21 March 1994 



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. Lazahdou-Dimitriadou 

Section of Zoology, Department of Biology, University of Thessaloniki, 
54006 Ttiessalonii<i, 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 tliat tlie 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 & 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. 



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 S. graeca, which is 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 



139 



140 



ELEUTHERIADIS & 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 & 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 1 1 -month-old (H = 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 1 1 -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 
|i 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 
fi 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 g). 

For each experimental group, batches of 
food and faeces were analyzed for С 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 С and N values for 
food and faeces. The weights recorded for 
each experimental group could then be used 
to obtain individual rates of total С and N 
ingestion (Tl) and egestion (NA = not assim- 
ilated) and, hence total С and N assimilation 
rates (ТА) for each snail (by subtracting the 
appropriate NA rate from Tl) 

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 1 1 -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 O2 sat- 
uration obtained for measurements in the 
presence of snails. O2 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-RELMED 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 С 


Total N 


Total С 


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 


5.79 ±1.85 


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 
& 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 in 8.5% HNO3, and the perios- 
tracum was removed. The tissue was dried to 
constant weight at 65°C. 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; Aldhdge 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 |il Ог/цд 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- 



iî 



îS 



0.8 



50.6- 



g0.4q 



0.2 



0.0. 



ií 



■ d 



1 3 11 

Age (months) 

FIG. 1 (A) Oxygen uptake (micrograms 02 per snail 
per hour) and (B) total nitrogen excretion (micro- 
grams N per snail per hour) for three ages (1 , 3 and 
1 1 months) of Bithynia graeca. Ш, females, D, 
males. Vertical bars represent standard errors. 



protein fraction. Equivalents for carbon mass 
consumed were derived from the appropriate 
amounts of CO2 evolved (0.536 цд С con- 
sumed/|il CO2). 

The relation of CO2 evolved to O2 con- 
sumed differs for proteins and nonproteins. 
The CO2 evolved from protein catabolism 
can be derived directly from the weight-spe- 
cific nitrogen excretion rate (4.75).il CO2 
evolved/jig N excreted). The CO2 evolved 
from nonprotein catabolism can be esti- 
mated by multiplying the weight-specific ox- 
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 & LAZARIDOU-DIMITRIADOU 



TABLE 2. Mean weight specific partitioning rates (micrograms С 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 


n 


ingested 


С ingested 


assimilated 


assimmilated 


С in RA С in RA 


in NRA 


С in NRA 


1 


— 


51 


1.70 


3.40 


0.95 


1.42 


0.53 


0.31 


0.42 


1.11 


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 


mai. 


23 


1.17 


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 (Tl), 
egestion (NA) and assimilation (ТА) 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 
for3-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 С 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 С 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 1 1 -month-old snails. The dif- 
ferences were statistically significant be- 
tween the 1 -month-old snails and the other 
age classes for protein С 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 С ingested (P < 0.05). There were no 
statistical differences for nonprotein С 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-REU\TED DIFFERENTIAL CATABOLISM IN THE SNAIL 



143 



TABLE 3. Comparisons between different age and sex groups (1 -month-old, 3-months-old or 
11 -month-old female (F) or male (M) snails) of Bithynia graeca concerning the protein С ingested, 
protein С assimilated and protein С in respired assimilation with Fisher LSD test (*:P < 0.05). 



1 



3 F 



3 M 



11 F 



11 M 



11 





Protein 


С 


ingested (PCi) 


— 








Protein 


С 






— 








assimilated 


(PCa) 










Protein 


in 


R.A 


(PRA) 


— 






M 


PCi 








0.174* 


0.190 


— 


M 


PCa 








0.334* 


0.366 


— 


M 


PRA 








0.395* 


0.257 


— 


F 


PCi 








0.174* 


— 




F 


PCa 








0.334* 


— 




F 


PRA 








0.390* 


— 




M 


PCi 








0.163* 


0.180 


0.180 


M 


PCa 








0.313 


0.347 


0.347 


M 


PRA 








0.385 


0.211 


0.251 


F 


PCi 








0.163* 


0.180 


0.180 


F 


PCa 








0.313 


0.347 


0.347 


F 


PRA 








0.386 


0.243 


0.249 



0.170 
0.327 
0.235 



&I 


A 




— 6- 






.* 4. 
" 0, 


Tl 

. 7M . . . ^ 


^S 




3 11 

Age (months) 



FIG. 2. Partitioning of nonprotein (D) and protein 
carbon Щ of Bithynia graeca. (A) Total ingestion 
(Tl) 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 С 
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 1 1 -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 & U\ZARIDOU-DIMITRIADOU 

TABLE 4. Mean assimilation efficiencies (percent) for various budget components. 



Age (months) 
Sex 



1 



3 
females 



3 
males 



11 
females 



11 
males 



С assimilation efficiency 
Protein С assimilation efficiency 
Nonprotein С assimilation efficiency 



51 


33 


39 


26 


23 


45 


38 


33 


42 


48 


57 


48 


46 


47 


54 


42 


33 


31 


39 


44 



TABLE 5. Gross and net growth efficiency averages (percent). 



Age 






Gross growth 


efficiency 




Net growth efficiency 




















(months) 


Sex 


n 


Total С 


Protein С 


Nonprotein С 


Total С 


Protein 


С 


Nonprotein С 


1 


— 


51 


30 


25 


32 


64 


04 




78 


3 


fem. 


33 


16 


14 


17 


42 


20 




52 


3 


mal. 


39 


08 


14 


05 


23 


30 




17 


11 


fem. 


26 


21 


02 


30 


50 


30 




77 


11 


mal. 


23 


29 


11 


38 


60 


44 




85 



young males and the opposite occured in 
older snails. The youngest snails had higher 
gross and net efficiencies for total С 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 (Eleuthehadis & 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 & 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 
& Buckley, 1983), and it is usually assumed 



that in males the greater kinetic energy ex- 
penditures are directed to extensive gene 
dispersion. 

It 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 ТА 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 & Shutov, 1979; Charrier & Daguzan, 
1980; Staikou & Lazahdou-Dimitriadou, 
1989; Lazahdou-Dimitriadou & Kattoulas, 
1991) and in freshwater gastropods (Aldridge 
et al., 1986). 



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, 1 1- 
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% forS. 
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 & Buckley, 1983). 
Burky (1971) reported 19% for N-RAЯA 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-RAH'A 
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 georgianas, 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 MARBACH, 1962, Clinical chem- 
istry, 8: 130. 

CHARRIER, M. & J. DAGUZAN, 1980, Consumma- 
tion alimentaire: production et bilan énergétique 
chez Helix aspersa (Müller) (Gastéropode pul- 
moné terrestre). Annalles de la Nutrition et de 
l'Alimentation, 34: 147-166. 

DANIEL, W. D., 1 991 , Blostatistics: a foundation for 
analysis in the health sciences. John Wiley & 
Sons, New York. 

D'ELIA, С F., P. A. STEUDLER & N. CORWIN, 

1977, Determination of total nitrogen in aqueous 
samples using persulphate digestion. Limnology 
and Oceanography 22: 760-764. 

ELEUTHERIADIS, N. & M. U\ZARIDOU-DIMITRIA- 
DOU, 1 992, Contribution to the ecology of Bithy- 
nia graeca, Westerlund, 1879 Macedonia, 
Greece. Abstracts of the Eleventh International 
Malacologlcal Congress, Siena, Italy, 1992, 175 
pp. 

ELEUTHERIADIS, N. & M. U\ZARIDOU-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. & J. P. BARKHAM, 1976, Ouan- 
titative study of feeding in woodland by the slug 
Arion ater. OIkos, 27: 168-173. 

U\ZARIDOU-DIMITRIADOU, M. & J. DAGUZAN, 

1978, Consummation alimentaire, production et 
bilan énergétique chez Euparypha pisana 
(Müller) (Gastéropode Pulmoné). Annalles de la 
Nutrition et de /' Alimentation, 32: 1317-1350. 

LAZARIDOU-DIMITRIADOU, M. & M. KATOULAS, 
1991, Energy flux in a natural population of the 
land snail Eobania vermiculata (Müll.) (Gas- 
tropoda: Pulmonata: Stylommatophora) in 
Greece. Canadian Journal of Zoology, 69: 881- 
891. 

McMAHON, R. F., 1975, Growth, reproduction, 
and bioenergetics variation in three natural pop- 
ulation of a freshwater limpet Laevapex fuscus 
(C. B. Adams). Proceedings of the Malacologlcal 
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- 
gla, 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 & UXZARIDOU-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, IVl. L. 
APLEY & A. 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. & M. LAZARIDOU-DIMITRIADOU, 
1 989, 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, 1 07: 1 33- 
150. 

TASHIRO, J. S. & 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; Imhe et al., 1990), 
food particle size (Levinton & DeWitt, 1989) 
and even taste (Daidorph & Thomas, 1988) 
has been described. In a related paper (Bren- 
delberger, 1992), I 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. 



Using a true food preference method 
{sensu Peterson & Renaud, 1989), I 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 
(Müller) (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°C, for at least two months be- 
fore the experiments began. During this ac- 
climation period, they were fed with lettuce 
{Radix, Bithynia) or Fontinalis antlpyretica (a 
moss) {Bythinella dunkeri) in non-limiting 
amounts. 



147 



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 cyanobactehum 
{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 {AInus gluti- 
nosa) 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 & Jürgens, 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, 1 973). 
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, i.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 




• * • • 
3<-.4 5.. 4 5-3 5-6 5-7 

food combinations 



2-8 



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 p < 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 1 1 .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 



Synechococcus elongatus 
Chlamydomonas reinhardii 
Sycamore maple, incubated 
Oak, incubated 
Alder, incubated 
Alder, ultrasonicated 
Alder, watered 
microorganisms from alder 



5.4 ±0.1 

6.0 ±0.9 

18.8 ±5.0 

24.4 ±0.8 

12.4 ±0.6 

14.3 ±0.8 

15.4 ±0.3 
12.2±1.2 



8.0 ±0.9 
7.3 ±3.9 
2.5 ±0.3 
2.0 ±0.3 
3.9 ± 0.2 
3.5 ± 0.2 
3.2 ±0.1 
2.9 ±0.6 



a maximum value of 24.4 for incubated oak 
leaves. Intermediate values were found for in- 
cubated alder and sycamore leaves, for wa- 
tered and ultrasonicated leaves of alder, and 
for microorganisms from alder. As the carbon 
content of these foods was generally fairly 
constant (between 43% and 48% of dry 
weight), changes in C:N ratios reflect changes 
in nitrogen content. This is shown in the fourth 
column of Table 1 . Absolute values were from 
2% to 8% nitrogen, with a minimum of 2% for 
oak and a maximum of 8% for cyanobacteria. 

As food with a low C:N-ratio and/or a high 
nitrogen content is generally considered to 
be better food (McMahon, 1974), I tried to 
correlate these data with the results of the 
food preference experiments. Figure 2 shows 
that there is a significant correlation between 
C:N ratio and food preference (solid line) for 
Radix peregra and for Bithynia tentaculata. 
The corresponding regressions are: 

Radix peregra: y = 21.421 -0.248x; n = 14; 
r = -0.532; Bithynia tentaculata: y = 
22.554-0.375x; n = 14; r = 0.722; (with y = 
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; 
r = 0.615; Bitliynia tentaculata: y = 
1.037-0.143x; n = 14; r = 0.876; (with y = 
%N-content and x = food preference value). 

The food preferences of Bytliinella dunl<eri, 
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 



301- 

20 

10 

■B 
"со 
^- 30 

20 

10 



sol- 



Radix 
V V, peregra 



I I I I I I L 



6 

H4 



О 



о 



Bythinella 
dunkeri 



«• • V V 

I I I I I I I 



Bithynia 
jî^ tentaculata 




6 

4 

42 



8 
6 
4 
2 



10 20 30 40 50 60 70 80 " 

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 is 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 a!., 1990), are helpful, but cannot 
be used directly to explain the performance 
of the animals in the field. 

