(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 



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 19 
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 



vwi_. Ol, iNw. I , 1989 



HARVARD 
UNIVERSITY 



MALACOLOGIA 



International 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 

Nineteenth Street and the Parkway 

Philadelphia, Pennsylvania 19103, U.S.A. 



EUGENE COAN 

California Academy of Sciences 

San Francisco, CA 



Co-Editors: 



Assistant Managing Editor: 

CARYL HESTERMAN 

Associate Editors: 



CAROL JONES 
Vasser College 
Poughkeepsie, NY 



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, President 
Museum of Comparative Zoology 
Cambridge, Massachusetts 

JOHN BURCH, Vice-President 

MELBOURNE R. CARRIKER 
University of Delaware, Lewes 

GEORGE M. DAVIS 
Secretary and Treasurer 

CAROLE S. HICKMAN 
University of California, Berkeley 



JAMES NYBAKKEN, President-Elect 
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 MALACOLOGICA Treasurer, UNITAS MALACOLOGICA 

Rijksmuseum van Natuurlijke Koninklijk Belgisch Instituut 

Historie voor Natuunwetenschappen 

Leiden, Netherlands Brüssel, Belgium 



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

ELMER G. BERRY, 
Germantown, Maryland 



Emeritus Members 

ROBERT ROBERTSON 

The Academy of Natural Sciences 

Philadelphia, Pennsylvania 

NORMAN F. SOHL 
U.S. Geological Survey 
Reston, Virginia 



Copyright © 1 989 by the Institute of Malacology 



1989 
EDITORIAL BOARD 



J. A. ALLEN 

Marine Biological Station 

Millport, United Kingdom 

E. E. BINDER 

Muséum d'Histoire Naturelle 

Genève, Switzerland 

A. J. CAIN 

University of Liverpool 
United Kingdom 

P. CALOW 

University of Sheffield 
United Kingdom 

A. H. CLARKE, Jr. 
Portland, Texas, U.S.A. 

B. С CLARKE 
University of Nottingfiam 
United Kingdom 

R. DILLON 

College of Charleston 

SC, U.S.A. 

C. J. DUNCAN 
University of Liverpool 
United Kingdom 

E. FISCHER-PIETTE 

Museum National d'Histoire Naturelle 

Paris, France 

V. FRETTER 
University of Reading 
United Kingdom 

E. GITTENBERGER 
Rijksmuseum van Natuurlijke Historie 
Leiden, Netherlands 

F. GIUSTI 
Université di Siena, Italy 

A. N. GOLIKOV 
Zoological Institute 
Leningrad, U.S.S.R. 

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. 

A. D. HARRISON 
University of Waterloo 
Ontario, Canada 

J. A. HENDRICKSON, Jr. 
Academy of Natural Sciences 
Philadelphia, PA, U.S.A. 

K. E. HOAGLAND 

Association of Systematics Collections 

Washington, DC, U.S.A. 

B. HUBENDICK 
Naturhistoriska Museet 
Göteborg, Sweden 

S. HUNT 

University of Lancaster 

United Kingdom 

R. JANSSEN 

Forschungsinstitut Senckenberg, 
Frankfurt am Main, Germany 
(Federal Republic) 

R. N. KILBURN 
Natal Museum 
Pietermaritzburg, South Africa 

M. A. KLAPPENBACH 

Museo Nacional de Historia Natural 

Montevideo, Uruguay 

J. KNUDSEN 

Zoologisk Institut & Museum 

Kobenhavn, Denmark 

A. J. KOHN 

University of Washington 

Seattle, U.S.A. 

Y. KONDO 

Bernice P. Bishop Museum 

Honolulu, Hawaii, U.S.A. 



A. LUCAS 

Faculté des Sciences 

Brest, France 

C. MEIER-BROOK 
Tropenmedizinisches Institut 
Tübingen. Germany (Federal Republic) 

H. K. MIENIS 

Hebrew University of Jerusalem 

Israel 

J. E. MORTON 

The University 
Auckland, New Zealand 

J. J. MURRAY, Jr. 
University of Virginia 
Charlottesville, U.S.A. 

R. NATARAJAN 
Marine Biological Station 
Porto Novo, India 

J. 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. 

W. F. PONDER 
Australian Museum 
Sydney 

R. D. PURCHON 

Chelsea College of Science & Technology 

London, United Kingdom 

Ql Z. Y. 

Academia Sínica 

Oingdao, People's Republic of China 

N. W. RUNHAM 

University College of North Wales 

Bangor, United Kingdom 



S. G. SEGERSTRÂLE 
Institute of Marine Research 
Helsinki, Finland 

G. A. SOLEM 

Field Museum of Natural History 

Chicago, U.S.A. 

F. STARMÜHLNER 

Zoologisches Institut der Universität 

Wien, Austria 

Y. I. STAROBOGATOV 
Zoological Institute 
Leningrad, U.S.S.R. 

W. STREIFE 
Université de Caen 
France 

J. STUARDO 
Universidad de Chile 
Valparaiso 

T. E. THOMPSON 
University of Bristol 
United Kingdom 

S. TILLIER 

Muséum National d'Histoire Naturelle 

Paris, France 

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

J. A. VAN EEDEN 
Potchefstroom University 
South Africa 

N. H. VERDONK 
Rijksuniversiteit 
Utrecht, Netherlands 

В. R. WILSON 

Dept. Conservation and Land Management 

Netherlands, Western Australia 

H. ZEISSLER 

Leipzig, Germany (Democratic Republic) 

A. ZILCH 

Forschungsinstitut Senckenberg 
Frankfurt am Main, Germany (Federal 
Republic) 



MALACOLOGIA, 1989, 31(1): 1-140 

AN ENDEMIC RADIATION OF HYDROBIID SNAILS FROM ARTESIAN SPRINGS 

IN NORTHERN SOUTH AUSTRALIA: THEIR TAXONOMY, PHYSIOLOGY, 

DISTRIBUTION AND ANATOMY 

By W.F. Ponder, R. Hershler*, and B. Jenkins, 
The Australian Museum, Sydney South, NSW, 2000, Australia 

CONTENTS 



INTRODUCTION 
The mound springs — a brief description 
Geomorphology and water chemistry 
Spring groups and complexes 
Climate 
MATERIALS AND METHODS 
Taxonomy 
Taxonomic rationale 
Materials 
Methods 
Characters 
Anatomy 
Physiology 
Materials 
Methods 
RESULTS 
Taxonomy 
Fonscochlea 

Fonscochlea (Wolfgangia) 
Trochidrobia 
Anatomy 
Anatomical description of 

Fonscochlea accepta 
Anatomical description of 
Trochidrobia punicea 
Physiology 
DISCUSSION 
Evolution and relationships of fauna 
Geological history 

Relationships of mound-spring inver- 
tebrates 
Evolution of species within mound 

springs 
Dispersal 

Environmentally-induced variation 
Ecology and behaviour 
Community structure 
Physiology 
Hydrobiid fauna 



Absence of fauna 

Conservation 
ACKNOWLEDGMENTS 
REFERENCES 
APPENDIX 1 

List of stations 

List of springs not sampled 

Stations at which no hydrobiids were 
collected 

Locality maps 
APPENDIX 2 

Tables of measurements 



ABSTRACT 

Artesian springs between Marree and 
Oodnadatta contain an endemic fauna of 
hydrobiid snails that have undergone an 
adaptive radiation in which habitat parti- 
tioning and size displacement are clearly 
evident. Ten new species in two new en- 
demic genera, Fonscochlea and Trochidro- 
bia, are described. Three of the species of 
Fonscochlea are divided into a total of six 
geographic forms, which are not formally 
named. Two geographic forms are restricted 
to single springs, the remainder being found 
in several springs, spring groups, or com- 
plexes of springs. Fonscochlea is divided in 
to two subgenera, Fonscochlea s.s. contain- 
ing five species and Wolfgangia with a single 
species. 

Both genera are represented in most 
springs, with up to five taxa present in single 
springs in the Freeling Springs Group and in 
some of the other springs in the northern part 
of the spring system. As many as four taxa 
are present in most other springs. The pat- 
tern of one or two sympatric species of Troch- 
idrobia, a large, amphibious species of 



•Present address, United States National Museum of Natural History, Washington, D.C., 20560 U.S.A. 

1 



PONDER, HERSHLER & JENKINS 



Fonscochlea, one large aquatic species of 
Fonscochlea and a small aquatic species of 
Fonscochlea is established in most of the 
springs in the area. Some of the factors lead- 
ing to the evolution and maintenance of this 
diversity are discussed. 

A subjective classification, based on shell, 
opercular and anatomical characters, was 
tested phenetically using discriminate analy- 
sis. 

Simple physiological experiments were 
carried out on some of the taxa to test for the 
effects of temperature, submergence, desic- 
cation, increased salinity, reduced dissolved 
oxygen, and responses to light. All taxa 
showed a wide range of tolerance to salinity 
and temperature but the small animals were 
more susceptible to desiccation than the 
large ones. Varying responses to light and 
submergence were obtained but all taxa 
showed reduced activity in deoxygenated 
water. 

The anatomy of the type species of both 
genera is described in detail. Fonscochlea is 
unique in having two equal-sized sperm sacs 
in the female that are probably derived from 
the bursa copulatrix and, as in Trochidrobia, 
which has a single sperm sac, the seminal 
receptacle is lost. 

The endemic snails, together with the un- 
usual endemic crustaceans sympatric with 
them, and their unusual community structure, 
give the springs special interest, both from 
the scientific and conservation viewpoints. 

Keywords: Mollusca, Hydrobiidae, springs, 
endemics, taxonomy, physiology, anatomy, 
speciation, sympatry, habitat partitioning 



INTRODUCTION 

The most nearly permanent type of water 
body in an arid environment is probably an 
artesian spring (Naiman, 1981). The habitat 
provided by an artesian spring in this situation 
is analogous to that of an island. Each spring 
is typically separated by arid land providing 
as marked a discontinuity of habitat as the 
sea does to terrestrial organisms. Artesian 
springs are typically permanent, within a mod- 
erate time scale, perhaps in the order of thou- 
sands to even millions of years for spring sys- 
tems but tens to hundreds of years for 
individual springs, and usually provide a rea- 
sonable diversity of habitats. Given these 
conditions one might expect genetic differen- 
tiation of populations in separate springs and 
some habitat partitioning allowing similar spe- 
cies to coexist. Studies of the faunas of ahd- 
zone artesian springs have sometimes re- 
vealed spectacular examples of speciation 



and habitat partitioning. The best docu- 
mented examples are of the fishes of the 
western deserts in the United States and 
northern Mexico (Minckley, et a!., 1986), par- 
ticularly of the Death Valley system (Soltz & 
Naiman, 1978). Studies of these fishes have 
provided insight into the nature of the speci- 
ation process (Turner, 1974; Soltz & Hirsh- 
field, 1981), biogeography relative to drain- 
age history (Hubbs & Miller, 1948; Hubbs et 
al., 1974; Smith, 1978) and adaptation to di- 
verse spring-fed habitats (Naiman & Soltz, 
1981). 

Natural water bodies in arid lands, such as 
springs, water in caves and marshes, are fre- 
quently refugia for relict biota. There are nu- 
merous examples, particularly amongst fishes 
and crustaceans, that are well documented. A 
spectacular example is the crocodiles In pools 
in the Ahaggar Mountains of Africa, now sur- 
rounded by vast desert areas (Cole, 1968). 
Springs sometimes support diverse faunas 
that might be partly relictual and partly en- 
demic radiations. The hydrobiid snails of the 
Cuatro Ciénegas Basin, Coahuila, Mexico, are 
presumably an example of such a fauna (Tay- 
lor, 1966a; Hershler, 1984, 1985). 

Radiations of hydrobiid snails In springs in 
temperate climates are also known, examples 
including those in Florida (Thompson, 1968) 
and parts of Europe (e.g., Radoman, 1983). A 
spectacular radiation of the related family Po- 
matiopsidae in Southeast Asia has been well 
documented by Davis (1979). 

Bayly and Williams (1973) note that ex- 
tremely little is known about the biology of 
Australian springs. This is certainly true for 
the artesian springs associated with the Great 
Artesian Basin. Before this study commenced 
the only animals that had been studied in de- 
tail in artesian springs in arid Australia were 
the fishes (Glover & Sim, 1978a; Glover, 
1982). Recent biological work is summarised 
by Ponder (1986). 

The artesian springs in the arid north of 
South Australia (Figs. 1 , 2) were only recently 
shown to contain a large and interesting biota 
(Mitchell, 1985; Symon, 1985; Ponder, 1985, 
1986). To date the only invertebrates de- 
scribed from these mound springs are a 
phreatoicid isopod {Phreatomerus ¡atipes 
(Chilton, 1 922), an ostracode, Nagarawa dirga 
(DeDeckker, 1979), and a macrostomid flat- 
worm, the first record of this order from Aus- 
tralia (Sluys, 1986). Both of the Crustacea are 
endemic to the springs and belong in mono- 
typic subfamilies. 



AUSTRALIAN SPRING HYDROBIIDS 


















FIG. 1. Various springs in the Lake Eyre Supergroup showing some of the morphological diversity. 

A. Blanche Cup Spring (Stns 8-12), a conical, calcareous mound spring with a crater-like pool. 

B. Aerial view of part of Hermit Hill Spring Complex showing part of a spring group (Finniss Swamp West) 
composed of small ground-level springs and some low sand mounds. 

С Almost extinct mound in the Blanche Cup Complex, in the Horse Spring Group (stn 748). Snails and 

crustaceans are abundant in small seeps such as this. 

D. The Bubbler Spring (stns 13-17), one of the largest flows in the Lake Eyre Supergroup. 



PONDER, HERSHLER & JENKINS 





j)'^ 






?4,j 




AUSTRALIAN SPRING HYDROBIIDS 



Gastropod molluscs were reported from the 
mound springs by Mitchell (1980, unpub- 
lished; 1985) who, on the advice of Dr. B. 
Smith, to whom the material was sent for iden- 
tification, recognized the presence of three or 
possibly four species referable to three or four 
genera. DeDeckker ( 1 979) also refers to these 
snails as undescribed endemics, on Smith's 
advice. We cannot find any earlier references 
to these species in the literature, despite their 
being conspicuous and abundant in most of 
the springs. A few of the early explorers no- 
ticed the small fish found in some springs (see 
review by Glover & Sim, 1978b). 