I 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 & Wal- 
ter, 1977; Lodge, 1986; Brendelberger, 
1992), although in general, pulmonates are 
regarded as generalist herbivores (Madsen, 
1992). 

Bytfiinella 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 
Bärlocher (1990). A possible explanation for 
this phenomenon is the increased availability 
of amino acids produced by aquatic hypho- 
mycetes during conditioning. 

Bitliynia 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 & Jürgens, 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 & Col- 
man, 1982). Consequently, the uptake of fil- 
terable green algae should be favoured 
whenever possible. The various dethtal 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- 
iflowing, 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 dunkeh, 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. LCidke, 
who traced numerous snail-runs. Is greatfully 
acknowledged. W. Lamport 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. 



BÄRLOCHER, F., 1990, Factors that delay coloni- 
zation of fresh alder leaves by aquatic hypho- 
mycetes. Archiv für Hydrobiologie, 119: 249- 
255. 

BEGON, M., J. L. HARPER & С R. TOWNSEND, 
1990, Ecology, 2nd ed. Blackwell, Boston. 

BLEAKNEY, J. S., 1989, Morphological variation in 
the radula of Placida denthtica (Alder & 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. 

BRENDELBERGER, H. & S. JÜRGENS, 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 (Kobelt) and its possi- 
ble ecological implications. Proceedings of the 
Malacological Society of London, 39: 203-215. 

CALOW, P., 1973, The food of Ancylus fluviatilis 
(MijIL), 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. G. & 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, R. I., 1992, Food selection in the 
land snail Limicolaria flammea Müller (Pulmo- 
nata, Achatinidae). Journal of Molluscan Studies, 
58: 49-56. 

FINDLAY, S. G. E. G. & 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, D. W., С R. MCCROHAN & S. 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, N. K. & H. B. N. HYNES, 1 971 , The fate 
of dead leaves that fall into streams. Archiv für 
Hydrobiologie, 68: 465-515. 

KNECHT, A. & J. E. WALTER, 1977, Vergleichende 
Untersuchungen der Diäten von Lymnaea auric- 
ularia und Lymnaea peregra (Gastropoda: Ba- 
sommatophora) im Zürichsee. Schweizerische 
Zeitschrift für Hydrologie, 39: 299-305. 

LEVINTON, J. S. & 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- 
ogy, 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 tine 
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, 1 03: 597-600. 

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. 

NEWMAN, R. M., Z. HANSCOM & W. C. 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 & D. С WATSON, 1990, Scraping 
a living: a review of littorinid grazing. Hydrobio- 
logia, 193: 117-138. 



PETERSON, С H. & P. E. RENAUD, 1989, Analysis 
of feeding preference experiments. Oecologia, 
80: 82-86. 

RICHARDSON, W. В., 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: K. M. 
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. S. & S. D. COLMAN, 1 982, 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 



MALACOLOGIA, 1995, 36(1-2): 155-169 

PATTERNS OF LAND SNAIL DISTRIBUTION IN A MONTANE HABITAT ON THE 

ISLAND OF HAWAir 

Robert H. Cowie^, Gordon M. Nishida^, Yves Basset^ & Samuel M. Gon, IIP 

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 Haw/aiian 
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- 
tain species of Achatinellinae (Hadfield & 
Mountain, 1981; Hadfield, 1986; Hadfield & 
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. It 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. 



155 



156 



COWIE ET AL. 




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 Alala," 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, 1 990): (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 
[Myoporaceae]), 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. Miliary 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 Ml -8 in the main area and AI -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 m 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 is often unknown. 
Therefore, it is frequently difficult to identify 
material to a recognized species. This is 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- 
opy height greater than 10 m. 

(3) Canopy closure: Scored in the field in 
12 classes, from 1 if the site was completely 
open, to 1 2 if the 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, i.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 & 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 is 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, 
1 7, 31 , 57) were included in the CCA, but the 
single site (site "x") 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 

Leptachiatina (Angulidens) anceyana Cooke, 
1910 

Leptachatina sp. A 

Leptachatina sp. В 

Leptachatina sp. С 

Leptachatina sp. 
PUPILLIDAE 

Nesopupa (Infranesopupa) subcentralis Cooke 
& Pilsbry, 1920 

Pronesopupa sp. 
SUCCINEIDAE 

Succinea konaensis Sykes, 1 897 
HELICARIGNIDAE 

Euconulus (Nesoconulus) sp. of. gaetanoi 
(Pilsbry & Vanatta, 1908) 

Philonesia sp. 
ZONITIDAE 

Nesovitrea hawaiiensis (Ancey, 1904) 

Striatura (Pseudohyalina) sp. of. meniscus 
(Ancey, 1904) 

ISthatura 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 infra- 
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 & Pilsbry, 1920, P. or/cta 
Cooke & 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. 



ISthatura 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 IStriatura 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 
& 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 IStriatura 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). In 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. 



IJ\ND SNAIL DISTRIBUTION PATTERNS IN HAWAII 



161 




FIG. 2. Partitioning of eigenvalues across the six 
axes extracted by the CCA. 



Lamellidea sp., Succinea konaensis and 
IStriatura 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 1 00 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 (r = -0.466, p = 0.069 and r = -0.039, 
p = 0.887, respectively, n = 16). The correla- 
tions for axes 3 and 4 were better (r = 

0.488, p = 0.055 and r = 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. 



CM 
(Л 

X 

< 




AXIS1 



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., В — Tor- 
natellides sp(p)., С — Leptachatina lepida, D — Leptachatina anceyana, E — Leptachatina sp. A, F — Lepta- 
chatina sp. B, G — Leptachatina sp. C, H — Nesopupa subcentralis, I — Pronesopupa sp., J — Succinea ko- 
naensis, К — Euconulus sp. cf. gaetanoi, L — Philonesia sp., M — Nesovitrea hawaiiensis, N — Striatura sp. cf. 
meniscus, О — IStriatura 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-comnnunities. 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) 



U\ND 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 


Axis 3 


Axis 4 


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 


1.12 


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 


-0.92 


-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. В 
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. 0, 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 



Lamellidea sp. 
Tornatellides sp(p). 
Leptachatina lepida 
Leptachatina anceyana 
Leptachatina sp. A 
Leptachatina sp. В 
Leptachatina sp. С 
Leptachatina sp. 
Nesopupa subcentralis 
Pronesopupa sp. 
Succinea i<onaensis 
Euconulus sp. cf. gaetanoi 
Philonesia sp. 
Nesovitrea hawaiiensis 
Striatura sp. cf. meniscus 
7Striatura sp. 
Vitrina tenella 



+ 


+ 





+ 


+ 


+ 





+ 


+ 


+ 


+ 


+ 


+ 


+ 


+ 


+ 









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. В 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 hawaiiensis 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- 
anoi 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 Leptachatlna 
sp. A, Leptachatina sp. C, Pronesopupa sp., 
Phllonesia sp., Nesovitrea hawaiiensis and 
IStriatura sp. The distributions of Leptacha- 
tina sp. A, Ptiilonesia sp. and Nesovitrea ha- 
waiiensis 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, 1 7 of the 23 at 
which Phllonesia sp. was found, and 19 of 
the 26 at which N. hawaiiensis was found, 
were on aa. All these associations are statis- 
tically significant (G tests; p < 0.005 in all 
cases). Leptachatina sp. С 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 IStriatura sp., plot 
among the pahoehoe sites on the ordination 
diagram. Of these two species, L. anceyana 
occurred at very few sites (see above), but 
IStriatura 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 Vithna 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 M 1-3). 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, Phllonesia 
sp. and Nesovitrea hawaiiensis 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 M 1-4) 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, Phllonesia sp. and 
Nesovitrea hawaiiensis, 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 1 2 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 prédation, 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 & 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 in- 
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 is the real variable, as- 
sociated with lava type, that is 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 little 
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 & 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 donn- 
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 & 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 all species, even if not 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. В (14 speci- 
mens), Euconulus sp. cf. gaetanoi (25), and 
perhaps for Leptachatina anceyana (46) and 
Leptactiatina sp. С (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, 
IStriatura 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 IStriatura 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. 



U\ND SNAIL DISTRIBUTION PATTERNS IN HAWAII 



167 



GENERAL CONCLUSIONS 

The land snail connmunity 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, 
1 992), 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 is 
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 Hawaiian en- 
demic family Amasthdae, which once num- 
bered over 400 species-group taxa (Cowie et 
al., 1994). Amastrids, and perhaps the genus 
Leptacliatina in particular, seem highly sus- 
ceptible to habitat modification (Christensen 
& Kirch, 1 986). The survey area is 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 préda- 
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- 
waii (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 Hawaiian envi- 
ronment. 



LITERATURE CITED 

ARAD, Z., 1 990, Resistance to desiccation and dis- 
tribution patterns in bush-dwelling snails. Jour- 
nal of Zoology, 221: 113-124. 

ARMSTRONG, R. 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, H. В., 1941, Zonitid snails from Pacific is- 
lands, parts 3 and 4. Bernice P. Bishop Museum 
Bulletin, 166: 205-370, pis. 43-65. 

BAUR, B. & 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., 1 981 , Ouantitative studies on some 
living British wetland mollusc faunas. Biological 
Journal of the Linnean Society, 15: 299-326. 

BORCARD, D., P. LEGENDRE & 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, 1 1 1 : 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, R. A. D., 1 992, 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, R. A. D. & P. J. DILLON, 1984, Habitat 
stability, population histories and patterns of 
variation in Cepaea. Malacologia. 25: 271-290. 

CHESSEL, D. & S. DOLÉDEC, 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, С С. & P. V. KIRCH, 1986, Non- 
marine mollusks and ecological change at Bar- 
bers Point, Oahu, Hawaii. Bishop Museum Oc- 
casional Papers, 26: 52-80. 

CLAQUE, D. A. & G. B. DALRYMPLE, 1987, The 
Hawaiian-Emperor volcanic chain. Part 1. Geo- 
logic evolution. United States Geological Survey 
Professional Papers, 1350(1): 5-54. 

COOKE, С M., Jr. & Y. KONDO, 1960, Revision of 
Tornatellinidae and Achatinellidae (Gastropoda, 
Pulmonata),ß/s/?op/Wuseumeu//ef/n, 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, R. H., 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 & С С. CHRIS- 
TENSEN, in press. Catalog of the native land and 
freshwater molluscs of the Hawaiian Islands. 
Bishop Museum Bulletin. 

COWIE, R. H. & J. S. 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. & 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. 

GAGNÉ, W. С & L. W. CUDDIHY, 1990, Vegeta- 
tion. Pp. 45-114 In W. L. WAGNER, D. R. HERBST & 

S. H. SOMMER, eds., Manual of the flowering plants 
of Hawaii, Vol. 1. University of Hawaii Press & 
Bishop Museum Press, Honolulu. 

GIAMBELLUCA, T. W., M. A. NULLETT & 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 próxima. Pacific Science, 43: 
1-16. 

HADFIELD, M. G., S. E. MILLER & A. H. CARWILE, 
1993, The decimation of endemic Hawaiian tree 
snails by alien predators. American Zoologist, 
33: 610-622. 



HADFIELD, M. G. & B. S. MOUNTAIN, 1 981 , 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. & P. M. VITOUSEK, 1994, Succes- 
sional development of a Hawaiian montane 
grassland. Biotropica, 26: 2-11. 

LEBRETON, J. D., D. CHESSEL, R. PRODON & N. 
YOCCOZ, 1988, L'analyse des relations espè- 
ces-milieu par l'analyse canonique des corres- 
pondences. Acta Oecologica/Oecologia Gener- 
alis, 9: 53-67. 

MORDAN, P. В., 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. 

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. 

PETERSON, D. W. & R. 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. 

ROHLF, F. J. & R. R. SOKAL, 1969, Statistical ta- 
bles. Freeman, San Francisco, xi + 253 pp. 

SEVERNS, R. M., 1981, Growth rate determination 
0^ Achatinella lila, a Hawaiian tree snail. Nautilus, 
95: 140-144. 

SMALLRIDGE, M. A. & G. С KIRBY, 1988, Com- 
petitive interactions between the land snails 
Theba pisana (Mijiler) and Cernuella virgata (Da 
Costa) from South Australia. Journal of Mollus- 
can Studies, 54: 251-258. 

SOKAL, R. R. & F. J. ROHLF, 1981 , e/omefAy, 2nd 
Edition. Freeman, San Francisco, xviii + 859 pp. 

SOLEM, [G.] A., 1976, Endodontoid land snails 
from Pacific Islands (Mollusca: Pulmonata: Sig- 
murethra). Part I. 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 & 
A. с VAN BRUGGEN. eds., World-wide snails. E. 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. & F. M. CLIMO, 1985, Structure 
and habitat correlations of sympatric New 
Zealand land snail species. Malacologia, 26: 
1-30. 

SOLEM, [G.] A., F. M. 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, D. TURNER & J. J. 

оду, 8:453-485. LOCKWOOD, 1992, The Mauna Loa environ- 

TER BRAAK, С J. F., 1986. Canonical correspon- mental matrix: foliar and soil nutrients. Oecolo- 

dence analysis: a new eigenvector technique for gia, 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 NORM AUS (PILSBRY 1900): 

THE EFFECTS OF LABORATORY STRESS 

Alan E. Stiven 

Department of Biology, University of North Carolina, Ctiapel 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 PGI) 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 & Bailey, 1982; Baur, 
1 988; Perry & Arthur, 1 991 ). However, the re- 
lationship of reproductive output to growth, 
body size and adult age is less well estab- 
lished (e.g. Carter & Ashdown, 1984; Baur, 
1988; Baur & Baur, 1990; Foster & 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 



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 rather than offspring of a single mating 
(Mallet et al., 1 986), when the sample consists 
of a young rather than an older cohort (Diehl 
& Koehn, 1985), or when the number of loci 
examined is large (Koehn et al., 1988). Expla- 
nations range from overdominance (Zouros et 
al., 1 980, 1 983), to the presence of null alleles 
(Foltz, 1986), and to different responses of 
genotypes to environmental stress (Koehn & 
Shumway, 1982; Hawkins et al. 1986; Holley 
& Foltz, 1 987). Attempts have also been made 
to assess the contribution of individual loci to 



171 



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 & Beaumont, 
1988: Borsa et al., 1992). 

Even though stress was predicted to en- 
hance the heterozygous advantage (Koehn & 
Shumway, 1982; Mitten & Grant, 1984), allo- 
zyme heterozygosity and fitness have been 
inadequately studied under stressful and op- 
timum conditions (Hoffmann & Parsons, 
1991). Most evidence supporting the impor- 
tance of stress has come largely from bivalves 
(e.g. Diehl & Koehn, 1 985; Gentili & Beaumont, 
1988 [but see Gaffney's 1990 reanalysis], 
Koehn & Bayne, 1989; Scott & 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 & 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., 1 993), 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 (< 2/1 
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 Hydrologie 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 & 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 & Gain, 
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 1 00-1 1 5 days of growth. Labora- 
tory temperature ranged from 21 °G to 24°C, 
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 a LiOH buffer. MPI 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 & 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, hatclning 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 



Clutch Size (no.) 

Hatching Success 

Survival 

Growth Rate (1 00 days) 



62.88 + 4.13 
0.91 ±0.02 
0.81 ±0.02 
0.68 ±0.03 



71.75 ±7.60 
0.91 ±0.02 
0.34 ±0.05 
0.41 ±0.03 



1.83 
0.15 
8.79 
5.74 



0.09 
0.88 
<0.0001 
0.0001 



gree of freedom using a pooling procedure 
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 
D = (Hq - He) / Hg, where H^ and Hg 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 Fgy 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 
P = 0.05. 



RESULTS 

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 1 989 "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, tc)f^i4 = 
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 



0.6 



0.5 



0.4 



0.3 



0.2 



MESODON GROWTH AND HETEROZYGOSITY 

0.8 



0.7 



175 





STRESSED 


п 


- 


- 




о 


- 


T ' 


» 


- 


Í 






1 


Regression F = 
г2 = 0.053 "• 
n = 192 

1 1 


10.62*" - 

1 



0.6 



0.5 



0.4 





UNSTRESSED 






(1 


, í T 
1 f 


(1 


1 


Regression F = 
г2 = 0.0004 ns 
n = 373 

1 1 1 


17 


ns 



12 3 4 

Number Heterozygous Loci 



12 3 4 

Nurnber 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; X^ = 6.03, P = 0.014). Two loci, MPI 
and АЬЛ, were also monomorphic in High- 
lands but not Coweeta snails, contributing to 
the lower P-value in the Highlands' popula- 
tion. The overall Fgr (Nei 1977) for all loci 



(0.121; 
0.001), 



was significant (X" 



= 35.4, P < 



Mean F|s and Р^ 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 r^ 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-, 190 = 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 183 = 26.62, P = < 0.0001 ; Fj 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 






Unstressed 






Differ 


Differ. 








Brood 


(%) F P r^ 


Brood 


(%) 


F 


P 


r^ 


C1 


no surviving homozygotes 


CI 


-2.6 


0.43 


0.52 


0.009 


C2 


+8.3 0.04 0.85 0.001 


C2 


+27.5 


3.78 


0.06 


0.078 


C3 


no surviving homozygotes 


C3 


-4.3 


1.01 


0.32 


0.032 


C4 


+20.6 0.36 0.57 0.056 


C4 


+11.8 


1.36 


0.25 


0.027 


C6 


-1.0 2.84 0.10 0.098 


C5 


no surviving homozygotes 


C7 


no surviving homozygotes 


C6 


+39.9 


4.07 


0.05 


0.085 


H3 


+20.0 5.93 0.02* 0.165 


C7 


+ 12.8 


4.42 


0.04* 


0.121 


H4 


+12.0 0.587 .451 0.022 


HI 


-1.9 


0.48 


0.49 


0.00 



vival annong broods in the stressed treat- 
ments (X^df=7 = /'З.З, P < 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 of the 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^ ^go = 
36.96, P = < 0.0001 ; Fi ^j, = 8.27, P = 0.004; 
Hetloci: Ft ^qo = 66.11,' P < 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 III 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 



was also performed on the unstressed data 
set. Two loci, ALG and AU\, 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 



о 

о 



0) 

э 
о 

О) 

N 
О 

!■ 
Ф 



(О 

> 

(О 
Q 

О 
О 



Ф 

(О 
ОС 



о 



MESODON GROWTH AND HETEROZYGOSITY 

Stressed 



177 



2.0 



1.5 



1.0 



«^ 0.5 

Ф 



0.0 



1.4 



1.2 



1.0 



0.8 



0.6 



0.4 



m HETL 
D GRO 



Co 



Hi 



Unstressed 



Ш HETL 
D GRO 



Co 



H 



2.0 



1.0 



0.0 



1.2 



0.8 



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, P < 0.0001; 
Fi,37i = 5.69, P = 0.018, respectively). 
Grov^h rate did not differ between shell color 
morphs, but was higher in mottled than in 
non mottled morphs (F-, 371 = 46.60, P < 
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 & 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 Hi sum 
of squares (SS). R^ == 0.103 and overall 
regression F^^^e = 4.28, P = 0.001. 



Locus 


H 


SS 


F 


P 


APA 


0.54 


0.148 


10.54 


0.001 *" 


PGI 


0.34 


0.068 


4.86 


0.029* 


PGM-2 


0.06 


0.010 


0.71 


0.400 


MPI 


0.02 


0.002 


0.14 


0.710 


APG 


0.55 


0.001 


0.07 


0.793 



*P < 0.05, "'P < 0.001 

of these studies, the positive association in 
Mesodon occurred during the vulnerable 
early life stage of the cohort, a time when 
most energy is being allocated to somatic 
growth in mollusks (Zouros et al., 1980). 
However, for M. 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 is small (5.3%), but corresponds 
to r^ 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 & 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 



(Л 



>4 


0.45 


Û 




о 
о 




CD 
CO 

ce 


0.4 


SI 




о 


0.35 



О 



ns 




1 



DHom HHet 



icick 



Г 

T 


IS 






y/ 




ns 




1 




T 


1 






''// 




1 






w 






7/, 



0.3-^ 

PGM-2 PGI MPI ALA-2 ALG 

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 P = 0.05, *** means significant at P < 0.001 (t-tests). 



X 

4—» 

(0 

о 

СП 

N 
О 

ф 

О) 

X 



0.5 



0.4 



0.3 



0.2 



0.1 



О 



Stressed 



COW 



HIGH 



Had Gyn 



unstressed 




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 (-D) for loci in the stressed and unstressed colnorts 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 


-0.710*** 


0.260 


0.169 


-0.110 




HIG 


0.017 


-0.740** 


0.037 


-0.098* 


0.293 




POOLED 


0.022 


-0.723*** 


0.193 


0.033 


-0.105 


Unstressed: 


COW 


0.009 


-0.332'** 


-0.022 


-0.074 


-0.205*** 




HIG 


-0.153 


-1.000*** 













POOLED 


-0.393-" 


-0.417*** 


-0.037 


-0.098* 


-0.398*** 



*P < 0.05, •• 0.05 < P < 0.01, ***P < 0.001 



TABLE 5. Comparisons of tissue and shell color properties of surviving young in the stressed and 
unstressed cohorts. 



Comparison 


Unstressed 


Stressed 


Tissue Color* 






Between sites 


ns 


ns 


Growth among colors 


ns 


F2i89 = 5.21, P = 0.006 
1 <2, 2 = 3 


Heterozygosity among colors 


ns 


F2 189 = 4.21, P = 0.016 
1 < 3, 2 = 3 


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 


Fi37i =42.61, P< 0.0001 
1 <2 


not recorded 


Tissue Mottling* 




Between sites 


X2 = 8.69, P = 0.003 


not recorded 


Growth between classes 


Fi 371 = 46.6, P < 0.0001 

2> 1 
Fi37i =6.69, P = 0.018 


not recorded 


Heterozygosity between classes 


not recorded 




1 >2 





'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^Ag = color 2, AgAg = 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 



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 



lern 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 & 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 Fgr 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|s and F|-r also reflect the 
high levels of homozygosity in the popula- 
tions, expected in a species with limited dis- 
persal and probably inbreeding. These Fgy 
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 1 1 
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, 1 977). 
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 

Support 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 Hydrologie 



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. & R. F. LEARY, 1986, Het- 
erozygosity and fitness in natural populations of 
animals. Pp. 57-76, In: м 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. B. & G. F. MCCRACKEN, 1986, 
Breeding system and population genetic struc- 
ture in pinilomycid slugs (Mollusca: Pulmonata). 
Biological Journal of the Linnean Society, 29: 
317-329. 

BAUR, В., 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, B. & A. BAUR, 1990, Experimental evidence 
for intra- and interspecific competition in two 
species of rock-dw/elling lands snails. Journal of 
Animal Ecology, 59: 301-315. 

BEAUMONT A. R., С M. BEVERIDGE & M. D. 
RUDD, 1983, Selection and heterozygosity 
within single families of the mussel Mytilus edulis 
(L). l\/lahne Biology Letters, 4: 151-161. 

BEAUMONT, A. R., E. M. GOSLING, С M. BEV- 
ERIDGE, M. D. RUDD & G. M. BURNELL, 1985, 
Studies on heterozygosity and size in the scal- 
lop. Pectén maximus (L.). Pp. 443-455, In: P. E. 
GIBBS, ed.. Proceedings 19th European Marine 
Biology Symposium. Cambridge University 
Press, England. 

BOOTH, С L, D. S. WOODRUFF & S. 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 & B. DELAY, 1992, Re- 
lationship between allozyme heterozygosity, 
body size, and survival to natural anoxic stress in 
the palourde Ruditapes decussatus L. (Bivalvia: 
Veneridae). Journal Expehmental Marine Biology 
and Ecology, 1 55: 1 69-1 81 . 

CAIN, 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, M. A. & M. 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, R. H. & A. J. CAIN, 1 983, Laboratory main- 
tenance and breeding of land snails, with an ex- 



ample from Helix aspersa. Journal of Molluscan 
Studies, 49:176-177. 