Some of the more accessible mound 
springs were visited in the latter part of the 
1970's by several biologists who made some 
collections, those of W. Zeidler of the South 
Australian Museum being the most signifi- 
cant. His collections and those sent to Dr. B. 
Smith were made available to one of us 
(W.F.P.) and field work was carried out in 
1981 by W.F.P. and Zeidler. The result of that 
field investigation, and an additional one the 
same year by Zeidler, showed the existence 
of an apparent endemic fauna of hydrobiid 
snails of considerable diversity. 

The available information on the mound- 
spring fauna was reviewed in an Environmen- 
tal Impact Statement (E.I.S.) for the Olympic 
Dam Project (Kinhill-Stearns Roger, 1982) 
and in a supplement to this E.I.S. (Kinhill- 
Stearns, 1983). The review in the supplement 
included some new information on the hydro- 
biid snails provided by two of us (W.F.P., 
B.W.J.). Because the Olympic Dam Project 
required water from a borefield located near a 
large spring complex at Hermit Hill (Fig. 2; 
Appendix 1, Fig. 62), further biological and 
hydrological studies were carried out to as- 
sess the importance of the flora and fauna 
associated with these springs. This paper has 
been developed from the report resulting from 
those studies. A summary of the results of the 
hydrobiid work appears in the report prepared 
for Roxby Management Services on the 
mound springs (Ponder & Hershler, 1984). 

The importance of the springs and the need 
for their conservation has been stressed by 
Casperton (1979), Harris (1981), Symon 
(1985) and Ponder (1985, 1986). This view 
has also been strongly supported by the evi- 
dence accumulated in the reports prepared 
as a result of the Olympic Dam Project (Kin- 
hill-Stearns Roger, 1982, Kinhill-Stearns, 
1983, 1984). The World Wildlife Fund has re- 
cently provided funds to fence some springs. 



The snails present in the mound springs 
are members of the Hydrobiidae, a world- 
wide family of prosobranch gastropods that 
are part of the large, predominantly marine 
superfamily Truncatelloidea. The hydrobiids 
were probably derived from brackish-water 
ancestors in the middle part of the Mesozoic 
(Ponder, 1988) and some members of the 
family are still restricted to brackish-water en- 
vironments. To date the family is known to be 
represented in Australia by about nine genera 
and approximately 35 named species, ex- 
cluding those from the mound sphngs, al- 
though recent unpublished work by W.F.P. 
shows that this fauna is actually much larger. 

The adaptations of organisms to the diverse 
and often extremely harsh aquatic environ- 
ments in deserts are of interest to physiolo- 
gists as well as ecologists and evolutionary 
biologists. While a variety of taxa are usually 
found in such waters, only the desert fishes 
are well studied in terms of their ecology and 
physiology (see summaries. Deacon & Minck- 
ley, 1974; Soltz & Naiman, 1978; Naiman & 
Soltz, 1981). In areas in which hydrobiid 
snails have radiated extensively in desert wa- 
ters, particularly spring systems of North and 
Central America (Taylor, 1966a, b; Hershler, 
1 985; Hershler & Landye, 1 988) and Australia 
(Ponder, 1986), their frequent local diversity 
and high densities suggest that they are 
trophically important members of desert 
aquatic communities. Yet there is a paucity of 
data concerning their ecology and virtually 
nothing is known of their physiology. Toler- 
ances to the environmental parameters that 
often achieve extreme levels in desert waters 
(e.g., salinity, temperature), have not been 
studied for any spring-dwelling hydrobiid spe- 
cies, although some work on South African 
species of Tomichia, of the related family Po- 
matiopsidae, has been done (Davis, 1981). 

This paper commences with an introduc- 
tory section outlining the main features of the 
mound springs. The rest of the paper is di- 
vided into three sections. The first deals with 
the taxonomy of the hydrobiid snails, followed 
by a detailed account of the anatomy of the 
type species of the two genera found in the 
springs. The results of the physiological work 
done in the field are presented in the third 
section. 

The mound springs — a brief description 

Geomorphology and water chemistry: The 
artesian mound springs of South Australia are 
aligned in an arc running from the far northern 



PONDER, HERSHLER & JENKINS 



part of the state at Dalhousie Springs, north of 
Oodnadatta, around the south of Lake Eyre to 
Lake Frome and Lake Callabonna on the 
eastern side of the Flinders Ranges. Addi- 
tional artesian springs are found in western 
Queensland and were found in the north-west 
of New South Wales, but these are now 
mostly extinct (personal observations by 
W.F.P. and M.A. Habermehl, pers. comm.), 
presumably as a result of water extraction 
from the basin by the pastoral industry. The 
springs are natural discharges from the aqui- 
fers formed from the Jurassic and Cretaceous 
sedimentary rocks of the Great Artesian Ba- 
sin (see Habermehl, 1980, 1982, for geolog- 
ical details). They occur in a variety of forms, 
the most common being small mounds result- 
ing from groundwater precipitates, mainly car- 
bonates, and fine sediments derived from the 
aquifer and confining beds. Wind-blown de- 
bris and plant material also contribute to the 
mound formation. The mounds are composed 
primarily of hard travertine or of sediment, or 
layers of both. They range from virtually flat to 
large mounds several tens of meters high. 
The larger mounds are the older springs, the 
ground-level springs the youngest (Ponder, 
1986: Fig. 4). More detailed descriptions of 
the springs are provided by Watts (1975), 
Habermehl (1982), Thomson and Barnett 
(1985), and Ponder (1986). The South Aus- 
tralian mound springs are the most active and 
numerous of the artesian springs fed by the 
Great Artesian Basin (Habermehl, 1982) and 
are now the best known biologically. The little 
that is known of Queensland artesian springs 
is summarised by Ponder (1986). 

Dalhousie Springs, to the north of Qodna- 
datta, yields about 95% of the natural dis- 
charge from the Great Artesian Basin in 
South Australia (Williams, 1979; Williams & 
Holmes, 1978). These springs are, however, 
outside the present study area, as are some 
small springs east of Marree to the north and 
east of the northern Flinders Ranges. Some 
of these springs contain endemic inverte- 
brates, including hydrobiids, and these will be 
dealt with elsewhere. The springs included in 
this report (Fig. 2; Appendix 1) are located 
mainly on the Warrina, Billa Kalina and Cur- 
dimurka 1 :250,000 map sheets and a few on 
the Oodnadatta sheet. They form a zone 
about 400 km long and as much as 20 km 
wide between Marree and Oodnadatta (Fig. 
2) and are referred to as the Lake Eyre group 
by Habermehl (1982) and the Lake Eyre Su- 
pergroup by Ponder (1986). 



The morphology of the springs in the Lake 
Eyre Supergroup is diverse (Fig. 1). The 
springs range from surface seeps (Fig. lb) to 
low, conical mounds (Fig. la, c) or even small 
hills. The mounds consist of sand, silt and 
clay, often cemented by carbonate and over- 
lain by layers of carbonate (Habermehl, 1980, 
1982). The cemented mounds often persist 
for considerable periods after the springs that 
formed them have ceased to flow, but the un- 
consolidated mounds erode rapidly. Some 
mounds have a crater-like, water-filled de- 
pression at the top (Fig. la), while others 
have rounded domes (Fig. 1c); both types 
typically have one or more outlets. Some of 
the larger, dome-like mounds (e.g., Kewson 
Hill and the Elizabeth Springs mound) have 
several small seeps issuing from them. 

Discharges from most of the springs are 
small, ranging from about 0.5 litre per sec- 
ond to 7.5 litres per second at the Bubbler 
Spring (Fig. Id) (Cobb, 1975; Williams, 1979; 
Habermehl, 1982). Despite this, discharge 
from some springs is sufficient to maintain 
flows for several hundred metres or, more 
rarely, a kilometre or more, providing a well- 
vegetated wetland habitat. Other springs 
have such a small discharge that they do not 
maintain an outflow, having only a pool or 
small swampy area at the head. Others are 
merely permanently damp patches that might 
flow occasionally. Some small springs in the 
Hermit Hill complex (Fig. lb) have been ob- 
served flowing on some occasions and are 
dry on others. The Lake Eyre Supergroup has 
a total estimated discharge of 100-200 litres 
per second (Habermehl, 1982), compared 
with 670 litres per second for Dalhousie 
Springs (A.F. Williams, 1974; Williams & 
Holmes, 1978). 

The depth of the water in the pools and 
outflows rarely exceeds 2-3 cm and is usu- 
ally only a few millimetres. The pools and out- 
flows usually contain sedges but rarely true 
aquatic vegetation apart from algae. The out- 
flows are usually narrow trickles with a firm, 
sandy base and, in the case of the hard 
mounds, calcareous rock. 

Our observations indicate that the area of 
outflow diminishes in summer, presumably 
owing to increasing evaporation, and some 
observations suggest that periods of high 
barometric pressure coincide with reduced 
water flow (C. Woolard, pers. comm.). 

Williams and Holmes (1978) have esti- 
mated that a spring with a small discharge 
typical of many of the springs in the Lake Eyre 



AUSTRALIAN SPRING HYDROBIIDS 



Supergroup, shown on the Curdlmurka map 
sheet, would take about 1 000 years to deposit 
sufficient calcium carbonate to build a hemi- 
spherical mound three metres high. On this 
basis some of the larger mounds, such as 
Kewson Hill, might, even with substantially in- 
creased flow rates, take several tens of thou- 
sands of years to form. Forbes (1961) has 
shown, however, that drilling on mounds in 
this vicinity reveals that a substantial portion 
of the mound is formed by the deposition of 
sand and clay rather than "limestone", sug- 
gesting that the calculations by Williams and 
Holmes (1978) might be invalid. 

Analyses of the water from the springs in 
the Lake Eyre Supergroup have been given 
by Cobb (1975), Williams (1979) and Kinhill- 
Stearns (1984) and summarized by Haber- 
mehl (1982). Sodium and bicarbonate are the 
major ions in springs in the eastern part of the 
Lake Eyre group whereas in springs in the 
western part the bicarbonate component is 
small and sodium and chloride ions predom- 
inate over calcium and sulphate. Total dis- 
solved solids in most springs range from 
2000-4000 ppm, with a few in excess of 5000 
ppm, and pH from about 7.1 to 8.1, although 
a field pH of up to 9.95 has been recorded in 
recent surveys. The temperatures in the 
spring vents are constant throughout the year 
and show a slight increase from east to west 
ranging from upper teens to mid-twenties (°C) 
in the east to upper twenties in the west. The 
salinity increases toward the discharge areas 
of the Great Artesian Basin. 

A few springs in the Lake Eyre Supergroup 
might not originate from the waters of the 
Great Artesian Basin aquifer, or show signif- 
icant mixing with sulphate-rich ground-water, 
as their hydrochemistry is atypical. These 
springs are located on the faulted edge of the 
basement rocks and include Kerlatroaboorn- 
tallina Spring, Taitón Springs, Edith Spring, 
Dead Boy Spring and Pigeon Hill Springs, the 
last two in the Hermit Hill Complex. None of 
these springs contains the typical mound 
spring invertebrates. 

Exploitation of the water from the Great Ar- 
tesian Basin has resulted in a drop of the Po- 
tentiometrie surface by several tens of metres 
in heavily developed areas (Habermehl, 
1980). Even by the turn of the century the 
sinking of bores near some springs had 
greatly reduced or extinguished their flow 
(Pittman & David, 1903). 

At present, a new steady-state condition 
appears to have been reached in which total 



recharge and discharge are approaching 
equilibrium again (Habermehl & Seidel, 1979; 
Habermehl, 1980), and little change is ex- 
pected to occur in the discharge rates of the 
springs provided no new well development 
takes place. 

Spring groups and complexes: The mound 
springs in the Lake Eyre Supergroup are not 
distributed evenly and for the purposes of this 
report can be divided into several major 
sphng complexes. Within each of these com- 
plexes spring groups can be identified. A 
spring complex can be defined as a large 
cluster of springs separated from adjacent 
spring clusters by several tens of kilometres. 
Smaller groups of springs, either within a 
complex or an isolated group, can be referred 
to as spring groups. For example. Hawker 
Springs can be called a spring group within 
the Mt. Margaret Spring Complex. In the Her- 
mit Hill Spring Complex there are several 
spring groups, e.g., Finniss Swamp West ( = 
West Finniss), Hermit Hill Springs Proper and 
Old Woman Springs. The following classifica- 
tion of spring complexes in the Lake Eyre Su- 
pergroup is essentially that proposed by Kin- 
hill-Stearns (1984) (Fig. 2). Table 1 lists the 
springs, grouped in complexes, containing 
hydrobiids. 

To facilitate discussion we have arranged 
these spring complexes into seven informal 
systems (Fig. 2), the arrangement being bi- 
ased towards the distribution of the hydrobiid 
fauna. Detailed maps for each spring area are 
given in Appendix 1 and these are referred to 
in the list below. 

1 . The Oodnadatta Springs. 

Mt. Dutton Spring Complex. The few small 
springs on the Oodnadatta Map Sheet that lie 
southeast of Oodnadatta (Appendix 1, Fig. 
63). 

2. The Freeling Springs: 

The Peake Hill Spring Complex. Includes the 
Freeling Springs and a few small springs to 
the north and northwest of Mt. Denison (Ap- 
pendix 1, Figs. 58, 63B). 

3. The Northern Springs: 

Mount Margaret Spring Complex. Includes 
the large, scattered group of springs to the 
east of Mt. Margaret, as well as the Peake 
and Denison Ranges (Appendix 1, Fig. 59). 

4. The North Western Springs: 

a) Nilpinna Spring Complex. A few scattered, 
small, springs to the west of the Marree- 
Oodnadatta Road and west of the Mt. Marg- 
aret Spring Complex (Appendix 1, Fig. 58). 