DAN, N. & S. E. R. 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. & R. K. KOEHN, 1985, Multiple-locus 
heterozygosity, mortality, and growth in a cohort 
of Mytilus edulis. Marine Biology, 88: 256-271 . 

EMBERTON, K. 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. & S. P. GARNER, 1987, Environ- 
mental stress and allozyme variation in Littorina 
littorea (Prosobranchia). Mahne Ecology 
Progress Series, 39: 129-136. 

FOLTZ, D. M., 1 986, Null 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 & R. K. SELANDER, 
1984, Genetic diversity and breeding systems in 
terrestrial slugs of the families Limacidae and 
Arionidae. Malacologia, 25: 593-605. 

FOLTZ, D. M., S. E. 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. M. & B. MCGEE, 1992, Multiple pa- 
ternity in Crepidula fornicata (Linaeus). The 
Veliger, 35: 12-15. 

GAFFNEY, P. M. & T. SCOTT, 1984, Genetic het- 
erozygosity and production traits in natural and 
hatchery populations in bivalves. Aquaculture, 
42: 289-302. 

GARTON, D. W., 1984, Relationship between mul- 
tiple locus heterozygosity and physiological en- 
ergetics of growth in the estuahne gastropod 
Thais haemastoma. Physiological Zoology, 57: 
530-543. 

GENTILI, M. R. & R. A. BEAUMONT, 1988, Envi- 
ronmental stress, heterozygosity, and growth 
rate in Mytilus edulis L. Journal of Expehmental 
Mahne Biology and Ecology, 1 20: 1 45-1 53. 



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. & A. G. CU\RK, 1989, Principles of 
population genetics, 2nd ed. Sinauer Associates, 
Sunderland, Massachusetts. 682 pp. 

HAWKINGS, A. J., B. L. BAYNE & A. J. DAY, 1 986, 
Protein turnover, physiological energetics and 
heterozygosity in the blue mussel, Mytilus edulis: 
The basis of variable age-specific growth. Pro- 
ceedings of tfie Royal Society of London, (B) 
229: 161-176. 

HOFFMAN, A. A. & P. A. PARSONS, 1991, Evolu- 
tionary genetics and environmental stress. Ox- 
ford University Press. Oxford, England. 284 pp. 

HOLLEY, M. 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 moliusks of the eastern United States. Fiel- 
diana, (n.s.) 24: 1-191. 

KOEHN, R. K. & 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 & 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. 

KOEHN, R. K. & S. 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 & K. R. FREEMAN, 1986, Genetics of 
growth in the blue mussel: family and enzyme- 
heterozygosity effects. Marine Biology, 92: 475- 
482. 

MITTON, J. B. & M. 0. 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 in 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.). Nethierlands 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 tfie Academy 
of National Sciences, Phiiladelphia, 52: 110- 
150. 

PILSBRY, H.A., 1940, Land Mollusca of North 
America. Academy of Natural Sciences of Pfiila- 
delphia. Monographs, 1 : 994 pp. 

SCOTT, T. M. & R. K. 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. & M. E. SOULE, 1 983, 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 & R. B. SEU\NDER, 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. IV. University Chicago Press, 
Chicago, Illinois, 580 pp. 

ZOUROS, E., 1987, On the relationship between 
heterozygosity and heteroses: An evaluation of 
the evidence from marine moliusks. Isozymes: 
current topics in biology and medical research, 
15: 255-270. Liss, New York. 

ZOUROS, E. & D. W. FOLTZ, 1987, The use of 
allelic isozyme variation for the study of hetero- 
sis. Isozymes: current topics in biology and med- 
ical research, 13: 1-59. Liss, New York. 

ZOUROS, E., M. ROMERO-DOREY & A. 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 & 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 & 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 



MAL7\C0L0GIA, 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 is 
evolutionahly 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, 1 971 : 
N1094) 



INTRODUCTION 

Stenzel (1 971 ) 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, 1 81 (Table 1 A). Polyphyly 
might be inferred from the subsequent cre- 
ation of some additional ostreoidean families 
[e.g. for Crassostrea and its allies by Scarlato 
& 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 
IB), 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, 1 985) 
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; 



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 IB). 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 phmitive-versus-derived character 
recognition (Henning, 1 966; Ax, 1 987; Funk & 
Brooks, 1990) to the oysters. A clear under- 



185 



186 



UWVRENCE 



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; Thassic-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; Thassic-Neogene 

*Subfamily Lophinae Vyalov, 1936; Thassic-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; Thassic-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 v\/itliin-taxon variability, is critical 
to tliis 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 & Kauf^man (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; Freneix, 1982; 
Harry, 1 985; 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 & 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 & 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 
with Crassostrea 



Genus, by numbers above 



2 


3 


4 


5 


6 


7 


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 
X 
X 
X 


X 
X 

X 



A. Denticles or chomata present 

B. Adult shell size smaller/larger 
С Greater valve massiveness 

D. Umbonal/ligamental areas different 

E. Conical shell form present 

F. External ornamentation different 

G. Muscle scar placement different 
H. Muscle scar geometry different 



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., 1 980; Harry & Dockery, 1 983).] 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. 



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 1 990, 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 1 B) apparently removed all of these 



Il 



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 Crassosirea -I ike genus Konbos- 
trea Chinzei, 1 986, or the similar-aged genera 
Soleniscostrea Chiplonkar & 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 Crassosirea-like genera that Malchus 
removed to the Liostreinae [Pseudoperna Lo- 
gan, 1899; Acutostrea Vyalov, 1936; Gyros- 
trea Mirkamalov, 1963; Indostrea Chiplonkar 
& 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 Crassosirea-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 1 1 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 virglnica and its History 

Gunter (1950) and Stenzel (1947, 1971) 
have summarized the nomenclatural history 
of the species now known as Crassostrea vir- 
glnica (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 1 00 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 
С gigantlssima (Finch, 1824) was the direct 
precursor of the extant C. virginica. Hopkins 
(1978), in a published abstract, suggested 



GENUS CRASSOSTREA 



189 




шаМ 



% 



.гзй#'^-^ 








FIGS. 1-2. Chomata on right valves of Crassostrea 
gigantisslma (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 dorsopostehor 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 C. virginica included the 
Late Cretaceous species C. glabra (Meek & 
Hayden, 1857). Among other reasons, Hop- 
kins chose C. glabra because of its v\/ide- 
spread occurrence in settings interpreted as 
brackish water, and he suggested that this 
ancestry may explain the extreme tolerance 
of С virginica, among oysters, to waters of 
lowered salinities. Unfortunately, there is 
no written record of Hopkins' views of the 
immediate precursor of С virginica. But, to 
assume static environmental tolerances 
through time is tenuous (in an evolutionary 
sense) and is 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- 
tisslma is the only crassostreine oyster avail- 
able to serve as a direct forerunner for the 



American oyster. This giant, fossil oyster is 
widespread in the Western Hemisphere and 
its synonymy when completed (Lawrence, in 
preparation) should approach at length that of 
its related Eurasian species C. 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, 1 971 : 
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 (Alimón, 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 C. virginica and C. gigantisslma. 
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 hdgelets 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, Л 13, 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 & 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 & 
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 readouts 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, systennatic 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 solenlscus 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 (1 964: HI 0) for their original 
description. Chomata in adult С 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 oys- 
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, 
1 992). 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 is not a 
distinguishing character of the genus Cras- 
sostrea. 

Stiell Size and IVIassiveness: 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, 1 971 : 
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 C. 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, 1 985: 1 50; 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 С 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 С 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, 40, 42, 54, 71, 72, 
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, 1 971 : fig. Л 01 , 1 b, 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 С 
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 



UWVRENCE 



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 & 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 & Badve (1 979) 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. In- 
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 is 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, "con- 
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/thgonal 
valves of C. 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 С 
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 II (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 II 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 II stages, and he pointed out that 
Crassostrea starts the prodissoconch II 
phase with a relatively small size. In compar- 
ison to Ostrea, prodissoconch II growth in 
Crassostrea contributes more significantly to 
the overall shape of mature larvae. Because 
left valve convexity differences appear to be- 
gin with prodissoconch II 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; 
Carhker & Palmer, 1 979; Waller, 1 981 ). These 
same and distinctive geometries also apply 
to species assigned to Saccostrea (Dina- 
mani, 1973, 1976). Antero-postehor 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 (1 971 : N1 1 35) argued for separating Sac- 
costrea from Crassostrea because "species 
of these two genera cannot be made to 
crossfertilize each other." More recently 
Chanley & Dinamani (1980) have questioned 
the gametic incompatibility of taxa assigned 
to Crassostrea and Saccostrea and have 
noted (Chanley & 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 is 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 



UWVRENCE 



family. Also, because of accepted subfamilial 
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 a!. (1 979b: 1 79) used this ob- 
servation to suggest, following the fossil 
record, that "the Saccostrea genus is 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 & Avise, 1 992: 
102). 

In the future, expanded taxonomic and 
geographic studies of crassostreine mito- 
chondrial DNA (Reeb & 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 С 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 С virginica, while mem- 
bers of the lineage to which С virginica be- 
longs, in Panamic regions, still bore chomata 
on both left and right valves (Woodring, 1982: 
612, pi. 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 bore 
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), = C. virginica; 
Newball & 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 is 
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 paléontologie 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 SCAR[J\TO & 

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 & Dautzenberg, 1920: 

471 
Dioeciostrea Orton, 1928: 320 
Crasostrea Koch, 1929: 6 {nonn. null.), fide 

Stenzel, 1971: N1128 
Dioeciostraea Thiele, 1934: 814 {nonn. null.) 
*Saxostrea Iredale, 1936: 269 
*Striostrea Vyalov, 1 936: 1 7 
Angustostrea Vyalov, 1936: 18 
*Acutostrea Vyalov, 1936: 18 
Crassostrea Vyalov, 1948: 23 {nom. null.) 
Somalidacna Azzaroli, 1958: 115 
Crassotrea Miyake & Noda, 1962: 599 (nom. 

null.) 



198 



LAWRENCE 



*Sanostrea Miyake & Noda, 1962: 599 {nom. 

null.) 
*Gyrostrea Mirkamalov, 1963: 152 
Indostrea Chipionkar & Badve, 1976: 245 
*Bosostrea Chipionkar & Badve, 1978: 106 
*Cussetostrea Chipionkar & Badve, 1979: 

443 
*Soleniscostrea Chipionkar & Badve, 1979: 

444 
*Stnostrea (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. S. 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. I Thank 
E. V. Coan, B. С Coull, G. M. Davis, A. C. 
Lawrence, and three anonymous reviewers 
for thoughtful comments on this work. This is 
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 Bullía 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'Oligocène e il Miocene 
della Somalia, stratigrafia, tettonica, paleontolo- 
gía (microforaminiferi, coralli, molluschi). Paleon- 
tographica Itálica, new series, 22: 1-143, 36 pis. 

BORN, I., VON, 1 778, Index rerum naturalium musei 
Caesarei Vindobonensis, Pars I. ma, Testacea. 
Verzeichniss der natürlichen Seltenheiten des К. 
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. K. HERSHBERGER & K. K. 
CHEW, 1979a, Population genetics of the family 
Ostreidae. I. Intraspecific studies of Crassostrea 
gigas and Saccostrea commercialis. Marine Bi- 
ology. 54: 157-169. 

BUROKER, N. E., W. K. HERSHBERGER & K. K. 
CHEW, 1979b, Population genetics of the family 
Ostreidae. II. 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 genehc 
groups treated of and their American Tertiary 
species. Part IV: 1 , Phonodesmacea: Nucula to 
Julia. 2, Teleodesmacea: Teredo to Ervilia. Wag- 
ner Free Institute of Philadelphia, Transactions, 
3(4): 571-947, pis. 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: 95-1 07. 

DOLLFUS, G.-F. & P. DAUTZENBERG, 1920, Con- 
chyliologie du Miocène moyen du Bassin de la 
Loire, pt. 1 : Pelecypodes (Suite et fin). Société 
Géologique de France, Mémoires, 27, Paléontol- 
ogie, 22(2-4): 379-500, pis. 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, pis. 15-20. 