8 PONDER, HERSHLER & JENKINS 

TABLE 1. Distribution of taxa in springs and spring complexes, x = present (living), s = shells only 





< 

E 


m 
E 


< 

E 
о 


CD 

E 
о 


< 

E 
о 


m 
E 
Ô 


о 

E 
о 


< 

E 
о 


CQ 

£ 
о 


о 

E 
о 


m 
















TD 


T3 


.2 
1 


2 




Ч 


я 

ч 


■S 


¡л 

s 


5 


1 


2 


с 




Я 


Я 
5" 






(11 


(Il 


CT 


Ü- 


о 


u 


о 


m 


СП 


(Я 




о 


J 


E 








N 


N 


« 


ю 


10 


ю 


ю 












a 


w 






SPRING OR SPRING GROUP 


u: 


u: 


u: 


LL 


u: 


UL 


u: 


u: 


u: 


LL 


ll 


ll 


h-; 


K-: 


h-; 


h-: 


SPRING COMPLEX 


Southern Springs 




































Welcome group 


X 








X 














X 


X 








Wangianna 


Davenport group 


X 








X 














X 


X 








Spring Complex 


Old Woman group 


X 










X 












X 


X 










West Finniss group 


X 










X 












s 


X 










Hermit Springs group 


X 










X 














X 










Old Finniss group 


X 










X 














X 








Hermit Hill 


Dead Boy Spring 












X 














X 








Spring Complex 


Sulphuric group 












X 














X 










Bopeechee Spring 












X 














X 










Venable Spring 


s 










s 












s 


s 










Priscilla Spring 


s 










s 












s 


s 








Lake Eyre 


Centre Island Spring 


s 
































Spring 


Emerald Spring 














X 




















Complex 


Middle Springs 





































Horse East group 
Horse West group 
Strangways Spring 
Mt. Hamilton Spring 
Blanche Cup group (785, 787) 
Blanche Cup Spring 
Blanche Cup group (786) 
Little Bubbler Spring 
The Bubbler Spring 



Blanche Cup 

Spring 

Complex 



Coward Springs Railway Bore 
Coward Springs group 
Kewson Hill group 
Julie group 
Elizabeth group 
Jersey group 



Coward Spring 
Complex 



Warburton group 
Beresford group 


X 
X 


X 
X 


X 
X 


X 
X 


Beresford Spring 
Complex 


South-Western Springs 
Strangways group 


X 


X 


X 


X 


Strangways Spring Complex 


Billa Kalina group 
Fenced Spring 
Welcome Bore Spring 


X 
X 

s 


X 
X 


X 
X 

s 


X 
X 


Old Billa Kalina 

Spring 

Complex 


Margaret Spring 
Francis Swamp group 
Loyd Bore spring 


s 

X 
X 


s 

X 
X 


s 

X 
X 


s 

X 
X 


Francis Swamp 

Spring 

Complex 



Northern Springs 
Bhnkley Spring 
Hawker group 
Twelve Mile group 
Outside group 
Fountain group 
Big Perry Spring 
Spring Hill Spring 



Mt. Margaret 

Spring 

Complex 



Freeling Springs 

Freeling group 

North of Freeling Spring 


x 


X 


X 


X 
X 


X Peake Hill Spring 
Complex 


Oodnadatta Springs 
Big Cadnaowie 


X 








Mt. Dutton Spring Complex 



AUSTRALIAN SPRING HYDROBIIDS 




MONTH 

FIG. 3. Temperature and evaporation data for Marree and Oodnadatta. Mean daily evaporation, for each 
month, is given only for Oodnadatta, 1968-1982. The temperature data are for Marree, 1957-1982, and 
Oodnadatta, 1940-1982 together (with error bars indicating the range that the two encompass) and consists 
of number of days/month with temperatures >40° C, number of days/month with temperatures >35° C, 
number of days/month with temperatures <0° C, and number of days/month with temperatures <2.2° С 



b) Lake Cadibarrawlrracanna Spring Com- 
plex. A widely scattered group of springs west 
of William Creek; the most westerly of all the 
spring complexes (Fig. 2; Appendix 1, Fig. 
58). 

5. The South Western Springs: 

a) Francis Swamp Spring Complex. A large 
group of springs south of William Creek (Ap- 
pendix 1, Fig. 60). 

b) Old Billa Kalina Spring Complex. A scat- 
tered group of springs south of Francis 
Swamp on the northern side of f\/largaret 
Creek (Appendix 1, Fig. 60). 

c) Strangways Spring Complex. A compact 
group of mostly extinct carbonate mounds to 
the east of Francis Swamp (Appendix 1, Fig. 
59). 

6. The Middle Springs: 

a) The Beresford Spring Complex. Two main 
springs associated with two very large, extinct 
mounds, North and South Beresford Hills (Ap- 
pendix 1, Figs. 60, 61). 

b) Coward Spring Complex (Appendix 1 , Fig. 
61) includes the springs between Coward 
Springs and Hamilton Hill. 



7. The Southern Springs: 

a) Lake Eyre Spring Complex. A few springs 
on the southern and southwestern sides of 
Lake Eyre South and on islands in this lake 
(Appendix 1, Figs. 61, 62). 

b) Hermit Hill Spring Complex. Several large 
groups of springs in the vicinity of Hermit Hill 
(Appendix 1, Fig. 62). 

c) Wangianna Spring Complex. Includes the 
Welcome and Davenport Spring Groups, as 
well as the degraded Wangianna Spring (Ap- 
pendix 1, Figs. 62, 63B). 

Climate: Basic meteorological data for this 
region are presented in Fig. 3. Note the fre- 
quency of summer days with >40° С temper- 
atures. Annual rainfall at Marree varied from 
39.3-379.9 mm for the 21 years between 
1957-1982, and at Oodnadatta from 54.3- 
465.8 mm for the 20 years between 1958- 
1982. Evaporation is exceedingly high, usu- 
ally >10mm/day (Fig. 3) and, for a given year, 
typically exceeds precipitation by a factor of 
10 or more (data for Oodnadatta and Marree 
were provided by the Bureau of Meteorology). 



10 



PONDER, HERSHLER & JENKINS 



MATERIALS AND METHODS 
Taxonomy 

Taxonomic rationale: Because the mound 
springs are isolated from one another, each 
population has the potential to contain a 
unique genome that, given sufficient time, iso- 
lation and selective pressure, could develop 
into separate taxa. It was impractical to ana- 
lyse all populations but a representative, non- 
random selection (Appendix 2, Tables 18-21) 
was made and these populations were 
treated as separate units in the statistical 
analyses to prevent bias towards the initial 
subjective split into species units. 

The method that we have used to distin- 
guish taxa is essentially phenetic. The phe- 
netic grouping of populations by discriminate 
analysis is used as an aid for recognizing taxa 
but because strict acceptance of phenetic 
classifications, we believe, can be mislead- 
ing, a subjective element was also intro- 
duced, generally on the side of caution. The 
rather large number of characters measured 
were statistically tested for differences be- 
tween the recognised taxa. Most taxa are dis- 
tinguished by at least one major set of char- 
acters (e.g., opercular, shell or reproductive) 
that are statistically significantly different (p < 
0.01) from the phenetically closest taxon. It is 
lur belief that the classification that we 
present is conservative and in all probability, 
by using techniques such as electrophoresis, 
genetic differences not easily recognised in 
the phenotype will be detected, and additional 
subdivision required. An electrophoretic pro- 
gram is planned that will test the classification 
adopted here and investigate some of the 
questions raised in the discussion. 

Cladistic methods were not applied in this 
study because species discrimination de- 
pended largely on measurements, which 
would lead to difficulty in adequately defining 
character states. 

Thorpe (1976) has discussed the practical 
and theoretical problems involved with sam- 
pling and analysing the phenetic differences 
among populations. He points out that there 
are two aspects to the problem of sampling, 
obtaining enough specimens to take account 
of local variation and surveying enough local- 
ities to represent the geographical area under 
consideration. We believe that our samples 
come close to meeting these requirements, 
especially as far as the shell and opercular 
data are concerned. Certainly the amount of 



variance obtained in most characters within 
even the wider-ranging taxa is generally 
small. 

There are some inherent problems in work- 
ing with hydrobiids because their shells are 
simple, unicoloured, rather featureless and 
small. Measurements of a number of shell pa- 
rameters provide a picture of the shell that 
can be statistically analysed to detect subtle 
differences that occur between taxa. The 
opercular characters of species of Fonscoch- 
lea have proved to be useful. The number and 
relative development of the pegs on the inner 
surface of the operculum are the most useful 
opercular characters. These pegs are appar- 
ently a mechanism to increase the surface 
area for the attachment of the columellar 
muscles. The anatomical characters were 
much more difficult and time-consuming to 
study and, consequently, smaller numbers of 
individuals were examined. Important and ob- 
vious anatomical differences occur between 
the species of Trochidrobia, but within the two 
primary groups of Fonscochlea the anatomi- 
cal differences are small and show high vari- 
ance. Non-quantified characters, such as the 
pigmentation patterns on the head, were con- 
sidered when constructing our classification, 
although in some taxa head-foot pigmentation 
showed considerable intra- and inter-popula- 
tion variation. Ratios were calculated using a 
number of measurements in all three data 
sets of shell, operculum, anatomy, in an at- 
tempt to reduce size-dependent differences 
and generate shape variables. These were 
used in the initial screening of the data to as- 
sist with the delineation of taxa. 

Species are recognized in those cases in 
which, first, there were one or more morpho- 
logical differences, which we judge to be sig- 
nificant, between the individuals in one taxon 
compared with the most similar taxon, and/or 
second, the taxa, recognisable by one or 
more differences, are sympatric and conge- 
neric. Sympatric in this sense is used to in- 
clude taxa living not only within the same 
spring but in closely associated springs 
(within a few hundred metres) in the same 
spring group (i.e. parapatry). 

Subspecies have not been recognised but 
geographic forms have been identified where, 
within a taxon recognised as a species, there 
are one or more differences judged to be of 
significance between allopatric populations, 
i.e. from different spring groups. These forms 
are apparently of infraspecific status but 
whether they should be formally named must 



AUSTRALIAN SPRING HYDROBIIDS 



11 



await an analysis using biochemical methods. 
Nevertheless we have set out a formal diag- 
nosis and description of each of these forms 
so that future investigation can more readily 
focus on some of the more important geo- 
graphic differences that occur in the species 
that we recognise. In each case in which 
more than one form is recognised, form A is 
the typical form. 

Materials: Specimens were collected by 
sifting sediment with a plastic hand sieve hav- 
ing a mesh size of approximately 1 mm, and 
by washing vegetation and solid objects 
(stones, bones, wood) into a bowl. Sieve con- 
tents were tipped into a bowl and excess wa- 
ter drained out. Snails and crustaceans usu- 
ally sank to the bottom of the bowl and were 
collected in bulk. Although care was taken, 
some of the crustaceans, but very few mol- 
luscs, were lost during this process by their 
floating out with the excess water. The mate- 
rial was preserved in 5-10% formalin neutral- 
ised with excess NaHCOa, after relaxation 
with menthol crystals for 10-12 hours. 

For most springs, separate collections were 
taken at the head of the spring, at the upper 
part of the outflow, and at the middle part of 
the outflow. Collections were also often taken 
at the lower outflow and elsewhere, depend- 
ing on the type and size of spring and amount 
of time available. Separate samples were 
sometimes taken from the water edge and 
middle of the flow, otherwise the sampling 
combined these zones. 

Before sorting, samples were sieved in the 
laboratory through a 1 mm mesh to minimize 
any size bias produced by use of hand sieves 
during collecting. Samples were sorted under 
a low-power binocular microscope. If the 
sample was especially large, it was subsam- 
pled by removing all animals from a portion of 
the sample after thorough mixing, until a max- 
imum of 600 individuals of any one species 
had been counted. The specimens were 
sorted into species and the counts of number 
of individuals for each species were used to 
give approximate percentage frequencies. 
Adults and subadults only were used in the 
percentage frequency analyses as identifica- 
tion of juveniles to species was difficult and 
time-consuming. Empty shells were ignored 
in counting. The results obtained by the ana- 
lyses of qualitative samples have several lim- 
itations that are discussed below. 

Most of the material on which this report is 
based is housed in the Australian Museum 



(AMS). The holotypes, some paratypes and 
some other representative specimens are in 
the South Australian Museum, Adelaide 
(SAM). A representative collection is housed 
in the United States National Museum of Nat- 
ural History, Washington, D.C. 

Methods: Series of 20-25 adult (occasion- 
ally more) snails were randomly selected from 
given samples for morphological analyses in 
the following manner. The sample was placed 
into a Petri dish, the bottom of which was di- 
vided into a grid of 50 equal-sized and num- 
bered squares. A random number table was 
used to select grid squares. All adult snails, 
excluding highly eroded specimens, were re- 
moved from each selected square until the 
desired number of specimens was obtained. 
Shells were measured with either a Wild dis- 
secting microscope (M5 or M7) fitted with an 
ocular micrometer, or with a Houston Instru- 
ments Hipad Digitizer linked to a Morrow Mi- 
crodecision (MD2) computer. For measure- 
ments using the former method, a shell was 
first affixed to a piece of plastic clay, apex 
pointing directly upwards, so that protoconch 
diameter (PD, Fig. 4c) could be measured and 
counts made of protoconch and teleoconch 
whorls (PW, TW). The shell was then reori- 
ented to the standard position, i.e. aperture 
facing upwards (Fig. 4a) and measurements 
made of shell height (SH), shell width (SW), 
aperture height (AH), aperture width (AW), 
and length of the body whorl (BW, Fig. 4a). For 
most shells measured using this method, a 
Wild M-5 microscope was used with 1 x eye- 
pieces, and 12 X (large species) or 25 x 
(small species) magnification for all shell fea- 
tures except protoconch diameter (50 x ). The 
variance in shell measurements using the oc- 
ular micrometer, as determined by repeated 
measurements of a given feature on a single 
specimen, was approximately 0.05 mm. 