DURVE, V. S., 1986, On the ancestry and distribu- 
tion pathways of three species of Indian oysters. 
Indian Journal of Marine Sciences, 1 5: 56-58. 

FINCH, J., 1824, Geological essay on the Tertiary 
formations in America. American Journal of Sci- 
ence, 7: 31-43. 

FISCHER, P., 1 886, 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. 
Oehiert, Fascicule 10. Paris, F. Savy. Pp. 897- 
1008. 

FRENEIX, S., 1972, Les Mollusques bivalves cré- 
tacés du Bassin côtier de Tarfaya (Maroc mérid- 
ional). Notes et Mémoires du Service Gé- 
ologiques (Maroc), 228: 49-255. 

FRENEIX, S., 1982, Disparité microstructurale du 
test entre Gryphaeostrea et Gyrostrea (Huîtres 
du Crétacé-Miocène). Nouvelle classification 
des genres. Annales de Paléontologie, 68: 223- 
240. 

FUNK, V. A. & D. R. BROOKS, 1990, Phylogenetic 
systematics as the basis of comparative biology. 
Smithsonian Contributions to Botany, 73: 1-45. 

GALTSOFF, P. 8., 1964, The American oyster 
Crassostrea virginica Gmelin. U. S. Fish and 
Wildlife Service, Fishery Bulletin, 64: 1-480. 

GALTSOFF, P. S. & R. 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 décimo 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, S. C. 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, H. W. & D. T. DOCKERY, III, 1983, Notes 
on the lower Oligocène 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-Königlichen Geologischen 
Reichsanstalt, 4, pis. 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, 1 1 : 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, pis. 20-24. 

JAWORSKI, E., 1913, Beiträge zur Stammesge- 
schichte der Austern. Zeitschrift fur Induktive 
Abstammungs-und Vererbunglehre, 9: 192-215. 

JIMENEZ, A. P., J. С. BRAGA & J. M. MARTIN, 
1991, Oyster distribution in the Upper Tortonian 
of the Almanzora Corridor (Almería, SE Spain). 
Geobios, 24: 725-734. 

KARL, S. A. & J. С 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. 

UXMARCK, J. B. P. DE M., 1801, Système des an- 
imaux sans vertèbres, 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, H/sfo /ле naturelle 
des animaux sans vertèbres, 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 espèces qui s'y rapportent, 6(1): 195-220, 
Paris. 



U\URAIN, M., 1980, Crassostrea gryphoides et С 
gingensis (Schlotheim, 1813). Deux expressions 
morphologiques d'une même espèce (Miocène, 
Bivalvia). Geobios, 13: 21-43. 

U\WRENCE, D. R., 1968, Taphonomy and infor- 
mation losses in fossil communities. Geological 
Society of America, Bulletin, 79: 1315-1330. 

LAWRENCE, D. R., 1975, Paleoenvironmental set- 
ting of Crassostrea gigantissima (Finch) commu- 
nities, Coastal Plain of North Carolina. South- 
eastern Geology, 1 7: 55-66. 

LAWRENCE, D. R., in preparation, The oyster 
Crassostrea gigantissima (Finch) in the Cenozoic 
of the Western Hemisphere. 

LEA, I., 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 pis. 

LITTLEWOOD, D. T. J. & S. K. DONOVAN, 1988, 
Variation of Recent and fossil Crassostrea in Ja- 
maica. Paleontology, 31: 1013-1028. 

LOGAN, W. N., 1899, Some additions to the Cre- 
taceous invertebrates of Kansas. Kansas Univer- 
sity Quarterly, (A) 8(2): 87-98, pis. 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: Ptehomorphia) Ägyptens (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. В., 1 871 , 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, F. B. & F. V. HAYDEN, 1 857, 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 В group in 
the extracts of Mollusca. II. On vitamin B6, inosi- 



GENUS CRASSOSTREA 



201 



toi, pantothenic acid, biotin, and niacin. Bulletin 
of the Japanese Society of Scientific Fisfieries, 
28: 597-601. 

MOORE, E. J., 1 987, Tertiary marine pelecypods of 
California and Baja California: Plicatulidae to Os- 
treidae. United States Geological Survey, Pro- 
fessional Papers, 1228-C: 1-53, 34 pis. 

NELSON, T. С, 1938, The feeding mechanism of 
the oyster. I. 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. & M. R. CARRIKER, 1983, System- 
atic relationship of the oysters Crassostrea rhizo- 
phorae and С 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. & 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 pis. 

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, е., 1971, Morphología de la charnela 
larvaria de Crassostrea angulata (Lmk.) en difer- 
entes fases de su desarrollo. Investigación 
Pesqueras, 35: 549-563. 

PASCUAL, е., 1972, Estudio de las conchas lar- 
varias de Ostrea stentina Payr, y Ostrea edulis L. 
Investigación Pesqueras, 36: 297-31 0. 

PERKINS, В. F., 1969, Rudist morphology. In: R. С. 
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. & A. P. BROWN, 1917, Oligocène 
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 espèces d'huîtres vivants 
actuellement dans le monde, défines par leurs 
coquilles larvaires ou prodissoconques. Étude 
des collections de quelques-uns des grands mu- 
sées d'histoire naturelle. Revue des Travaux de 
l'Institut des Pêches Maritimes, 31: 127-199. 

REEB, 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, H. G. & A. HARBISON, 1942, Miocene 
invertebrate fauna of New Jersey. Academy of 
Natural Sciences of Philadelphia, Proceedings, 
94: 167-250, pis. 7-22. 

RUDWICK, M. J. S., 1 961 , 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, 1 molluschi dei terreni terziarii del 
Piemonte e della Liguria. Bollettino dei Musei di 
Zoología e di Anatomía Comparata della R. Uni- 
versita, Torino, 12(298): 99-102. 

SCARU\TO, O. A. & Ya. I. STAROBOGATOV, 
1 979a, The system of the suborder Mytileina (Bi- 
valvia). Pp. 22-25, in I. 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. A. & Ya. I. 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, с с. 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 Beiträge zur 
Naturkunde {B) 1: 1-57. 

SOHL, N. F. & 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 pis. 

STENZEL, H. В., 1947, Nomenclatural synopsis of 
supraspecific groups of the family Ostreidae 
(Pelecypoda, Mollusca). Journal of Paleontology, 
21: 165-185. 

STENZEL, H. В., 1959, Cretaceous oysters of 
southwestern North America. International Geo- 
logical Congress, 20th, Mexico, 1956, [Traba- 
jos], El Sistema Cretácico, 1: 15-37. 

STENZEL, H. В., 1971, Oysters. In: R. С 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- 
hne and Freshwater Research, 5: 132-168, pis. 
1-11. 



202 



LAWRENCE 



THUNBERG, С. P., 1793, Tekning och beskrifning 
pâ en stör Ostronsort ifrân Japan. Kungliga 
Svenska Vetenskapsakademiens Handlingar, 14: 
140-142. 

TORIGOE, K., 1981, Oysters in Japan. Journal of 
Science of Hirosfiima University, (В) [1 -Zoology] 
29: 291-419. 

VEATCH, J. О. & 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 l'Acad- 
émie des Sciences de l'URSS, Nouvelle Serie, 
4(13): 17-20. 

VYALOV, O. S., 1948, Printsiply klassifikatsii sem- 
eyistva Ostreidae. Lvovskoe Geologictieskoe 
Obslichestvo, Trudy, Paleontologictieskaya 
Seriya, 1: 30-40. 

WALLER, T. R., 1981, Functional morphology and 
development of veiiger larvae of the European 
oyster, Osfrea edulis Linne. Smithsonian Contri- 
butions to Zoology, 328: 1-70. 

WARD, L. W., D. R. LAWRENCE & B. W. BU\CK- 
WELDER, 1978, Stratigraphie revision of the 
middle Eocene, Oligocène, and lower Miocene- 
Atlantic Coastal Plain of North Carolina. United 
States Geological Sun/ey, 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 pis. 

WILKES, J., 1810, Encyclopaedia Londinensis; or, 
Universal dictionary of arts, sciences, and litera- 
ture. Vol. 5. London, J. Adiard. 

WILLIAMS, A. & D. J. POWELL, 1965, Morphology. 
In: R. С 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 Cuspidahidae; 
Additions to families covered in P 306-E; Addi- 
tions to gastropods; Céphalopode). United 
States Geological Survey, Professional Papers, 
306-F: 541-759, pis. 83-124. 

YONGE, С M., 1960, Oysters. London, Collins, 
209 pp. 

YONGE, С 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 



MAIJ\COLOGIA, 1995, 36(1-2): 203-208 



LETTER TO THE EDITOR 



CLARIFICATION AND EVALUATION OF TILLER'S (1989) 
STYLOMMATOPHORAN MONOGRAPH 

Kenneth C. Emberton^ & Simon Tillier^ 



When KE (Ennberton, 1991) performed an 
independent phylogenetic analysis of 17 of 
the 1 1 1 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 Rüdiger Bieler for his useful com- 
ments on an earlier draft. 

Timer's (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. ОС = 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. 1:0C). 

3. SC = gastric crop: 1 = cylindrical; 2 = 
median portion inflated; 3 = anterior region 
inflated; 2' = funnelform, widening from 
esophagus to stomach; = 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: PS). 

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 x right 
cerebral ganglion width (Fig. 2: CC). 

1 0. 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 Manns et de Malacologie, Muséum National d'Histoire Naturelle, 55 Rue Buffon, 
75005 Pans, FRANCE 



203 



204 



вм 





PS 



2' 




EMBERTON & TILLER 
ОС 






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 
CPD CPR 



205 




<w:n> 



•a 





PLG 




fcv3 



to) 







« 



fc^ 



FG 




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; 1 1 . 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 = from 1.5 to 2.5 (Fig. 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); = 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); = 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). 

1 7. 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 (SO) in semislugs could not be 
placed within a transition series, it was rede- 
fined as state and eliminated from phylo- 
genetic analysis. 

ST's "factor analysis" is not the statistical 
method known to most American workers as 
factor analysis, but the method of Benzéch 
(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 & 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, 
CPR, PLD, PLG); (b) apply arbitrary outpoints 
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 ОС, 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. 
Ill, 484, 512). 

As discussed more recently by Tillier & 
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 Delaftre 
(1988), which does not permit reversals, 



STYLOMMATOPHORAN PHYLOGENETICS 



207 



which may not find the most parsimonious 
tree(s), and which KE believes is a phenetic 
rather than cladistic method. Obviously, the 
more recent availability of more efficient cla- 
distic algorithms has made this part 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 by 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; (c) 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; (d) 
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. 

DELû^TTRE, P., 1988, Sur la recherche des filia- 
tions en Phylogenese. Revue Internationale de 
Systémique, 2: 479-504. 

EMBERTON, К. С, 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. 
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: 327-352. 

FARRIS, J. S., 1988, Hennig86, Version 1.5. James 
S. Farris, Port Jefferson Station, New York, NY 
11776. 

FÉNELON, J. P., 1981, Qu'est-ce que l'analyse des 
données? Lefonen, Paris: 311 pp. 

JAMBU, M. & M. 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- 
mones stylommatophores (Mollusca). Thèse de 
Doctorat d'Etat, Muséum 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., M. MASSELOT, H. PHILIPPE & A. 
TILLIER, 1992, Phylogénie moléculaire des Gas- 
tropoda (Mollusca) fondée sur le séquencage 
partiel de TARN ribosomique 28S. Comptes- 
Rendus de l'Académie des Sciences, Paris, série 
III, 314: 79-85. 

TILLIER, S., MASSELOT, M., GUERDOUX, J. & A. 
TILLIER, in press, Monophyly of major gastro- 
pod taxa tested from partial 28S rRNA se- 



208 



EMBERTON & TILLER 



quences, with emphasis on Euthyneura and hot- 
vent limpets Peltospiroidea. The Nautilus. 
TILLER, S. & W.F. PONDER, 1992, New species of 
Smeagol from Australia and New Zealand, with a 
discussion of the affinities of the genus (Gas- 
tropoda: Pulmonata). Journal of Molluscan Stud- 
ies. 58: 135-155. 



Revised Ms. accepted 1 January 1994 



The editor-in-chief of Malacologia welcomes let- 
ters that comment on vital issues of general im- 
portance to the field of Malacology, or that com- 
ment on the content of the journal. Publication is 
dependent on discretion, space available and, in 
some cases, review. Address letters to: Letter to 
the Editor, Malacologia, care of the Department 
of Malacology, Academy of Natural Sciences, 
19th and the Parkway, Philadelphia, PA 19103. 



MALACOLOGIA, 1995, 36(1-2): 209-215 



INDEX 



Page numbers in italics indicate figures of 
taxa. No new taxa appear in this number 
of IVIalacologia. 

Acavidae 50, 64 

Acavoidea 64 

Acavus 206 

A с ha tin a 68 

A с hat i ne I la 59 

Achatinellidae 66, 159 

Achatinellinae 155 

Achatinelioidea 66 

Achatinida 48, 63, 65 

Achatinoidea 63 

Acroptychia 69 

acuta, Physa 39, 84, 86, 87 

Acutostrea 187, 188, 191, 194, 197 

Adula (Botula) 1 

fa lea ta 5 
Aegopis 49 
Aequipecten 1 5 
afficta, Helix 121, 124 
af ficta, Helix (Helicigona) 112, 7 7 5, 1 24 
alabamiensis, Ost rea 1 93 
albicans, Littoraria 93 
a I bo I a bris, Neo he Их 65 
alexandrina, Biomphalaria 84 
Allodiscus granum 66 

и rq и hart i 66 
Allogona profunda 65 
Amastridae 159 
Amepelita xystera 69 
amnicum, Pisidium 30-32, 35, 37-39 
Ampelita 54, 64, 67 
A m pel it a (A m pel it a) lamarei 70, 72, 73, 

74 
Ampelita (Eurystyla) 72 

julii 70, 12-1 A^ 

soulaiana 70, 72-74 
Ampelita {Xystera) 71,72 

fulgurata 70, 72, 73, 75 

xystera 70, 70, 11-1 A 
Ampelita fulgurata 71, 72, 74, 75 

julii 69, 71, 72, 74, 75 

lamarei 71,72 

soulaiana 69, 71, 72, 75 

xystera 71, 72, 74, 75 
amphibulima, Helicophanta 70, 1^-1 A 
Amyqdalum 1 

anceyana, Leptachatina 1 61 -1 64 
anceyana, Leptachatina (Angulidens) 159 
Ancylus 39 
Ancylus fluviatilis 30, 31, 34, 36, 3-40, 

86, 145, 152 
andrewsae, Mesodon 172 
andrewsae, Mesomphix 181 
Anguispira 206 
Anguispira mord ax 65 



angula ta, Crassostrea 1 96 

Angustostrea 1 97 

apicina, Xero tricha 1 34 

Appalachina say ana 65 

appressa. Patera 65 

Archaeogastropoda 64, 66 

Archaeopulmonata 64 

A re ua tu la 5, 12 

Argopecten irradians 1 5, 23, 24 

a riel, Phrixgnathus 66 

Ariophantinae 64 

Ashfordia 1 34 

Assimineidae 63 

Athoracophorus bitentaculatus 66 

bar bata. Helix 1 1 5 

bar bata. Helix (Helicigona) 1 1 5 

Basommatophora 133 

beata, Ca na riel I a hispidula var. 119, 120, 

129 
beata, Caracollina 1 1 8 
beata, Helicodonta (Caracollina) 1 1 8 
beata. Helix (Caracollina) 1 1 8 
beata. Helix (Gonostoma) 112, 116, 118, 

119 
Belgrandia 3 9 
Belgrandiella 3 9 

bertheloti, Canariella hispidula 1 1 8 
bertheloti, Canariella hispidula var. 1 1 9, 

120, 128 
bertheloti. Helix 112, 116, 117-119 
bertheloti, Helix (Gonostoma) 1 1 7 
be rt he Iota, Helix 1 1 7 
Biomphalaria alexandrina 84 
Biomphalaria glabrata 84, 86, 87 
bitentaculatus, Athoracophorus 66 
Bithinella 1 49 
Bit hy nia 1 49 

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 

si m i la ris 5 9 
Bradybaenidae 133 
bu ce i ne I la, Caviella 66 
Buliminidae 63 
Buliminoidea 63 
Bull nus t ru ne at us 84 
Bythinella dunkeri 1 47-1 52 
eahobasensis, Ostrea 1 90 



209 



210 



INDEX 



Camaenidae 59 
Canariella 111-137 

discobole 1 2 1 

discobolus 121, 722, 123, 725, 131 

eutropis 122, 123, 725, 127-131, 130 

everia 118 

fortuna ta 1 1 8 

gomerae 122, 123-124, 725, 72S, 
131 

hispidula 115-122, 722, 131 

hispidula bertheloti 1 1 8 
var. beata 1 19, 120, 725 
var. bertheloti 119, 120, 725 
var. fortunata 119, 120, 725, 725 
var. hispidula 114, 11 9-1 20, 725 
vax. lanosa 119, 120, 725, 725 
war. subhispidula 119, 120, 725 

leprosa 118, 123, 722, 123, 725, 
126-127, 130, 131 

multigranosa 1 1 5 

planaria 116, 122, 123, 124, 750, 
131, 134 

pthonera 1 27 
caps ella, Paravitrea 65 
Caracollina beata 1 1 8 

everia 1 1 8 

gomerae 1 2 3 

planaria 1 24 
Cardium 23 
Carocolla hispidula 112, 113, 115, 116, 

118 

planaria 112, 7 7 5, 124 
Cary с hi um clappi 64 

nannodes 64 
Caseolus 1 34 
catascopium, Lymnaea 87 
Caucasocressa 1 34 
Caviella bu с ci ne Ha 66 

roseveari 66 
celinda, Therasiella 66 
Cepaea 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 
с/7/эГа, С///е//э 127 
С/7/е//э 132, 133 

c///aía 127 
Ciliellidae 132 
Ciliellinae 132 
Ciliellopsis 133, 134 
cinnamomea, Botula 5, 8 
Cionella morseana 64 
clappi, Carychium 64 
clappi, Vertigo 65 



С lav ato r 67 

moreleti 69, 70, 71, 73-75 
Clithon oualaniensis 97-109, 7 00 
Cochlicopoidea 64 
collisella, Ventridens 65 
columbiensis, Crassostrea 196 
Columella simplex 65 
cone av um, Haplotrema 65 
conella, Phrixgnathus 66 
contracta, Gastrocopta 65 
cookiana, Geminopora 66 
Corbicula fluminea 30-32, 35, 37-40 
cores ia, Délos 66 
corneus, Planorbarius 79-89 
coronadoi, Neohoratia 39 
corticaria, Gastrocopta 65 
С ras os t rea 1 97 
Crassos trea 185-202 
Crassostrea (Euostrea) 1 97 
Crassostrea angula ta 1 96 

columbiensis 196 

gigantissima 188, 755, 190-194 

g'/yas 1 96 

gryphoides 189, 196 

rhizophorae 1 97 

soleniscus 1 88 

virginica 24, 188-189, 196 
Crassostreinae 185-202 
Crass o trea 197 
С re tig en a 1 34 
cuccullata, Saccostrea 1 96 
cumberlandiana, Glyphyalinia 65 
cupreus, Mesomphix 65 
Cu pu leí la 49 

Cussetostrea 187, 188, 193, 194, 198 
Cyathopoma 63 
Cyclophoridae 48, 49, 63 
Cyclophoroidea 63, 66 
Cyrnotheba 1 34 
cytora, Cytora 66 
Cytora cytora 66 
Cytora torquilla 66 
darwinianus, Brachidontes 2, 5, 10 
Oe/os со res i a 66 

jeffreysiana 66 
demissus. Modiolus 1 
denotata, Xo lot rem a 65 
Dioeciostraea 1 97 
Dioeciostrea 1 97 
Discidae 49, 65 
discobola, Canariella 121 
discobolus, Canariella 121, 722, 123, 

725, 131 
discobolus, Helicodonta (Caracollina) 121 
discobolus, Helix 112, 121 
discobolus. Helix (Anchistoma) 121 
discobolus, Helix (Gonostoma) 121 
Discus nigrimontanus 65 

pa tu lus 65 
Dolichonephra 207 
Dreissena polymorphe 15-27 
dunkeri, Bythinella 147-152 



INDEX 



211 



duryí, 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 

Ellobioidea 64 

Endodontidae 155 

Enidae 63 

Enneinae 63 

e ri go ne, Phrixgnathus 66 

eta, IVIocella 66 

Euconulidae 49, 65 

Euconulus fulvus 65 

gaetanoi 161-163 
Euconulus (Nesoconulus) gaetanoi 159, 

160 
Eug ¡andina 5 9 
Eurystala 69 
eutropis, Canariella 122, 123, 125, 127- 

131, 130 
eutropis, 14 élis 112 
eutropis. Helix 122, 127 
eve ri a, Canariella 118 
everia, Caracollina 1 1 8 
everia, Helicodonta (Caracollina) 1 1 8 
everia, Helicodonta (Caracollina) hispidula 

118 
everia. Helix 112, 7 75, 118, 1 20 
everia. Helix (Anchistoma) 1 1 8 
Exogyrinae 186 
falca ta. Adula (Во tu la) 5 
Fectola infecta 66 

mira 66 

unidentata 66 
Ferrissia 39 

rivularis 145 
filosa, Littoraria 92 
Flammulina с hi ron 66 

pe г dita 66 
fluminea, Corbicula 30-32, 35, 37-40 
fluviatilis, Ancylus 30, 31, 34, 36, 3-40, 

86, 145, 152 
fon tina lis, Physa 86 
fortuna ta, Canariella 1 1 8 
fortuna ta, Canariella hispidula var. 119, 

120, 725, 725 
fortuna ta, Gonostoma 1 1 8 
fortunata, Helicodonta (Caracollina) 1 1 8 
fortúnate. Helix 112, 114, 7 75, 117, 119 
fortunata. Helix (Anchistoma) 1 1 8 
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 



gaetanoi, Euconulus (Nesoconulus) 159, 

160 
Gastrocopta contracta 65 

corticaria 65 

pentodon 65 
Gastrodonta interna 65 
Gastrodontinae 65 
Gasulliella 134 
Geminopora cookiana 66 
georgianus, Viviparus 87 
Georissa pure has i 66 
gigantissima, Crass os trea 188, 7 55, 190- 

194 
gr/gras. Crass os trea 1 96 
giveni, Phenacohelix 66 
g la bra ta, Biomphalaria 84, 86, 87 
Glyphyalinia cumberlandiana 65 

rimula 65 
gomerae, Canariella 122, 123-124, 725, 

72S, 131 
gomerae, Caracollina 1 23 
gomerae, Helicodonta (Caracollina) 1 23 
gomerae. Helix (Gonostoma) 112, 722, 

123 
G o пах is 1 33 
Gonostoma fortunata 1 1 8 

hispidula 1 1 8 
gouldi. Vertigo 65 
gracilis, Lamellidea 159 
gracilis, Lithophaga 5, 8 
graeca, Bithynia 139-146 
granosissimus. Modiolus demissus 9 
granum, Allodiscus 66 
Grassostrea 1 97 
greenv\/oodi, Rhytida 66 
Gryphaea 1 97 
Gryphaeidae 185, 186 
Gryphaeinae 186 
Gryphaeostreinae 186 
gryphoides, Crassostrea 189, 196 
Giy/e//a 63 
Guppya sterkii 65 
guyanensis. My te Ha 1 , 2 
Gyraulus 39 