For measurements using the digitizing pad, 
shells were oriented in the positions de- 
scribed above and placed under a Wild M-5 
dissecting microscope. The shell image was 
projected onto the digitizing pad by a drawing 
apparatus attached to the microscope. Shell 
features were measured by placing the cur- 
sor, equipped with a cross-hair, over stan- 
dardized points of the shell in a predeter- 
mined sequence, with coordinate data sent to 
the computer at these points by pressing the 
cursor button, using the point, not stream, 
mode. In addition to the six meristic variables 
listed above, the width of the first half-whorl of 



12 



PONDER, HERSHLER & JENKINS 
■Ab 




FIG. 4. Shell and operculum, showing various measurements. 

A. Shell. AH, aperture height; AW, aperture width; BW, height of body whorl; SH, shell height; SW, shell 
width; WB, width of body whorl. 

B. Shell showing measurements taken for convexity calculation (see methods). 
С Protoconch. PD, protoconch diameter. 

D. Operculum, inner side. OL, opercular length. 

E. Operculum, side view. PC, length of calcareous area; PH, peg height. 

F. Palliai cavity, showing selected measurements of palliai structures. 

A, anus; CA, distance from anus to ctenidium; CG, capsule gland; CO, distance between posterior end of 
osphradium and posterior end of ctenidium; CT, ctenidium; LC, length of ctenidium; ML, maximal length of 
palliai cavity; MM, minimal length of palliai cavity; MW, width of palliai cavity; OS, osphradium; R, rectum; 
RO, renal opening. 



the body whorl (WB, Fig. 4a) and convexity of 
the penultimate whorl (CV; see below) were 
also measured using the Hipad. The Hipad 
was significantly more accurate than the 
above method, with repeated measurements 
varying by less than 0.02 mm. After a shell 
was measured it was cracked and the snail 
sexed by examination of the anterior portions 
of the genital tracts. 

After sexing, opercula were removed from 
the same groups of snails used for shell mea- 
surements. Because measurements taken of 
the opercula of species of Trochidrobia did 
not provide useful data, these have been ex- 
cluded from the analyses. The following 
methods apply to the opercula of species of 
Fonscochlea. Opercula were measured using 



a Wild M-5 dissecting microscope equipped 
with an ocular micrometer, with 10 x eye- 
pieces and 50 X magnification. Opercula 
were first fixed flat onto a piece of plastic clay 
with the side that was attached to the foot 
facing upwards. The opercular length was 
measured (OL, Fig. 4d) and the calcareous 
pegs were counted. Then the opercula were 
stood on edge, with the pegs projecting be- 
neath the operculum (Fig. 4e), enabling the 
length of the calcareous deposit (PC) and the 
height of the tallest peg (PH) to be measured. 
Specimens were dissected after their shells 
were dissolved in Bouin's solution. Dissec- 
tions were done while the animals were 
pinned out in a black wax-bottomed dish filled 
with a solution of 50-70% Bouin's solution 



AUSTRALIAN SPRING HYDROBIIDS 



13 



and water. Palliai and head structures were 
measured after the palliai roof and visceral 
coil were removed from the head/foot/neck. 
The digestive gland and gonad were then 
measured, followed by the other reproductive 
organs and stomach. All measurements were 
made, in the latter part of the study, with a 
crossed measuring reticule, divided into 200 
segments on each line, in a 25 x eyepiece 
using 31 X magnification on the Wild M-7, or 
25 X magnification on the Wild M-5. In the 
early stages of the project a single line reti- 
cule, divided into 120 segments, in a 10 x 
eyepiece, was used at 31 x magnification on 
the Wild M-7. All measurements were con- 
verted into millimeters and used for calcula- 
tion of ratios by the computer (see below). 

The mean, standard deviation and variance 
were calculated for each attribute by sex for 
each population, using the microcomputer. All 
data files generated from the microcomputer 
were reformatted into data matrices based 
upon species and attribute groups and trans- 
mitted via a modem to disk storage on a main- 
frame computer, initially the CSIRO Cyber 
computer but more recently the NSW Data 
Processing Bureau Burroughs 7700. The Sta- 
tistical Package for the Social Sciences was 
used to generate descriptive statistics (sub- 
program BREAKDOWN), test homogeneity of 
variances with both Bartlett's and Cochran's 
C-test, and perform two-tailed, single classifi- 
cation analyses of variance with the subpro- 
gram ONEWAY for each attribute. Missing 
data were ignored. In the cases in which 
groups of populations displayed significant 
heterogeneity of variance for given attributes, 
the data were transformed using either a log 
or arcsine transformation prior to analysis of 
variance. Student-Newman-Keuls test (SNK) 
and the Scheffe test were used to compare 
means using 0.05 and 0.001 probability lev- 
els. For all tests, significance was checked 
using the tables of critical values in Rohlf and 
Sokal (1969). Tests for sexual dimorphism 
were carried out using the subprogram ONE- 
WAY on selected attributes for all species 
groups at probability levels of 0.05 and 0.001 . 
Because some characters in some species 
proved to be sexually dimorphic, the male and 
female data were analysed separately. 

Multivahate analysis was undertaken using 
dischminate function analysis (MDA) (here- 
after referred to as discriminate analysis) us- 
ing the BIOSTAT package of programs (Pi- 
mentel & Smith, 1986). Because there are 
problems in using ratios in multivariate anal- 



yses (Brookstein etal., 1985) and closely cor- 
related measurements a reduced set of mea- 
surements was used in the discriminate 
analyses [Fonscochlea: shell: SH, SW, AH, 
TW; operculum (not used with F. zeidleri): OL, 
PH, PC, PN; Trochidrobia: SH, SW, AH, AW, 
BW, TW, PD]. Discriminate analyses were run 
for each species group at the population level 
with sexes separate, and populations grouped 
into species and/or geographic forms of spe- 
cies with sexes separate and sexes combined. 
Anatomical data sets were run in the same 
way with two species groups in which ana- 
tomical data were used phmanly to discrimi- 
nate some of the species and geographic 
forms {Trochidrobia spp.; female genital mea- 
surements: GO, CG, AG, ВС, WB, DB, CV, 
DV; "large aquatic" species of Fonscochlea; 
palliai measurements: LC, WC, FC, AC, HC, 
LO, WO, DO, CO, with sexes combined be- 
cause of small numbers for each station). 

Because of space constraints the univari- 
ate statistical analyses of the measurement 
data are not provided, nor are the details of 
the measurements obtained for every popu- 
lation. In the case of those data utilized in 
discriminate analysis, however, the results of 
an SNK test (P<0.05) are given for each 
character. It is hoped to utilize further the ex- 
tensive set of measurement data in conjunc- 
tion with a planned electrophoretic program. 
A summary of the measurement data is given 
in Appendix 2, Tables 18-21. 

Characters: For descriptions of the taxa 
and analyses of morphological variation, the 
characters listed below were quantified for 
samples of snails from given populations. 

The characters of the shell that were mea- 
sured (Fig. 4A-C) are: 

Maximal diameter of protoconch (PD). 

Number of protoconch whorls (PW). 

Number of teleoconch whorls (TW). 

Shell height (SH), maximal length of shell 
along shell axis. 

Shell width (SW), maximal width of shell 
perpendicular to shell axis. 

Length of body whorl (BW), length from the 
suture, at junction of penultimate and body 
whorls. 

Width of body whorl (WB), maximal diame- 
ter of first half-whorl of body whorl. 

Height of aperture (AH), maximal length 
parallel to shell axis. 

Width of aperture (AW), maximal width per- 
pendicular to shell axis. 

Convexity (CV), shortest distance from line 



14 



PONDER, HERSHLER & JENKINS 



connecting sutures at junction between pen- 
ultimate and body whorls to most abaxial 
point on whorl outline (Fig. 4B:c-d), divided by 
length of line connecting the two sutures (Fig. 
4B:a-b). 

The following ratios were generated from 
the shell measurements and used in the data 
analysis: protoconch diameter/shell height 
(PD/SH); shell width/shell height (SW/SH); 
aperture height/shell height (AH/SH); aper- 
ture height/length of body whorl (AH/BW); ap- 
erture width/width of body whorl (AW/WB); 
and an estimation of the degree to which the 
outer lip of the aperture protrudes beyond the 
outline of the junction of the penultimate and 
body whorl (WB/SW). 

The opercular characters determined were: 

Opercular length (OL), the maximal length 
of the operculum. 

Number of opercular whorls (OW); deter- 
mined for species of Trochidrobia only. 

Number of pegs (PN) (i.e. number of sep- 
arate calcareous projections); determined for 
species of Fonscochlea only, as were the fol- 
lowing opercular characters. 

Maximal height of pegs (PH), including 
thickness of operculum itself. 

Length of calcareous smear (PC), length of 
calcareous deposit associated with pegs. 

Several anatomical characters were deter- 
mined. All measurements are maximal 
widths, lengths etc. unless otherwise stated. 
Characters of the head/foot and general body 
are: 

Length of snout (LS), distance from eye to 
snout tip. 

Length of tentacles (LT), distance from eye 
to tentacle tip. 

Length of buccal mass (BM), measured af- 
ter removal from snout. 

Length of radular sac behind buccal mass 
(RS), length of portion of radular sac protrud- 
ing from posterior end of buccal mass. 

Length of digestive gland (LD), measured 
along its mid-upper surface following the coil. 

Length of gonad (LG), measured as above. 

Length of the digestive gland anterior to go- 
nad (DG). 

In the case of the palliai cavity all measure- 
ments were taken with the palliai cavity re- 
moved and flattened out (Fig. 4F). Characters 
are: 

Maximal and minimal lengths of palliai cav- 
ity (ML, MM), distance from renal opening to 
given points along edge of cavity (Fig. 4F). 

Width of palliai cavity (MW), taken as width 
of cavity approximately perpendicular to rec- 



tum (large species of Fonscochlea) (Fig. 4F) 
or as width along mantle edge (small species 
of Fonscochlea, and Trochidrobia spp.). 

Number of ctenidial filaments (FC). 

Length of ctenidium (LC), following curva- 
ture of ctenidium (Fig. 4F). 

Width of ctenidium (WC), maximal width 
along long axis of filaments. 

Gill apex (AC), width of ctenidium from left 
side to position of filament apex. 

Filament height (HC), height of a filament at 
widest part of ctenidium. 

Length and width of osphradium (LO, WO). 

Distance between posterior tip of osphra- 
dium and postehor tip of ctenidium (CO) (Fig. 
4F). 

Shortest distance between osphradium and 
edge of palliai cavity (DO). 

Distance between ctenidium and anus 
(CA), measured as shortest distance between 
anterior end of ctenidium and left side of anus 
(Fig. 4F). 

Shortest distance between anus and man- 
tle edge (MA). 

Characters of the stomach are: 

Length (SL), taken as entire length of stom- 
ach, including style sac, for Trochidrobia and 
small species of Fonscochlea, and length of 
stomach excluding style sac portion for large 
species of Fonscochlea. 

Length of style sac (SS). 

Height of anterior stomach chamber (AS). 

Height of posterior stomach chamber (PS). 

Many characters of the genital system were 
measured. 

Whereas small variations due to reproduc- 
tive state could not be assessed in this ana- 
lysis, all individuals for which genital charac- 
ters were measured appeared to be sexually 
mature. Immature or parasitized specimens 
were rejected. 

Characters of the male genitalia are: 

Length and width of prostate gland (PR, 
PW). 

Length of palliai portion of prostate gland 
(PP), that part protruding into palliai cavity. 

Length of penis (PL). 

Characters of the female genitalia are: 

Length of glandular oviduct (GO). 

Length of capsule gland (CG) and albumen 
gland (AG). 

Length of genital opening (GP). 

Length and width of bursa copulatrix (ВС, 
WB). 

Length of duct of bursa copulatrix (DB). 

Length and width of "seminal receptacle" 
(SR, WR), only for Fonscochlea. 



AUSTRALIAN SPRING HYDROBIIDS 



15 



Length of duct of "seminal receptacle" 
(DR), only for Fonscochlea. 

Length of coiled portion of oviduct (CV), 
length of coiled section posterior to "seminal 
receptacle" (Fonscochlea) or bursa copulatrix 
(Trochidrobia). 

Maximal and minimal diameters of coiled 
portion of oviduct (DV, MO). 

Length of oviduct between seminal recep- 
tacle and bursa copulatrix (BS); Fonscochlea 
only. 

Length of free portion of ventral channel 
(VC), that portion anterior to duct of bursa 
copulatrix. 

For species of Fonscochlea, the following 
groups of anatomical ratios were used: a) pal- 
liai ratios: LC/SH (SH is shell height), LO/SH, 
FC/SH, MM/SH, HC/SH, MA/SH, CA/SH, MW/ 
MM, LO/LC, HC/WC, AC/WC, WC/LC, WO/ 
LO; b) general ratios: BM/SH, BM/RS, LT/LS, 
LD/SH, LG/LD; c) stomach ratios: SS/SL (see 
comments above under SL), PS/AS; d) male 
genital ratios: PL/SH, PP/SH, PP/PR; e) fe- 
male genital ratios: AG/SH, CG/SH, CG/AG, 
BC/AG, DB/AG, SR/BC, CV/GO, VC/CV, VC/ 
AG, BS/OD (0D= CV + VC), OV/GO (0V = 
CV + VC + BS). For Trochidrobia, the palliai 
ratios, stomach and general ratios, and male 
genital ratios were precisely the same as 
those for Fonscochlea, except that shell width 
(SW), rather than shell height, was used for 
scaling. The female genital ratios generated 
for Trochidrobia were AG/SW, CG/SW, CG/ 
AG, BC/AG, DB/AG, CV/GO, VC/CV, VC/AG, 
DV/MO, DB/BC, and DV/VC. 

Anatomy 

Two species are described in detail, F. ac- 
cepta (form A), from Welcome Springs, and T 
punicea, from Blanche Cup Spring and Fin- 
niss Springs. Some supplementary informa- 
tion is given for F. zeidleri from Blanche Cup 
Spring. 

The specimens were dissected by the 
same methods used to obtain the anatomical 
measurements above). Specimens fixed in 
Bouin's solution were sectioned in paraffin at 
about 6 microns and stained with Mallory's 
Triple Stain. 

Physiology 

Matenals: The following snail species (with 
localities) were used in the experiments: Troch- 
idrobia punicea (Finniss Springs), Fonscoch- 
lea cónica (Welcome Springs), Fonscochlea 



variabilis form A (Blanche Cup, Coward 
Springs Railway Bore), Fonscochlea accepta 
form В (Finniss Springs), Fonscochlea ac- 
cepta form A (Welcome Springs), Fonscoch- 
lea aquatica form A (Blanche Cup) and of. 
form A (Kewson Hill) and Fonscochlea zeid- 
leri form A (Finniss Springs, Blanche Cup, 
Kewson Hill and Coward Springs Railway 
Bore). These species represent the majority 
of those found in the southern and middle 
groups of springs found between Marree and 
Oodnadatta. 

The springs from which the material studied 
was collected were, for logistical reasons, all 
in the southern half of the spring system be- 
tween Marree and Oodnadatta (see Appendix 
1 for detailed maps and station details). 
These were, in east-west order: 

Welcome Springs (Stn 756), a moderately 
large spring with a low mound. A small pool 
near the head is a few cm deep and there is a 
shallow (< 1cm), rather long outflow. The 
substrate is a mixture of calcareous rock, 
sand and mud. Sedges are moderately com- 
mon and filamentous algae are abundant. 

Finniss Springs (Stn 693), a small spring 
with a very low sand mound. The substrate is 
sand and mud. Sedges are common and fil- 
amentous algae are present. 

Blanche Cup Spring (Stn 739), a conical 
calcareous mound with a pool at the top (Fig. 
la). The outflow is shallow and mainly broad 
and flows over calcareous rock but the pool 
contains mainly mud. Sedges line the pool 
edges and filamentous algae are abundant in 
the pool and in the outflow. 

Coward Springs Railway Bore (Stn 743), a 
very large swamp issuing from a large pond 
with the bottom composed mainly of silt. The 
water depth is generally in excess of several 
cm where the specimens were collected, in 
the vicinity of the pond outflow. Large sedges 
and rushes line the edges of the pool and 
outflow. Filamentous algae are abundant. 
This is the only known case in which the 
mound spring snails have become estab- 
lished in a bore drain. It is also the only known 
locality at which F. zeidleri is aquatic as well 
as amphibious. Fonscochlea aquatica is not 
found here and T punicea is uncommon. 

Kewson Hill Springs (Stn 742), one of sev- 
eral small springs issuing from this hill. They 
trickle down the steep hillside in narrow out- 
flows where they form a series of small ter- 
races (Ponder, 1986), each containing water 
a few mm deep. There is no vegetation apart 
from some filamentous algae. 



16 



PONDER, HERSHLER & JENKINS 



Methods: All experiments were conducted 
in a makeshift laboratory set up in a large tent 
(5 X 4 m) in the field between August 27 and 
September 9, 1983. Snails from given popu- 
lations were collected and then held in water 
in aerated plastic containers (16 x 16 cm) for 
one to three days before being used in the 
experiments. When possible, water from the 
spring from which a given sample of animals 
was collected was used for holding both the 
animals and for the experiments (Blanche 
Cup, Welcome Spring, Coward Springs Rail- 
way Bore). In instances in which a large water 
sample could not be obtained owing to shal- 
low water and/or low discharge, water from a 
nearby spring or bore was used. In the case 
of Finniss Springs, the water was taken from 
a bore about 7 km southwest of Hermit Hill 
and the water used for the experiments with 
F. aquatica from Kewson Hill was taken from 
the Blanche Cup Spring. Full analyses of 
the water from these localities is given in Kin- 
hill-Stearns (1984). A running record of the 
laboratory environment (air temperature, hu- 
midity) was kept. To avoid introducing age- 
related differences, only adult snails, i.e. 
those possessing a complete and thickened 
peristome, were used for the experiments. 

A major problem encountered in physiolog- 
ical experiments involving shelled gastropods 
is determining when individuals are dead. Re- 
traction of the snail into its shell usually oc- 
curs before death in response to unaccept- 
able conditions. For most of the experiments 
the activity of the snails was used as an indi- 
cator of their tolerance to the conditions being 
presented. Given the time constraints inher- 
ent in the project, the customary replicates of 
each experiment could not be done. We pre- 
ferred to use the available time to run each 
experiment for all of the taxa. The detailed 
methods of each type of experiment are given 
below. 

In the desiccation experiments animals 
from given populations were placed in a se- 
ries of 9-cm Petri dishes. Ten specimens 
were placed in each dish. The dishes were of 
three types: those lined with dry filter paper 
and without a lid (hereafter referred to as dry); 
those lined with moist filter paper and with a 
lid (moist); and those half-filled with water and 
with a lid (wet). The moist and wet tests 
served as controls. A total of 21 dishes, seven 
sets of each of the three types, was set up for 
each population tested. A separate set of 
dishes was checked after periods of one, two, 
four, six, 12, 24, and 48 hours from the be- 



ginning of the experiment. As the moistened 
dishes tended to dry out, despite having lids, 
they were frequently examined and re-moist- 
ened whenever necessary. To check for sur- 
vival of snails in a set of dishes, the dishes 
were first flooded, if dry or moist, with water. 
The number of animals in each dish that were 
active 10 minutes after flooding was noted. A 
similar check for active animals was made 
one hour after flooding. Animals inactive after 
one hour were considered dead. Death was 
confirmed for the snails by tests carried out in 
some of the early runs: shells were gently 
crushed to expose the animal, placed under a 
dissecting microscope, and the mantle was 
not seen to retract when prodded. 

In the salinity experiments table salt was 
added to the appropriate spring water to ob- 
tain solutions of six, nine, 12, and 24 %o. The 
salinities of these solutions were tested using 
an optical refractometer. Each of these solu- 
tions, as well as a normal sample of the spring 
water, for which a zero salinity reading was 
obtained using the refractometer, serving as a 
control was added to a glass jar of about 380 
cc brimfull capacity, which was then capped 
with a plastic lid to exclude air from the jar as 
much as possible. Ten specimens were 
placed into each of these five jars. After inter- 
vals of one, two, three, six, 12, and 24 hours, 
each of the jars was examined, but not 
opened, and the number of active or clinging 
snails counted. Mortality was not tested. The 
salinities for each of the water sources used, 
calculated from the conductivities given by 
Kinhill-Stearns (1 984), are shown in Table 1 2. 

In the experiments with deoxygenated wa- 
ter, water from the appropriate spring was 
boiled for two to three minutes in a glass bea- 
ker and then poured very gently, to prevent 
reoxygenation, into each of five 25 cc test 
tubes. Rubber stoppers were then gently in- 
serted into each of the tubes. The tubes were 
cooled and then 20 snails were placed into 
each of them, as well as into a sixth tube con- 
taining well-oxygenated spring water as a con- 
trol. The tubes were then again firmly stop- 
pered, with an effort made to exclude air 
bubbles. After intervals of one, two, four, six, 
and 20 hours, a tube with deoxygenated water 
was checked in the following manner. First the 
number of active specimens in the tube was 
counted. Then the specimens from the tube 
were placed into a dish with oxygenated water. 
The number of active specimens in the dish 
was counted after periods of ten minutes and 
one hour. Specimens inactive after one hour 



AUSTRALIAN SPRING HYDROBIIDS 



17 



were considered dead. At the end of each of 
the five time periods, the control tube was ex- 
annined as well, but not opened, and the num- 
ber of active individuals in the tube counted. 

The purpose of the temperature experiment 
was to determine activity of animals at various 
temperatures. Twenty specimens were 
placed into each of two 275 cc jars, half-filled 
with water. One jar was slowly heated by 
placing it into a steam-heated, water-filled 
dish. The jar was periodically removed from 
the water bath, the temperature of the water 
in the jar noted, and the number of active in- 
dividuals in the jar counted when the desired 
temperatures were reached. The process 
was continued until such a temperature was 
reached at which all specimens became inac- 
tive. A similar method was used to determine 
tolerance to low temperatures: the second jar 
was placed into a small freezer and periodi- 
cally removed to check the temperature and 
count the active animals. Again, the experi- 
ment was terminated when all specimens be- 
came inactive. The jars were not aerated dur- 
ing the experiments. Mortality was not tested 
and no attempt to achieve acclimation was 
made. 

In determinations of submergence toler- 
ance a 380 cc jar was filled to the brim with 
water and 20 snails were added. The jar was 
then capped with a lid that had a small hole in 
it so that an aerator tube could pass through it 
into the jar. An aerator stone was attached to 
the end of the tube. At intervals of one, two, 
four, 15, 24, 48, and 72 hours, the jar was 
examined and the number of active snails 
counted. In experiments of submergence/ 
non-submergence preference a plastic plate 
was used (diameter of 220 mm), with a flat 
circular bottom (diameter of 150 mm), steeply- 
sloping sides (approximately 60° width of 
13 mm), and a slightly-sloping rim (approxi- 
mately 1 0° width of 22 mm). The dish was filled 
with water to the lower edge of the rim. Fifty 
snails were placed in the dish and left for three 
hours. At the end of this time period the num- 
bers of specimens found on the bottom of the 
dish, on the steep slope and on the broad rim 
(out of the water) were counted. 

In determinations of response to light a 200 
X 200 X 15 mm clear perspex box, with 
tightly-fitting lid, was constructed for use in 
this experiment. Three lines were drawn 
across the width of the box in order to divide 
the box lengthwise into four equal zones. One 
hundred snails were placed in the box to- 
gether with water. The water level in the box 



was then topped off and the lid placed on top, 
with a smear of petroleum jelly added to tfie 
sides to provide a seal. Care was taken to 
exclude any air bubbles from the box. Half of 
the box, containing two entire zones, was 
covered with a dark plastic sheet and then an 
Olympus dissecting microscope lamp was 
placed 2 cm above the mid-line at the uncov- 
ered end of the box. The lamp was oriented 
so that its beam was perpendicular to the 
plane of the box. The lamp was then turned 
on, to level 6 on the transformer, and the en- 
tire apparatus, box and lamp, was covered 
with a black plastic sheet to exclude other 
light. After one hour both the dark sheet and 
the sheet covering one half of the box were 
removed, and the numbers of animals in each 
of the four zones were quickly counted. The 
numbers of snails found in the light and light- 
middle zones were combined, as were those 
found in the dark and dark-middle zones, in 
order to obtain sufficiently high frequencies 
for the statistical analysis of these results. For 
most of the populations tested, two separate 
runs were done. The box was thoroughly 
washed and all grease removed between 
runs of this experiment. 

To test for differences in results between 
runs, populations or species, the following 
statistical tests were used (following Siegel, 
1956): Fisher's Exact Test, when the experi- 
ments involved fewer than 20 animals or 
when expected frequencies in cells were 
fewer than five; and The Chi-Square Test of 
Independence, with continuity correction, 
when the experiments involved 20 or more 
animals with expected frequencies in the cells 
exceeding five. Null hypotheses were re- 
jected when the significance level was less 
than or equal to 0.05. 



RESULTS 



Taxonomy 



The hydrobiids occurring in the Lake Eyre 
Supergroup are formally described in this 
section. Two new genera, Fonscochlea with 
six species and Trochidrobia with four spe- 
cies, are erected, with a new subgenus, Wolf- 
gangia, of Fonscochlea, containing one spe- 
cies. Geographic forms are recognised in four 
of the species of Fonscochlea, these being 
formally described but not named. 

A summary of measurement details is 
given in Appendix 2, Tables 18-21. 



18 



PONDER, HERSHLER & JENKINS 



TABLE 2. Tests for sexual dimorphism in shell 
height (SH) and shell width (SW). The asterisk 
indicates a significant difference, at the level 
indicated, between males and females for all 
pooled measurements for the taxon. 







SH 


SW 


Species 


.05 


.001 


.05 .001 


F. accepta form A 


. 


. 


. 


F. accepta form В 


* 


* 


* 


F. accepta form С 








F. acuática form А 




* 


* 


F. aquatica form В 




* 




F. variabilis form A 




* 


. 


F. variabilis form В 




* 


* 


F. variabilis form С 




. 


* 


F. billakallna 






* 


F. cónica 


* 


* 


* 


F. zeidleri form А 






* 


F. zeidleri form В 


* 


* 




Г. punicea 


* 


* 


* 


T. smithi 








T. minuta 






* 


T. inflata 









Type species: Fonscochlea accepta n.sp. 

Distribution: Artesian springs between Маг- 
гее and Oodnadatta, northern South Austra- 
lia. 

Diagnosis: Shells (Figs. 5-7, 14, 19, 22, 
23, 25) of known species snnall to large for 
family (1.3 mm long), non-umbilicate, ovate- 
conic to ovate, smooth or with weak axial 
rugae formed from enlarged growthlines. Pro- 
toconch (Fig. 9) of about one and one-half 
whorls, minutely pitted, the pits sometimes ar- 
ranged into spiral rows (subgenus Wolfgan- 
gia). Aperture rather large relative to shell 
length (AH/SH >0.4), oval, thickened when 
mature, without external varix; outer lip 
slightly prosocline to slightly opisthocline. Peri- 
ostracum thin, sometimes developing weak 
ridges that coincide with the growthlines and, 
sometimes, spiral scratches. 

Operculum (Fig. 8) corneous, oval, flat, of 
few whorls, nucleus eccentric, inner surface 
with small calcareous smear and/or calcare- 
ous pegs. 





FIG. 5. Shells of Fonscochlea accepta. 

a. Fonscochlea accepta form A, holotype. Welcome Springs (003). 

b. Fonscochlea accepta form B. Old Finniss Springs (694) (SAM, D. 17918). 
с Fonscochlea accepta form С Emerald Springs (703) (SAM, D. 17919). 



Those species shown to be sexually dimor- 
phic in size (at P<0.01) are listed in Table 2. 
Because most of the species showed evi- 
dence of dimorphism the morphometric data 
for each sex were treated separately. Some 
additional data are provided below. 