Gyrostrea 187, 188, 192, 198 
Haines ia 54, 63 
Halolimnohelcinae 132 
Haplohelix 132, 134 
Haplotrema concavum 65 
Haplotrematidae 65 
hawaiiensis, Neso vitrea 159, 162-164 
hectori, Huanodon 66 
Helicarionidae 1 59 
Helicida 48, 64, 65 
Helicinoidea 64 
Helicodonta 1 34 

plana ria 1 24 

sa/fer/ 112, 7 7 5, 118, 119 
Helicodonta (Caracollina) beata 118 

discobolus 1 2 1 

everia 118 

fortunata 1 1 8 



212 



INDEX 



gomerae 1 23 
Helicodontidae 130, 132 
Helicodontoidea 1 30 
Helicoidea 130, 132 
Helicophanta 54, 64, 67 

amphibulima 70, 1^-1^ 
H élis eut ro pis 1 1 2 
Heiisoma duryi 84-86 

trivolvis 87 
Helix afficta 121, 124 

af ficta pi ana ri a 1 24 

bar bata 1 1 5 

bertheloti 112, 116, 117-119 

berthelotii 1 1 7 

discobolus 112, 121 

eutropis 122. 127 

everia 112, 7/5, 118, 120 

fortunata 112, 114, 116, 117, 119 

hispidula 115 

lanosa 127 

/e/7s 115 

/epA-osa 112, 126, 127 

plana n'a 1 24 

piñonera 112 
A/e//x (Anchistoma) discobolus 121 

everia 1 1 8 

fortunata 118 
/Уе//х (Caracollina) beata 1 1 8 
/Уе//х (Ciliella) lanosa 112, 117, 119, /22 

leprosa 126 
/Уе//х (Gonostoma) beata 112, 7 76, 118, 

119 

bertheloti 1 1 7 

discobolus 1 2 1 

gomerae 1 12, 722, 123 

hispidula 1 1 7 

рэл/у/ 1 1 2 
/Уе//х (Helicigona) afficta 112, 7 7 6, 1 24 

barba ta 1 1 5 

/e/7s 115 
/Уе//х (H i s pide I la) lanosa 1 1 7 

leprosa 1 26 
АУе//х (Macularia) plutonia 1 1 2 
АУе//х (Ochthephyla) multigranosa 1 1 2 
Helixarionidae 49, 64 
Helixarionoidea 64, 65 
Hendersonia occulta 64 
hispidula, Canariella 115-122, 722, 131 
hispidula, Canariella hispidula var. 7 74, 

119-120, 725 
hispidula, Carocolla 112, 113, 115, 116, 

118 
hispidula, Gonostoma 1 1 8 
hispidula, Helix 1 1 5 
hispidula, Helix {Gonostoma) 1 1 7 
hochsteteri, Liarea 66 
Но rati a 39 

Horatia sturmi 30, 31, 33, 36, 38-40 
Huanodon hectori 66 

pseudoleiodon 66 
Hydrocenidae 66 
Hydrocenoidea 66 



Hydromiidae 1 34 

Hygromia (Ciliella) lanosa 1 1 8 

leprosa 1 26 
Hygromiidae 111-137 
Hygromiinae 1 34 
Hygromioidea 132 
ide, Suteria 66 
im it at rix, Leptachatina 1 60 
Indostrea 187, 188, 191, 193, 194, 198 
infecta, Fee tola 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. De los 66 
jenkinsi, Potamopyrgus 30-35, 38-40 
/i;///, Ampelita 69, 71, 72, 74, 75 
/t////, Ampelita (Eurystala) 70, 12-1 û^ 
Kalidos 64, 69 
Kaliella 54, 64 
/r/V/, Serpho 66 
konaensis, Leptachatina 1 60 
konanensis, Succinea 159, 162, 163 
Konbostrea 187, 188, 192-194, 198 
Laevapex fuscus 145 
lamarei, Ampelita 71, 72 
lamarei, Ampelita (Ampelita) 70, 72, 73, 

74 
Lamellidea 159, 161-163, 166 

gracilis 1 59 

novoseelandica 66 

oblonga 1 5 9 

peponum 1 5 9 
lanosa, Canariella hispidula var. 119, 

120, 725, 725 
lanosa. Helix 1 27 
/эл705а, АУе//х (Ciliella) 112, 117, 119, 

722 
lanosa. Helix (Hispidella) 1 1 7 
lanosa, Hygromia (Ciliella) 1 1 8 
Laoma leimonias 66 

mariae 66 

marina 66 
lapidaria, Pomatiopsis 64 
lateumbilicata. Para laoma 66 
latissimus, Vitrizonites 65 
leimonias, Laoma 66 
/e/7s, f/e//x 1 1 5 
/eA7s, ^e//x (Helicigona) 1 1 5 
/eA7s, Lindholmiola 1 1 5 
/ер/£Уа, Leptachatina 160, 162, 163 
lepida, Leptachatina (Leptachatina) 1 59 
leprosa, Canariella 118, 123, 722, 123, 

725, 126-127, 730, 131 
leprosa, Helix 112, 126, 127 
leprosa. Helix (Ciliella) 1 26 
leprosa. Helix (Hispidella) 126 
leprosa, Hygromia (Ciliella) 126 



INDEX 



213 



Leptachatina 159, 160, 164 

anceyana 161-164 

i m ¡tat rix 1 60 

konaensis 1 60 

lepida 160, 162, 163 
Leptachatina (Angu/idens) anceyana 159 
L ep ta cha tina (Leptacha tina ) lepida 1 5 9 
Liarea hochsteten 66 
Liareidae 49, 66 
Liguus 5 9 
Limnoperna 1,11 

fortunei 5 
Lindholmiola lens 1 1 5 
Liostreinae 185, 188 
Lithophaga 1 2 

gracilis 5, 8 

nasuta 5 
Littoraria 97 

albicans 93 

//■/osa 92 

/>7 termedia 91-95 

luteola 93 

pallescens 92 

philippiana 92 

s cabra 93 

stngata 94 
Littonna 97, 98 
Littorinoidea 63 
Lophinae 185, 186 
/iv/s/, Pseudoamnicola 30, 31, 33, 36, 38, 

39 
luteola, Littoraria 93 
lymaniana, Pronesopupa 1 60 
Lymnaea catascopium 87 

obrussa 84 

palustris 145 

peregra 29-31,34,36,38-40,84 

stagnalis 23, 29, 40, 84, 87 

trunca tu la 39, 84 
Macrochiamydinae 64 
macula ta, Mocella 66 
major. Neo helix 49 
Malaga rion 64 
mariae, Laoma 66 
marina, Laoma 66 
Melanopsis 30-33, 39, 40 
meniscus, Striatura 160, 162, 163 
meniscus, Striatura (Pseudohyalina) 1 59, 

160 
Mercuria 3 9 
meridonalis, Striatura 65 
Mesodon andrewsae 172 

norm a I is 171-184 

zaietus 65 
Mesodontini 65 
Mesogastropoda 63, 64, 66 
Mesomphix andrewsae 181 

cupreus 6 5 

inomatus 65 

perlaevis 65 

subplanus 1 8 1 
Metafruticicola 1 34 



metcalfei, Modiolus 4, 5 
Microcystinae 64 
Microcystis 64 
/77/ЛЭ, Fee tola 66 
Mocella eta 66 

macula ta 66 
modiolus. Modiolus 5, 7 
Modiolus 1,12 

dem is s us 1 

demissus granosissimus 9 

metcalfei 4, 5 

modiolus 5, 7 

squamosus 9 

s tria tu lus 5 

undula tus 5, 8 
moellendorffi, Phrixgnathus 66 
Monacha 1 34 
Montserratina 1 33 
mordax, A ng и is pira 65 
moreleti, Clavator 69, 70, 71, 73-75 
morseana, Cionella 64 
Mulinia 1 8 1 

multidentata. Para vit rea 65 
multigranosa, Ca na rie I la 1 1 5 
multigranosa. Helix (Ochthephyla) 1 1 2 
Mus culis ta 1,10 

senhausia 5, 1 1 
Musculus 1,10 
Mytella charruana 1 , 4, 5, 1 0-1 2 

guyanensis 1 , 2 

speciosa 1 

strigata 1-14; 3, 5, 7 

tumbezensis 1 
M/íZ/ívs 1 , 11, 12 

ec/í7//s 2, 4, 5, 7-11, 22-24, 181 
nannodes, Carychium 64 
nasuta, Lithophaga 5 
nemo ral is, Cepaea 98, 180, 182 
Neohelix albolabris 65 

major 49 
Neo ho rati a 39 

coronadoi 3 9 

schuelei 39 
neozelandica, Therasiella 66 
Nesopupa subcentralis 162, 163 
Nesopupa (Infranesopupa) subcentralis 

159 
Nesovitrea hawaiiensis 159, 162-1 64 
nigrimontanus. Discus 65 
normalis, Mesodon 1 7 1 -1 84 
novoseelandica, Lamellidea 66 
Nu celia 97 

oblonga, Lamellidea 159 
obrussa, Lymnaea 84 
occulta, Hendersonia 64 
Oestophora 132 
Omphalo trop is 6 3 
Ortheurethra 64, 66, 206 
orycta, Pronesopupa 1 60 
Osfrea 196 
Ost rea alabamiensis 1 93 

cahobasensis 1 90 



214 



INDEX 



Ostreidae 185, 195 
Ostreinae 185, 186 
Otinidae 206 

oua /aniens is, Clithon 97-109, 700 
ova lis, Succinea 65 
Palaeolophidae 185, 186 
Palaeolophinae 1 86 
pallescens, Littorana 92 
palustris, Lymnaea 145 
Paralaoma lateumbilicata 66 

serratocostata 66 
Para vitrea caps ella 65 

multidentata 65 

subtil is 65 
parry i. Helix {Gonostoma) 112 
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 
peregra, Radix 147-152 
per la ev is, Mesomphix 65 
perna, Perna 1 1 
Perna 1 2 

решэ 1 1 

viridis 5, 8, 11 
Phenacohelix giveni 66 

p/'a/a 66 
philippiana, Littoraria 92 
Philonesia 159, 160, 162, 163 
Phrixgnathus a riel 66 

conella 66 

elaioides 66 

erigone 66 

moellendorffi 66 

poecilosticta 66 
P/7/sa 39 

aciyfa 39, 84, 86, 87 

fontinalis 86 
pilsbryi, Cha ropa 66 
P//ív/a 64 

piscinalis, Valvata 30-34, 38, 39 
Pisidium amnicum 30-32, 35, 37-39 
p/i//a, Phenacohelix 66 
planaria, Canariella 116, 122, 123, 124, 

730, 131, 134 
planaria, Caracollina 1 24 
planaria, Carocolla 112, 7 7 6, 1 24 
planaria, Helicodonta 1 24 
planaria. Helix 1 24 
planaria. Helix afficta 1 24 
Planorbarius 39 

CO meus 79-89 
planorbis, Planorbis 79-89 
Planorbis 29 

planorbis 79-89 
plutonia. Helix (Macularia) 1 1 2 
poecilosticta, Phrixgnathus 66 
Poecilozonites 49 



Poly gy relia 206 
Polygyridae 49, 65 
Polygyrinae 65 
Polygyroidea 65 
polymorpha, Dreissena 1 5-27 
Pomatiasidae 63 
Pomatiopsis lapidaria 64 
Po fa/77 op /rg'iys 3 9 

jenkinsi 30-35, 38-40 
profunda, Allogona 65 
Pronesopupa 159-163 

ly m a ni ana 1 60 

orycta 1 60 

se ri cata 1 60 
Pronesopupa (Sericipupa) 1 60 
Prosobranchia 63, 64, 66 
pseudanguicula, С ha ropa 66 
Pseudoamnicola 39 

/iv/s/ 30, 31, 33, 36, 38, 39 
pseudoleiodon, Huanodon 66 
Pseudoperna 187, 188, 191, 194, 197 
pthonera, Canariella 127 
pthonera. Helix 1 1 2 
pugetensis, Striatura 1 60 
Pulmonata 63, 64, 66 
Punctidae 49, 65, 66 
Punctoidea 64-66 
Punctum blandianum 65 
Pupillidae 159 
Pupilloidea 65 
pure has i , Georissa 66 
Pycnodonteinae 1 86 
Rachis 63 