Family Hydrobiidae 

GENUS FONSCOCHLEA n. gen. 

Derivation: Fons (Latin), a spring; cochlea 
(Latin), a snail (fem.). 



Radula (Fig. 10) with rectangular central 
teeth, cusp formula ^'^^„V|^p^ - lateral teeth 
2-4 + 1 +2-4. Inner marginal teeth with 8-1 5 
cusps, outer marginal teeth with 17-25 
cusps. 

Head-foot (Figs. 1 1 , 24a-g,i) typical of fam- 
ily. Cephalic tentacles slightly tapering to par- 
allel-sided; weakly and inconspicuously cili- 
ated on ventral surfaces. Snout well 
developed, slightly shorter to slightly longer 
than tentacles. Pigmentation heavy to light. 



AUSTRALIAN SPRING HYDROBIIDS 



19 




FIG. 6. Shells of species of Fonscochlea. a-d,i. Fonscochlea accepta form B. a. Finniss Swamp West 
(690)(AMS, С 152978). b. Sulphuric Springs (735) (AMS, C.I 52979). с Hermit Hill Springs (711) (AMS, 
С 152980). d. Old Woman Spring (733) (AMS, C.I 52981). i. Old Finniss Springs (710) (AMS, С 152982). 
e-h. Fonscochlea zeidleri iorm A. e. Elizabeth Springs (024) (AMS, C.I 52975). f-h. Blanche Cup Spring 
(008) (AMS, C.I 52977). 



pigment granules black and white. No acces- 
sory tentacles. 

Palliai cavity (Fig. 4F) with well-developed 
ctenidium, osphradium oval, about three to 
four times as long as broad; its posterior ex- 
tremity situated near posterior end of ctenid- 



ium. Ctenidium about 3-4.5 times length of 
osphradium. 

Alimentary canal typical of family. Stomach 
(Figs. 43a, 44b, 45) with anterior and poste- 
rior chambers, single digestive gland opening 
and no caecal appendage. 



20 



PONDER, HERSHLER & JENKINS 




FIG. 7. Shells of species of Fonscochlea. 

a. Fonscochlea ze/d/er/ form A, Strangways Springs (030) (AMS, C.1 52992). 

b. Fonscochlea zeidleri form B, Big Cadnaowie Spring (661) (AMS, C.I 52993). 

с Fonscochlea aquatica cf. form A, very squat variety, Kewson Hill Springs (742) (AMS, C.1 52994). 

d. Fonscochlea billakalina, paratype, Old Billa Kalina Spring (026) (AMS, C.1 52995). 

e. Fonscochlea variabilis form B, The Fountain Spring (032) (AMS, C.1 52996). 

f. Fonscochlea aquatica form B, Freeling Springs (665) (AMS, C.1 52997). 

g. Fonscochlea accepta form A, Welcome Springs (003) (AMS, C.1 52998). 

h. Fonscochlea accepta form B. Old Finniss Springs (694B) (AMS, C.1 52999). 
i. Fonscochlea accepta form C, Emerald Springs (703) (AMS, C.1 53000). 
Scale: 0.5mm. 



AUSTRALIAN SPRING HYDROBIIDS 



21 




FIG. 8. Opercula of species of Fonscochlea. 

a. Fonscochlea zeidleri form B, Big Cadnaowie Spring (661 ). 

b. Fonscochlea zeidleri ^orm A, Coward Springs Railway Bore (018). 
с Fonscochlea aquatica cf. form A, Kewson Hill Springs (742). 

d. Fonscochlea billakalina. Old Billa Kalina Spring (026). 

e. Fonscochlea variabilis form B, The Fountain Spring (032), 

f. Fonscochlea aquatica form B, Freeling Springs (665). 

g. Fonscochlea accepta form B, Old Finniss Springs (694B). 
h,i. Fonscochlea accepta form A, Welcome Springs (003). 
Scale: 0.1mm. 



22 



PONDER, HERSHLER & JENKINS 




FIG. 9. Protoconchs of species of Fonscochlea. 

a. Fonscochlea accepts form A, Welcome Springs (003). 

b. Fonscochlea accepta form С Emerald Springs (703). 
c-d. Fonscochlea zeidlen form A, Strangways Springs (030). 

e. Fonscochlea aquatica form A, Outside Springs (039). 

f. Fonscochlea cónica. Welcome Springs (003). 
Scale: d = 0.01mm; all others = 0.1mm. 



AUSTRALIAN SPRING HYDROBIIDS 



23 




FIG. 10. Radulae of Fonscochlea. 

a. Fonscochlea ze/d/er/ form B, Big Cadnaowie Spring (661). 

b. Fonscochlea zeidleri form A, Coward Springs Railway Bore (018). 
с Fonscochlea accepta form B, Old Finniss Springs (694B). 

d. Fonscochlea accepta form C, Emerald Springs (703). 

e. Fonscochlea variabilis form B, The Fountain Spring (032). 

f. Fonscochlea aquatica form B, Freeling Springs (665). 
Scale: 0.01mm. 



Female reproductive system (Figs. 12, 27, 
47) with two sperm sacs, i.e. anterior bursa 
copulatrix and posterior "seminal receptacle", 
and coiled oviduct lying on inner (left) side of 



albumen gland, sperm sacs and major ovi- 
duct folds being opposite posterior part of 
gland or partly extending behind it. Coiled ovi- 
duct an unpigmented, coiled or undulating 



24 



PONDER, HERSHLER & JENKINS 




FIG. 1 1 . Dorsal views of heads of large species of Fonscochlea; all from living material. 

a. Fonscochlea zeidleri form A, Kewson Hill Springs. 

b. Fonscochlea zeidleri form A, Welcome Springs. 

с Fonscochlea aquatica form A, Blanche Cup Spring. 

d. Fonscochlea accepta form A, Welcome Springs. 

e. Fonscochlea aquatica cf. form A, Kewson Hill Springs. 

f. Fonscochlea accepta form B, Old Finniss Springs. 
Scale: 0.25mm. 



AUSTRALIAN SPRING HYDROBIIDS 



25 




FIG. 12. Female genitalia of species of Fonscochlea. 

a. Fonscochlea zeidleri form B, Big Cadnaowie Spring. 

b. Fonscochlea zeidleri form A, Old Finniss Spring. 

с Fonscochlea aquatica form A, Blanche Cup Spring. 

d. Fonscochlea accepta form A, Welcome Springs. 

e. Fonscochlea accepta form C, Emerald Springs. 

f. Fonscochlea accepta form B, Old Finniss Springs. 

g,h. Fonscochlea accepta form A, Davenport Springs; detail of sperm sacs and tfieir ducts shown in h. 
ag, albumen gland; be, bursa copulatrix; eg, capsule gland; cv, coiled oviduct; go, oviduct opening; mcp, 
posterior limit of palliai cavity; sr, seminal receptacle; vc, ventral channel; vcp, posterior extension of ventral 
channel. 
Scale: 0.25mm. 



muscular tube extending from immediately 
behind posterior palliai wall, where its initial 
section forms U-shaped, glandular loop, to 



loop posteriorly around sperm sacs at, 
or just behind, albumen gland. Gonopericar- 
dial duct represented by tissue strands only. 



26 



PONDER, HERSHLER & JENKINS 



Oviduct between sperm sacs very short to 
moderately long, forming U-shaped loop. 
Anterior to bursal duct, which opens to ovid- 
uct opposite posterior part of albumen gland, 
muscular oviduct either runs straight to ven- 
tral channel or thrown into loop. Bursa 
copulatrix and "seminal receptacle" approxi- 
mately equal in size and with ducts markedly 
shorter than length of sacs. Both sperm sacs 
similar histologically and rather thick-walled. 
Capsule gland approximately equal in size to 
albumen gland or slightly smaller or larger. 
Ventral channel well defined, with conspicu- 
ous ciliated lateral fold. Genital opening sub- 
terminal. 

Male reproductive system with vas deferens 
complexly coiled beneath anterior part of tes- 
tis. Palliai and visceral vas deferens enter and 
leave prostate gland in middle section. Pros- 
tate gland extends into palliai wall, as slight 
bulge in some species to about half its length 
in others. Palliai vas deferens narrow, tubular, 
and lying beneath epithelium of right side of 
palliai floor, undulating as it passes across 
neck and enters base of penis. Penis (Fig. 46) 
with swollen, unpigmented base bearing 
prominent concentric creases; distal two thirds 
smooth and tapering to point, often pigmented 
and muscular. Penial duct similar to palliai vas 
deferens, i.e. very narrow, ciliated and with 
only very thin muscle layer; straight in distal 
part of penis, undulating in proximal part. Pe- 
nial pore simple. 

Egg capsules hemispherical, attached to 
substrate. 

Nervous system (Fig. 43b) with typical hy- 
drobiid pattern: cerebral ganglia separated by 
short commissure, left pleural ganglion at- 
tached to suboesophogeal ganglion and right 
pleural ganglion separated from supra- 
oesophageal ganglion by long connective. 

See anatomical section below for further 
details of anatomy. 

Remarks: The distinctive features of this 
genus include the equal-sized sperm sacs, 
the short ducts connecting these sacs to the 
oviduct and the position at which they enter 
the oviduct. In most hydrobiids the bursal duct 
opens to the oviduct opposite the anterior end 
of the albumen gland, not the posterior end as 
in Fonscochlea. The pegged operculum, and 
the shell of some of the smaller species, re- 
semble states seen in the Australian species 
of Hemistomia sensu lato (Ponder, 1982). 
This genus, and the related genus Tatea T. 
Woods, 1 879, can be distinguished from Fons- 
cochlea in having a more "typical" hydrobiid 



reproductive system (Ponder, 1982). In these 
genera the seminal receptacle is thin-walled 
and much smaller than the bursa copulatrix, 
and the bursal duct opens to the oviduct in the 
region near the anterior end of the albumen 
gland. In most other respects these three 
genera are similar. 

Subgenus Fonscochlea s.s. 

Diagnosis: Shell (Figs. 5, 6a-d, i, 7c-i, 14b, 
d, 19, 22, 23, 25) thin to moderately thick, 
aperture with thin to slightly thickened peris- 
tome. Protoconch microsculpture (Fig. 
9a,b,e,f) of irregular, shallow pits. 

Operculum (Fig. 8c-i) with prominent pegs, 
weak pegs or pegs absent. 

Radula (Fig. lOc-f) as for genus. (Table 3) 

Head-foot (Figs. Ilc-f, 20a-g, i) with ce- 
phalic tentacles slightly longer than snout. 

Female genital system (Figs. 12c-h, 27) as 
for genus except that the oviduct between the 
ventral channel and the bursal duct is always 
bent or folded and the sperm sacs lie behind 
(to the right of) the coiled oviduct and their 
ducts emerge from their dorsal sides. 

Male system as for genus. 

Remarks: The typical subgenus includes 
five of the six known taxa of Fonscochlea. It 
encompasses two radiations, one of small 
species and the other of large species, all of 
which are aquatic. 

Group 1 : the large aquatic species. 

Fonscochlea accepta n.sp. 

Derivation: accepta (Latin), welcome, a ref- 
erence to the type locality. 

Diagnosis: Shell about 2.4 to 3.8 mm long, 
with about 2.5-3.6 convex (convexity ratio 
0.08-0.25) teleoconch whorls. Aperture with 
thin peristome, outer lip slightly prosocline. In- 
ner lip narrow, loosely attached to parietal 
wall. Operculum with strong pegs. 

Shell (Figs. 5, 6a-d,i, 7g-i; 9a,b), see diag- 
nosis. Colour dark brown. 

Operculum (Fig. 8g,i) with several, usually 
3-4, strong pegs. 

Radula (Fig. 10c,d) as for genus (see Table 
3 for details). 

Head-foot (Fig. 11d,f), see under descrip- 
tions of the forms of this species below. 

Anatomy typical of subgenus. Described in 
more detail in the anatomical section below. 

The typical form of this species is described 



AUSTRALIAN SPRING HYDROBIIDS 



27 



TABLE 3. Cusp counts from radular teeth of species of Fonscochlea and Trochidrobia. Missing counts 
from the outer marginal teeth are the result of not being able to make accurate counts from the available 
preparations. 





Central tooth 
No. of No. of 


Lateral tooth 


Inner 
marginal tooth 


Outer 




No. of ■ No. of 


marginal tooth 




lateral 


basal 


inner outer 


No. of 


No. of 


Species 


cusps 


cusps 


cusps cusps 


cusps 


cusps 


F. accepta form A 


3-4 


1 


3-4 3-4 


9-10 


24-25 


F. accepta form В 


3 


1-2 


2-3 3-4 


9-12 


— 


F. accepta form С 


4 


1-2 


3 3-4 


10-13 





F. acuática form А 


3-4 


1 


2-3 2-4 


7-10 


— 


F. aquatica form В 


2-3 


1 


3 3 


8-9 


21-25 


F. variabilis form A 


4-6 


1-2 


2-3 2-4 


12-15 





F. variabilis form В 


3-4 


1-2 


2-3 2-3 


9-12 





F. variabilis form С 


2-4 


1-2 


2 2-3 


9-11 





F. billakalina 


3-4 


1-2 


2-3 2-4 


10-12 





F. cónica 


4-6 


1-2 


3 3-4 


14-18 


— 


F. zeidleri form А 


2-3 


2 


2-3 3 


9-13 


17-21 


F. zeidleri form В 


2-3 


2 


2-3 3 


9-10 


20-21 


Г. punicea 


5-8 


1-2 


3-6 4-6 


24-31 


— 


T. smith! 


6-7 


1 


4-5 5-6 


23-25 


— 


T. minuta 


4-7 


1-2 


4-6 6-7 


22-24 


— 


T. inflata 


6-8 


1 


5-6 5-7 


18-23 


— 



below as "form A" where a holotype is des- 
ignated for the species. 

Localities: Southern Springs: Welcome, 
Davenport, Hermit Hill and Emerald Springs 
(Fig. 13). 

Remarks: Three geographically separated 
forms are recognised. Discriminate analysis 
did not convincingly separate two of these us- 
ing shell and opercular characters but reason- 
able discrimination was achieved using palliai 
data. The forms are primarily distinguished by 
differences in their ctenidia and unquantified 
differences, including tentacle shape and pig- 
mentation and habitat preference. 

This species has a range of about 80 km 
with the typical form occupying about a 25 km 
range, separated from the Hermit Hill popula- 
tions (form B) by about 12 km and those in 
turn separated from Emerald Spring, the lo- 
cality of the third form, by about 40 km. 

This species is the "large aquatic" species 
of the Southern Springs. It is generally abun- 
dant in the pool at the head of the springs and 
in their outflows. It can sometimes be seen 
clustering on the sides of the outflows but it is 
not amphibious and, if emergent, is covered 
by a film of water. 

Fonscochlea accepta form A. 

(Figs. 5a, 7g, shell; 9a, protoconch; 8h,i, 
operculum ; 1 1 d , head-foot; 43a, 44b, stomach ; 



43b, nervous system; 46a, penis; 12d,g,h, fe- 
male genitalia. 

Diagnosis: Tends to have longer and more 
numerous ctenidial filaments (Table 18B) 
than F. accepta form В and shorter filaments 
than F. accepta form С Radular sac longer, 
and ratio of buccal mass to radular sac (BM/ 
RS) smaller, than in both other forms. Also 
differs from F. accepta form В in pigmentation 
and morphology of cephalic tentacles. 

Shell (Figs. 5a, 7g; 9b, protoconch) as for 
species, but not so broad relative to length as 
F. accepta form С See Table 18A for mea- 
surement data. 

Operculum (Fig. Bh) as for species. See Ta- 
ble 18A for measurement data. 

Radula as for species. See Table 3 for data. 

Head-foot (Fig. lid) black on sides of foot 
and on neck and snout. Tentacles parallel- 
sided or taper slightly distally and lightly to 
darkly pigmented, except for pale median 
stripe most obvious in individuals with darker 
tentacles. An indistinct red-brown patch on 
outer dorsal side of tentacles just in front of 
eyes present and few dense white pigment 
cells lie above eyes. 

Anatomy (Figs. 12d,g,h, female genitalia; 
43a, 44b, stomach; 43b, nervous system; 
46a, penis) as for species. See Tables 18B-E 
for measurement data. 

Type material: holotype (Fig. 5a) (SAM, 
D.I 791 7, stn 003); and paratypes (003, AMS, 



28 



PONDER, HERSHLER & JENKINS 




AUSTRALIAN SPRING HYDROBIIDS 



29 



C.1 52848, many, C.1 52998, 1, figured; 756A, 
AMS, C.1 52849, many; 756B, AMS, 
С 152850, many; 756C, AMS, C.1 52851, 
many). 

Dimensions of holotype: length 3.26 mm, 
width 1.83 mm, length of aperture 1.43 mm. 

Localities: Welcome Springs (002, 003, 
754A-D, 755A-D, 756A-C); Davenport 
Springs (004, 005, 752A,C, 753A,B (Fig. 13). 

Remarks: The populations at Welcome and 
Davenport Springs do not seem to show any 
significant differences in any of the non-gen- 
ital characters measured but there are some 
differences in measurements in the female 
genitalia. In particular BS/OD, CV/GO and OV/ 
GO are significantly different. It is possible, on 
more detailed analysis, that these popula- 
tions, which are more than 20 km apart, will 
be shown to be separable. 

Fonscochlea accepta form B. 

Figs. 5b, 6a-d,i, 7h, shell; 11f, head-foot; 12f, 
female genitalia; 8g, operculum; 10c, radula 

Diagnosis: Ctenidial filaments fewer and 
shorter than in other two forms, and ctenidium 
tends to be shorter, although these differ- 
ences not consistently significantly different 
for all populations. Radular sac shorter, and 
ratio of buccal mass to radular sac (BM/RS) 
larger, than in both other forms of F. accepta. 
Cephalic tentacles with reduced or absent 
median stripe and not tapered. 

Shell (Figs. 5b, 6a-d,i, 7h) generally similar 
to form A but some individuals approach F. 
accepta form С in shape. See Table 1 8A for 
measurement data. 

Operculum (Fig. 8g) as for species. See Ta- 
ble 18A for measurement data. 

Radula (Fig. 10c) as for species. See Table 
3 for data. 

Head-foot (Fig. 1 1f) similar to that of F. ac- 
cepta form A but median stripe on tentacles 
reduced or absent and tentacles usually 
slightly swollen distally, or if not, parallel- 
sided (i.e. not tapered). 

Anatomy (Fig. 12f) as for species. See Ta- 
bles 18B-E for measurement data. 

Voucher material: primary voucher speci- 
men (Fig. 5b) (SAM, D.I 791 8, stn 694B); ad- 
ditional material from same station (694B, 
AMS, С 152852, many, С 152999, 1, figured; 
693A, AMS, C.1 52853, 36; 693B, AMS, 
C.152854, 50; 693C, AMS, C.152855, 10; 
694A, AMS, C.1 52856, 10; 694C, AMS, 
C.152857, 16). 

Dimensions of primary voucher specimen: 



length 3.17 mm, width 1 .86 mm, length of ap- 
erture 1.38 mm. 

Localities: Hermit Hill Complex: Hermit Hill 
Springs (711A-D, 712); Old Finniss Springs 
(693A-C, 694A-C, 710); Old Woman Springs 
(733A-E); Finniss Swamp West (690A-C, 
691 A-D, 730); Dead Boy Spring (689); Sul- 
phuric Springs (735); Bopeechee Springs 
(692A,B). Shells, possibly referable to this 
form, are known from Priscilla (686) and Ven- 
able (687) Springs (Fig. 13). 

Remarks: This form is distinguished from F 
accepta form A in ctenidal characters, a 
shorter radular sac, and tentacle shape. The 
smaller gill seen in F accepta form В might 
have evolved in response to the generally 
small springs found in the Hermit Hill area. 
This form also differs behaviourly from form A, 
preferring the shallow water in the outflows to 
the deeper water in pools, whereas F accepta 
form A is found in pools in large numbers. 

Using discriminate analysis on a subset of 
shell measurements and opercular measure- 
ments, populations of this form did not sepa- 
rate well from F accepta form A, although 
partial separation is achieved (Figs. 15, 16; 
Table 4). Palliai measurements, however, 
produced a clear separation from form A and 
the next form (Figs. 17, 18; Table 4). 

Fonscochlea accepta form С 

(Figs. 5c, 7i, shell; 9b, protoconch; lOd, rad- 
ula; Fig. 12e, female genitalia) 

Diagnosis: Shell with relatively shorter spire 
than many other populations, but this not con- 
sistent. Gill filaments longer, typically twice as 
long, and more numerous than those of F 
accepta form В. Similar, but less pronounced, 
differences between this form and F accepta 
form A, with ratios of ctenidial length/shell 
length (LC/SH) and length of ctenidial fila- 
ments to shell length (HC/SH) larger than in 
both other forms. Distance between anus and 
ctenidium (CA) and ratio of this distance over 
shell length (CA/SH) larger than in other two 
forms. Radular sac intermediate in length be- 
tween other two forms. Head-foot (not ob- 
served in living material) similar to F accepta 
form A in having well-developed, unpig- 
mented dorsal stripe on tentacles. 

Shell (Figs. 5c, 7i; 9b, protoconch) as for 
species except for a relatively larger aperture 
(mean of AH 1 .52, males; 1 .51 , females; com- 
pared with 1.31-1.46 mm for the other two 
forms). AH/BW is larger in most individuals 
than in the other two forms (mean 0.62, com- 



30 



PONDER, HERSHLER & JENKINS 



TABLE 4. Summary of results of discriminate analysis of the forms of the large aquatic species of 
Fonscochlea. The numbers are the Euclidean (taxonomic) distances between the groups. 





F.ac.A 


F.ac.B 


F.ac.C 


F.aq.A 


Faq.A(r) 


F.aq.ci.A 


F.aqf.B 




















Right side: 


F. accepta form A 


X 


0.460 


0.598 


1.611 


1.477 


2.519 


1.010 


Female, shell 
& operculum 






0.470 


0.131 


1.472 


1.274 


2.693 


1.042 


Male, shell & 
operculum 



F accepta form В 


0.459 
0.198 


X 


0.503 
0.442 


1.762 
1.570 


1.742 
1.521 


2.418 
2.517 


1.302 
1.229 


F. accepta form С 


0.375 
2.722 


0.370 
2.889 


X 


1.286 
1.484 


1.328 
1.298 


1.964 
2.685 


0.950 
1.063 


F. aquatica form А 
(combined) 


1.550 


1.667 


1.326 


X 


— 


— 


0.771 
0.521 


F. aquatica form A 
(restricted) 


1.384 
9.365 


1.630 
9.533 


1.272 
6.756 





X 


1.842 
2.119 


0.507 
0.372 


F. aquatica cf. form A 


2.606 
0.396 


2.463 
0.539 


2.261 
2.402 


— 


1.972 


X 


2.029 
2.020 


F. aquatica form В 


1.025 
3.630 


1.253 
3.797 


0.901 
1.169 


0.637 


0.420 
5.737 


2.004 
3.271 


X 



Left top — shell + operculum combined sexes 
Left bottom — palliai combined sexes 

pared with 0.57-0.58). See Table 18A for 
measurement data. 

Operculum as for species. See Table ISA 
for measurement data. 

Radula (Fig. 1 0d) as for species. See Table 
3 for data. 

Head-foot similar to that of F. accepta form 
A as far as can be judged from preserved 
material. 

Anatomy (Fig. 12e, female genitalia) as for 
species. See Tables 16B-E for measurement 
data. 

Voucher material: primary voucher speci- 
men (Fig. 5c) (SAfVI, D.I 791 9, stn 703A); ad- 
ditional material from same station (703A, 
AMS, C.I 52858, many, C.I 53000, 1, figured; 
703B, AMS, C.I 52859, 60). 

Dimensions of primary voucher specimen: 
length 3.10 mm, width 1 .90 mm, length of ap- 
erture 1.40 mm. 

Locality: Emerald Springs (703A,B). 

Remarks: This population is recognised as 
a separate form because it differs from the 
other two forms, particularly F. accepta form 
B, in gill characters, as described above. It 
appears to have head-foot characters similar 
to those of F. accepta form A, but differs from 
F. accepta form В in this respect, and also 
differs in the distance of the anus from the 
mantle edge from both of the other forms. Dis- 
criminate analysis on palliai measurement 



data readily separates this form (Figs. 17, 18; 
Table 4). 

This form lives in the upper outflow of a 
large, isolated spring in swiftly flowing water 
that reaches a depth of as much as several 
centimeters. It is common in the roots of dense 
vegetation around the fenced spring head at 
the uppermost part of the outflow but relatively 
rare on the downstream side of the fence 
where it appears to require shelter beneath 
debris such as wood. This suggests that, un- 
like the other two forms, which are commonly 
seen in the open, this form is strongly pho- 
tonegative. 

Emerald Springs is unusual in containing 
only one species of hydrobiid. This locality is 
widely separated, by about 40 km, from other 
populations of F. accepta, the nearest being 
those in the vicinity of Hermit Hill (F. accepta 
form B). 

Fonscochlea aquatica n.sp. 

Derivation: a reference to the aquatic habit of 
this species, in contrast to F. zeidleri. 

Diagnosis: Shell large for genus (2.6 to 
4.8 mm long), with 2.1-3.7 teleoconch 
whorls. Aperture with thin peristome and or- 
thocline to opisthocline outer lip. Inner lip 
broad and firmly attached to parietal wall. 
Operculum with weak or absent pegs. 



AUSTRALIAN SPRING HYDROBIIDS 



31 



Shell (Figs. 7c,f; 14b,d; 53c,e; 9e, proto- 
conch) as for diagnosis. Colour yellowish- 
brown to chocolate or reddish-brown. 

Operculum (Fig. 8c,f) with pegs weak to 
moderately strong, or absent altogether. 

Radula (Fig. 1 0f) as for genus. See Table 3 
for details. 

Head-foot (Figs. 11c,e) with pale, tapering 
cephalic tentacles and the darkly-pigmented 
head and snout. 

Anatomy (Fig. 12c, female genitalia) typical 
of subgenus. Similar to F. accepta, differ- 
ences being mainly size-related. 

The typical form of this species is described 
below as "form A" where a holotype is des- 
ignated for the species. 

Localities: Middle, South Western, North- 
ern and Freeling Springs (Fig. 13). 

Remarks: This species can be divided into 
two geographic forms, possibly subspecies, 
which are separated on shell and opercular 
characters. It differs from F. accepta in its 
larger size (SH) and most other shell mea- 
surements are significantly different in nearly 
all populations and, consequently, many 
other size-related characters. They also differ 
in apertural details and in the relatively 
weaker to absent pegs on the operculum; PH/ 
OL, PC/OL and PN/OL are all significantly dif- 
ferent in most populations. The ratio AH/BW 
(aperture height/body whorl) is significantly 
larger in F. aquatica than in F accepta in 
nearly all populations. This species separated 
well from F accepta in discriminate analysis 
using shell and opercular measurements 
(Figs. 15, 16; Table 4). 

Fonscochlea aquatica form A. 

(Figs. 7c, 14d, 53c, e, shell; 9e, protoconch; 
8c, operculum; 11c,e, head-foot; 12c, female 
genitalia) 

Diagnosis: Shell with 2.10-3.63 (mean 
3.24, males; 3.26, females) weakly to moder- 
ately convex teleoconch whorls (convexity ra- 
tio 0.16-0.24; mean 0.17, males; 0.18, 
females). Aperture oval with inner lip attached 
to parietal wall over most of length. Colour 
yellowish to chocolate brown. Operculum with 
calcareous smear 0-0.4 mm long (mean 
0.22 mm, males; 0.21, females). 

Shell (Figs. 7c, 14d, 53c,e; 9e, protoconch) 
as for diagnosis. See Table 18A for measure- 
ment data. 