/?эс^/х peregra 147-152 
recurvum, Ischadium 9 
rhizophorae, Crassostrea 1 97 
Rhytida greenv^oodi 66 
Rhytididae 66 
Rhytidoidea 65, 66 
rimula, Glyphyalinia 65 
Rissoidea 63, 64 
rivularis, Ferrissia 145 
roseveari, Caviella 66 
Roy bel Ha 49 

saccellus, Saccostrea 1 96 
Saccostrea 187, 750, 194-197 

cue cull at a 1 96 

saccellus 196 
Sas'i/a 206 

sa/fer/, Helicodonta 112, 7 7 6, 118, 119 
Sanos t rea 1 98 
Saxostrea 197 
sayana, Appalachina 65 
scabra, Littoraria 93 
Schileykiella 132-134 
schuelei, Neohoratia 39 
senhausia, Musculista 5, 1 1 
Sept i fer 1 1 

sericata, Pronesopupa 1 60 
Serpho kivi 66 
serrata, T herasiel la 66 
Sesarinae 64 



INDEX 



215 



Sigmurethra 63, 65 

similaris, Bradybaena 59 

simplex, Columella 65 

Sítala 64 

Soleniscostrea 187, 188, 194, 198 

soleniscus, Crassostrea 1 88 

solisianus, Brachidontes 2, 5, 10 

Somalidacna 1 97 

soulaiana, Ampelita 69, 71, 72, 75 

soulaiana, Ampelita {Eurystala) 70, 72-74 

speciosa, Mytella 1 

squamosus. Modiolus 9 

stagnalis, Lymnaea 23, 29, 40, 84, 87 

stenotrema, Stenotrema 65 

Stenotrema edvardsi 65 

stenotrema 65 
sterkii, Guppya 65 
Streptaxidae 50, 63 
Streptaxinae 63 
Streptostele 63 
s tri at и lus. Modiolus 5 
Stria tu га 159, 160 

meniscus 160, 162, 163 

meridonalis 6 5 

pugetensis 1 60 
Striatura (Pseudohyalina) meniscus 1 59, 

160 
strigata, Littoraria 94 
strigata, Mytella 1-14; 3, 6, 7 
Striostrea 187, 192-194, 197 
Striostrea (Parastriostrea) 1 98 
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, /25 
subhispidula, Helicodonta (Caracollina) 

hispidula 1 1 8 
subhispidula. Helix (Anchistoma) hispidula 

118 
subhispidula. Helix (Gonostoma hispidula 

119 
subhispidula. Helix (Gonostoma) hispidula 

112, 722 
subplanus, Mesomphix 181 
subtilis, Pa ravit rea 65 
Su bu lina 54, 63 
Subulinidae 49, 63, 65 
Succinea konanensis 159, 162, 163 

ova I is 65 
Succineidae 1 59 
Succineoidea 65 
S и te ha i de 66 
Szentgalia 1 34 

íe/7e//a, WfA/>7a 159, 161-163, 166 
tentaculata, Bithynia 29-34, 38, 39, 87, 

145, 147, 148, 150-152 
Thalassohelix ziczac 66 
Theodoxus 39 



Therasiella celinda 66 

neozelandica 66 

serrata 66 
7/7 ysanophora 206 
Tomatellaha 159 
Tornatellides 159, 160, 162, 163 
tridentata, Triodopsis 65 
Triodopsinae 65 
Triodopsini 65 
Triodopsis tridentata 65 

vulgata 65 
Trissexodon 1 3 2 
trivolvis, H el i so m a 87 
Tropidophora 54, 63, 69 
truncatula, Lymnaea 39, 84 
trunca tus, Bull nus 84 
tumbezensis, Mytella 1 
Turkostrea 1 88 
Tyrrheniella 133 
Tyrrheniellina 133, 134 
undula tus. Modiolus 5, 8 
unidentata. Fee to la 66 
(Ул/о 30-32, 35-40 
urquharti, Allodiscus 66 
Valvata piscinalis 30-34, 38, 39 
Vent rid ens со His ella 65 
Vertiginidae 48, 65 
Vertigo clappi 65 

gouldi 65 
Vicariihelicinae 1 32 

virginica, Crassostrea 24, 188-189, 196 
viridis. Pern a 5, 8, 11 
Vitreini 65 

V/fr/>7a fe/7e//a 159, 161-163, 166 
Vitrinoidea 65 
Vitrizonites latissimus 65 
Viviparus georgianus 87 
vulgata, Triodopsis 65 
Xanthonychidae 133 
Xenostrobus 1 , 5 
Xerotricha apicina 1 34 
Xolo trema denotata 65 
xystera, Amepelita 69 
xystera, Ampelita 71, 72, 74, 75 
xystera, Ampelita {Xystera) 70, 70, 72- 

74 
z a le tus, M es о don 65 
ziczac, Thalassohelix 66 
Zonitidae 49, 65, 159 
Zonitinae 65 
Zonitini 65 



VOL 36, NO. 1-2 1995 



MALACOLOGIA 



International Journal of Malacology 
Revista Internacional de Malacologia 
Journal International de Malacologie 
Международный Журнал Малакологии 
Internationale Malakologische Zeitschrift 



Publication dates 

Vol. 28, No. 1-2 19 January 1988 

Vol. 29, No. 1 28 June 1988 

Vol. 29, No. 2 16 Dec. 1988 

Vol. 30, No. 1-2 1 Aug. 1989 

Vol. 31, No. 1 29 Dec. 1989 

Vol. 31, No. 2 28 May 1990 

Vol. 32, No. 2 7 June 1991 

Vol. 33, No. 1-2 6 Sep. 1991 

Vol. 34, No. 1-2 9 Sep. 1992 

Vol. 35, No. 1 14 July 1993 
Vol. 35, No. 2 2 Dec. 1993 



VOL 36, NO. 1-2 MALACOLOGIA 1995 

CONTENTS 

R. ARAUJO, J. M. REMÓN, D. MORENO & M. A. RAMOS 

Relaxing Techniques for Freslnwater Molluscs: Trials for Evaluation of Different 
Methods 29 

JOST BORCHERDING 

Laboratory Experiments on the Influence of Food Availability, Temperature and 
Photoperiod on Gonad Development in the Freshwater Mussel Dreissena 
Polymorpha 15 

HEINZ BRENDELBERGER 

Dietary Preference of Three Freshwater Gastropods for Eight Natural Foods of 
Different Energetic Content 1 47 

L M. COOK & J. BRIDLE 

Colour Polymorphism in the Mangrove Snail Littoraria intermedia in Sinai 91 

KATHERINE COSTIL & JACOUES DAGUZAN 

Effect of Temperature on Reproduction in Planorbarius corneus (L.) and Plan- 

orbis planorbis (L.) Throughout the Life Span 79 

ROBERT H. COWIE, GORDON M. NISHIDA, YVES BASSET & SAMUEL M. GON, III 

Patterns of Land Snail Distribution in a Montane Habitat on the Island of 
Hawaii 1 55 

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 139 

KENNETH С EMBERTON 

Distributional Differences Among Acavid Land Snails Around Antalaha, Mada- 
gascar: Inferred Causes and Dangers of Extinction 67 

KENNETH С EMBERTON 

Land-Snail Community Morphologies of the Highest-Diversity Sites of Mada- 
gascar, North America, and New Zealand, with Recommended Alternatives to 
Height-Diameter Plots 43 

KENNETH С EMBERTON & SIMON TILLIER 

Clarification and Evaluation of Tillier's (1989) Stylommatophoran Mono- 
graph 203 

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) 97 

MIGUEL IBÁÑEZ, ELENA PONTE-LIRA & MARÍA R. ALONSO 

El Género Canariella Hesse, 1918, y su Posición en la Familia Hygromiidae 
(Gastropoda, Pulmonata, Helicoidea) Ill 

DAVID R. LAWRENCE 

Diagnosis of the Genus Crassostrea (Bivalvia, Ostreidae) 1 85 

ALAN E. STIVEN 

Genetic Heterozygosity and Growth Rate in the Southern Appalachian Land 

Snail Mesodon normalis (Pilsbry 1900): The Effects of Laboratory Stress 171 

MARÍA VILLARROEL Y JOSÉ STUARDO 

Morfología del Estomago y Partes Blandas en Mytella sthgata (Hanley, 1843) 
(Bivalvia: Mytilidae) 1 



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) 



MALACOLOGIA 



1995 



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. 



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. 

1 0. Use the metric system throughout. Mi- 
cron should be abbreviated ¡am. 

1 1 . 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 
V2 or Уз. Letters and numbers should be at 
least 3 mm 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 цт," not "1 63 цт"). 
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 be used sparingly, and 
vertical lines not at all. 

1 8. References cited in the text 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 type- 
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 in a museum where they 
may eventually be reidentified. 

21. Submit each manuscript in triplicate. 
The second and third copies can be repro- 
ductions. 

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 after publica- 
tion. 



REPRINTS AND PAGE COSTS 

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 in page proof, the cost is U.S. 
$3.00 or more. 

24. When an article is 10 or more printed 
pages long, MALACOLOGIA requests that an 
author pay part of the publication costs if 
grant or institutional support is available. 



SUBSCRIPTION PURCHASE 

25. Effective Jan. 1995, personal subscrip- 
tions are U.S. $36.00. Institutional, agency or 
dealership subscriptions are U.S. $55.00. 
Back issues and single volumes: $36.00 for 
individual purchaser; $55.00 for institutional 
or agency purchasers. There is a one dollar 
handling charge per volume for all purchases 
of single volumes. Address inquiries to the 
Subscription Office. 



VOL. 36, NO. 1-2 MAL7\C0L0GIA 1995 

CONTENTS 

MARÍA VIUJ\RROEL Y JOSÉ STUARDO 

Morfología del Estomago y Partes Blandas en Mytella strigata (Hanley, 1843) 
(Bivalvia: Mytilidae) 1 

JOST BORCHERDING 

Laboratory Experiments on the Influence of Food Availability, Temperature and 
Photoperiod on Gonad Development in the Freshwater Mussel Dreissena 
Polymorpha 15 

R. ARAUJO, J. M. REMÓN, D. MORENO & M. A. RAMOS 

Relaxing Techniques for Freshwater Molluscs: Trials for Evaluation of Different 
Methods 29 

KENNETH С EMBERTON 

Land-Snail Community Morphologies of the Highest-Diversity Sites of Mada- 
gascar, North America, and New Zealand, with Recommended Alternatives to 
Height-Diameter Plots 43 

KENNETH С EMBERTON 

Distributional Differences Among Acavid Land Snails Around Antalaha, Mada- 
gascar: Inferred Causes and Dangers of Extinction 67 

KATHERINE COSTIL & JACQUES DAGUZAN 

Effect of Temperature on Reproduction in Planorbarius corneus (L.) and Plan- 

orbis planorbis (L.) Throughout the Life Span 79 

L M. COOK & J. BRIDLE 

Colour Polymorphism in the Mangrove Snail Littoraria intermedia in Sinai 91 

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) 97 

MIGUEL IBÁÑEZ, ELENA PONTE-LIRA & MARÍA R. ALONSO 

El Género Canariella Hesse, 1918, y su Posición en la Familia Hygromiidae 
(Gastropoda, Pulmonata, Helicoidea) Ill 

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 — — 139 

HEINZ BRENDELBERGER 

Dietary Preference of Three Freshwater Gastropods for Eight Natural Foods of 
Different Energetic Content 147 

ROBERT H. COWIE, GORDON M. NISHIDA, YVES BASSET & SAMUEL M. GON, III 

Patterns of Land Snail Distribution in a Montane Habitat on the Island of 
Hawaii 1 55 

ALAN E. STIVEN 

Genetic Heterozygosity and Growth Rate in the Southern Appalachian Land 

Snail Mesodon normalis (Pilsbry 1900): The Effects of Laboratory Stress 171 

DAVID R. LAWRENCE 

Diagnosis of the Genus Crassostrea (Bivalvia, Ostreidae) 1 85 

KENNETH С EMBERTON & SIMON TILLIER 

Clarification and Evaluation of Tillier's (1989) Stylommatophoran Mono- 
graph 203 



3 2044