Operculum (Fig. 8c) with 1-4 (mean 2.80, 
males; 2.57, females) pegs, 0.02-0.29 mm 



(mean 0.10 mm, males; 0.11 mm, females) 
high. See Table 18A for measurements. 

Radula as for species. See Table 3 for data. 

Head-foot (Fig. 11c,e) as for species; dor- 
sal cephalic tentacles uniformly lightly to 
darkly pigmented, sometimes with narrow, 
short unpigmented stripe bordered with dark 
lines. 

Anatomy (Fig. 12c, female genitalia) as for 
species. See Tables 18B-E for dimensions. 

Type material: holotype (Fig. 14d) (SAM, 
D.I 7920, 009); and paratypes (008, AMS, 
C.I 52860, 2; 685, AMS, C.I 52861, many; 
739, AMS, C.1 52862, many). 

Dimensions of holotype: length 4.27 mm, 
width 2.45 mm, length of aperture 1 .86 mm. 

Localities: Middle Springs: Horse Springs 
East (747A,B, 748A-C), Horse Springs West 
(746A,B), Mt. Hamilton Homestead (006), 
Strangways Spring (745A), Blanche Cup 
Spring (008, 685,739), Little Bubbler Spring 
(744A-C), Bubbler Spring (013), unnamed 
springs, Blanche Cup Group (786, 787), Cow- 
ard Springs (019, 764A-C), Kewson Hill 
Springs (740, 741, 742A,B, 765), Elizabeth 
Springs (766A-F, 767A,B, 771 A-C), Julie 
Springs (772A-D, 773A,B), Jersey Springs 
(683A,B, 769A,B, 770A), Warburton Spring 
(681 A-C, 682), Beresford Spring (028). 

South Western Springs: Billa Kalina 
Springs (026, 723A-D, 759A, 761 A-C, 
762A,B, 763A,B), Francis Swamp (717B,C, 
720A,B, 721 A-C), Strangways Springs (007, 
029-030, 678A,B, 679A-C). Shells only from 
Margaret Spring (722). 

Northern Springs: Brinkley Springs (677), 
Hawker Springs (670B,C, 671 , 672A-D, 673), 
Fountain Spring (031-033), Twelve Mile 
Spring (036,037), Big Perry Spring (034), 
Outside Springs (038-040, 041) (Fig. 13). 

Remarks: This form is the large aquatic 
species living in the Middle, South Western 
and Northern Springs, replacing F accepta, 
which occurs in the Southern Springs. 

Specimens from the Kewson Hill Springs 
and, to a lesser extent Elizabeth, Jersey and 
Julie Springs, tend to have stunted shells 
(Figs. 7c, 53c) and smaller gills with fewer 
filaments than have other populations of this 
form. The only important characters consis- 
tently separating these populations are peg 
height (PH) and the length of the calcareous 
smear (PC) and these, together with the val- 
ues of PH/OL and PC/OL, are significantly dif- 
ferent from those of all other populations of F 
aquatica. Peg number also tends to be less, 
but not consistently so. The non-opercular dif- 



32 



PONDER, HERSHLER & JENKINS 






FIG. 14. Shells of species of Fonscochlea 

a. Fonscochlea zeidleri form A, holotype. Coward Springs (764) 

b. Fonscochlea aquatica forrn B. Freeling Springs (665) (SAM D 17921) 

c. FonscocWeaze/d/er/ form B. Big Cadnaowie Spring 661) (SAM D 17916) 

d. Fonscochlea aquatica form A. holotype. Blanche Cup Sp ing (009) 



AUSTRALIAN SPRING HYDROBIIDS 



33 




FIG. 15. Plot of group centroids, using the first two canonical axes, obtained from discriminate analysis of 
populations of large aquatic species and forms of Fonscochlea using shell and opercular measurements. 
Males and females of each population are, for the purposes of this analysis, treated as distinct populations. 
The axes contain the following percentages of the variance of the variables used: first (horizontal) axis: SH, 
50.15%; SW, 41.40%; AH, 74.33%; TW, 53.49%; OL, 91.57%; PH, 78.06%; PC, 35.01%; PN, 38.94%. 
Second (vertical) axis: SH, 0.18%; SW, 19.15%; AH, 6.39%; TW, 2.14%; OL, 4.03%; PH, 13.72%; PC, 
47.09%; PN, 0.06%. a, F. accepta form A; с, F. aquatics form В; f, F. accepta form В; к, F. acuática cf. form 
A; q, F. aquatica form A, typical; t, F. accepta form С 



34 



PONDER, HERSHLER & JENKINS 




FIG. 16. Plot of group centroids, using first and third canonical axes, obtained fronn discriminate analysis of 
populations of large aquatic species and forms of Fonscochlea using shell and opercular measurements. 
Males and females of each population are, for the purpose of this analysis, treated as distinct populations. 
The axes contain the following percentages of the variance of the variables used; first (horizontal) axis: SH, 
50.15%; SW, 41 .40%; AH, 74.33%; TW, 53.49%; OL, 91 .57%; PH, 78.06%; PC, 35.01%; PN, 38.94%. Third 
(vertical) axis: SH, 0.42%; SW, 0.04%; AH, 0.32%; TW, 0.13%; OL, 0.52%; PH, 5.62%; PC, 7.36%; PN, 
10.35%. a, F. accepta form A; с, F. aquatica form В; f, F. accepta form В; к, F. aquatica cf. form A; q, F. 
aquatics form A, typical; t, F. accepta form С. 



ferences are not consistent within the geo- 
graphic area in which the form occurs. The 
opercular and palliai differences are important 



but we do not judge them to be of sufficient 
magnitude to regard this form as a species, 
given the degree of overlap with typical F. 



( 



AUSTRALIAN SPRING HYDROBIIDS 



35 



aquatica form A. These differences are as 
great as or greater than those between some 
groups of populations recognised here as dis- 
tinct geographic forms but because these 
populations do not occupy a geographic area 
clearly separate from that of F. aquatica form 
A, it is not formally differentiated. These pop- 
ulations are recognised in the discussion be- 
low as F. aquatica cf. form A but are included 
in the diagnosis of form A above. They form a 
separate group when opercular and shell data 
are lumped together using discriminate ana- 
lysis (Figs. 15,16). The Kewson Hill popula- 
tion (stn 741) in particular, has most shell 
measurements significantly different from all 
other populations of this species (including 
683 and 767) and also differs from all popu- 
lations (except stn 679) in the ratio BW/WH, 
but not in other shell ratios. Discriminate ana- 
lysis using palliai measurements also sepa- 
rates the Jersey-Elizabeth-Kewson Hill popu- 
lations from typical F. aquatica form A (Figs. 
17, 18). 

Somewhat surprisingly, there do not ap- 
pear to be any consistent differences be- 
tween the populations in the Northern and 
Blanche Cup Springs; despite their consider- 
able separation, these group very closely in 
all the analyses. It is suggested below that the 
presence of F. aquatica form A in springs of 
the Middle Springs might be due to a rela- 
tively recent introduction to some of those 
springs, but that the form in the springs be- 
tween Jersey Springs and Kewson Hill might 
be an earlier stock that differentiated at an 
infraspecific level. Biochemical evidence is 
required to determine the status of these pop- 
ulations. 

Fonscochlea aquatica form B. 

(Figs. 7f, 14b, shell; 8f, operculum; 10f, rad- 
ula) 

Diagnosis: Shell with 3.0 to 3.7 (mean 3.30, 
males; 3.33, females) teleoconch whorls, with 
more convex (convexity ratio 0.18-0.25; 
mean 0.21, males; 0.23, females) teleoconch 
whorls than is usual in typical form. Aperture 
more nearly circular than in typical form, with 
inner lip attached to parietal wall over shorter 
distance. Colour reddish to orange-brown. 
Operculum with calcareous smear (0.26- 
0.60 mm; mean 0.39 mm, males; 0.37 fe- 
males) longer than in typical form. 

Shell (Figs. 7f , 1 4b), see diagnosis. See Ta- 
ble 18A for measurements. 

Operculum (Fig. 8f) as for species. Calcar- 



eous smear longer than in typical form. See 
Table 18A for measurements. 

Radula (Fig. lOf) as for species. See Table 
3 for data. 

Head-foot as for species (preserved mate- 
rial only examined) except for distinct, dark, 
black to dark grey, dorsal stripe on tentacles 
of most individuals; rarely with short white 
stripe. 

Anatomy as for species. See Tables 1 8B-E 
for measurement data. 

Voucher material: primary voucher speci- 
men (Fig. 14b) (SAM, D.I 7921, 665A); addi- 
tional material from this station (665A, AMS, 
С 152863, many, С 152997,1, figured; 665B, 
AMS, C.I 52864, many; 665C, AMS, 
C.I 52865, many); 664A, AMS, С 152866, 
many; 664B, AMS, C.I 52867, many; 664C, 
AMS, C.I 52868, 32; 045, AMS, C.I 52869, 
25; 046, AMS, C.I 52870, many. 

Dimensions of primary voucher specimen: 
length 4.59 mm, width 2.47 mm, length of ap- 
erture 1.98 mm. 

Localities: Freeling Springs (042-044, 
045-046, 663, 664B,C, 665A-C). 

Remarks: The Freeling Springs form of F 
aquatica is consistently and readily distin- 
guishable at sight from specimens in the 
springs farther southeast, the more convex 
teleoconch whorls and reddish colour in par- 
ticular, being distinctive features. The circular 
aperture is probably correlated with the more 
convex whorls and the shorter area of attach- 
ment of the inner lip of the aperture to the 
parietal wall. 

This form separates well from related taxa 
by discriminate analysis using shell and oper- 
cular measurements (Figs. 15, 16) and is also 
separated from F aquatica form A using pal- 
liai measurements (Figs. 17, 18). 

Discrimination of the large aquatic taxa of 
Fonscochlea and their forms was tested us- 
ing discriminate analysis on shell and opercu- 
lar measurements and palliai measurements. 
The results showed that all groups could be 
discriminated using these data, with 85% of 
all measured individuals (n = 625) being clas- 
sified correctly with the shell + opercular 
data and 78% of the specimens (n = 1 03) us- 
ing the palliai measurements. The Euclidian 
(taxonomic) distances between the groups 
are given in Table 4. With shell and opercular 
data the greatest distance score when sexes 
were treated as separate populations was 
2.69 between F aquatica cf. form A and F 
accepta form A. All of the painwise compari- 
sons between F aquatica and F accepta 



36 



PONDER, HERSHLER & JENKINS 




FIG. 17. Plot of group centroide, using first two canonical axes, obtained from discriminate analysis of 
populations, sexes combined, of large aquatic species and forms of Fonscochlea using pallia! measure- 
ments. The axes contain the following percentages of the variance of the variables used: first (horizontal) 
axis: LC, 16.32%; WC, 77.51%; FC, 52.54%; AC, 78.58%; HC, 59.24%; LO, 46.76%; WO, 29.30%; DO, 
1.12%, CO, 37.69%. Second (vertical) axis: LC, 19.95%; WC, 5.70%; FC, 30.44%; AC, 1 .57%; HC, 33.13%; 
LO, 0.04%; WO, 6.27%; DO, 47.12%, CO, 8.90%. a, F. accepta form A; c, F. aquatica form B; f, F accepta 
form В; к, F. aquatica cf. form A; q, F. aquatica form A, typical; t, F. accepta form С 

\ 



AUSTRALIAN SPRING HYDROBIIDS 



37 




FIG. 18. Plot of group centroids, using first and third canonical axes, obtained from discriminate analysis of 
populations, sexes combined, of large aquatic species and forms of Fonscochlea using pallia! measure- 
ments. Thie axes contain the following percentages of the variance of the variables used: first (horizontal) 
axis: LC, 16.32%; WC, 77.51%; PC, 52.54%; AC, 78.58%; HC, 59.24%; LO, 46.76%; WO, 29.30%; DO, 
1.12%, CO, 37.69%. Third (vertical) axis: LC, 6.38%; WC, 0.62%; FC, 0.01%; AC, 12.12%; HC, 2.07%; LO, 
4.27%; WO, 36. 1 3%; DO, 1 0.62%, CO, 0.30%. a, F. accepta form A; c, F. aquatica form B; f, F. accepta form 
B; k, F aquatica cf. form A; q, F aquatica form A, typical; t, F accepta form С. 



scored >0.95 (all but one >1.0, the lowest 
distance score between females of F. aquat- 
ica form В and F. accepta form C). All of the 
pairwise comparisons between the groups 
within A. aquatica scored >0.37 (all but one 
>0.5, the lowest distance score between 
males of F. aquatica forms A and B). Within 
F. accepta all groups scored >0.13 (all but 
one >0.44, the lowest distance score be- 
tween males of F. accepta forms A and C). 
Using palliai data the distance scores be- 
tween F. aquatica and F. accepta were 
>0.39 (all but two >1.0, the lowest scores 



between F. aquatica cf. form A and F. ac- 
cepta forms A and B, reflecting the reduced 
gill in this form of F. aquatica). The greatest 
scores (>9.3) were between F. aquatica form 
A and F. accepta forms A and B. Within F. 
accepta the forms had distance scores 
>0.19, this score being between forms A and 
B, form С having a score of >2.7 when con- 
trasted with the other two forms. The groups 
within F. aquatica separated with scores 
>3.2, that between form A and cf. A being 
5.7. 
SNK tests (5% level) using pooled data, 



38 



PONDER, HERSHLER & JENKINS 



combined and separate sexes, for each vari- 
able used in the discriminate analyses gave 
these results. 

Shell and opercular characters: 

SH — Combined sexes: significantly differ- 
ent for both species and all forms except F. 
accepta form С and F. accepta form A. Sep- 
arate sexes: the same result except for F. 
aquatica form A, F. aquatica cf. form A and F. 
aquatica form В overlapping. The means for 
this character were not significantly different 
between males and females except for the 
two forms of F. aquatica (females larger). 

SW — Combined sexes: means significantly 
different for F accepta form В and F. accepta 
form A + F. accepta form С. Separate sexes: 
F. accepta form В, F. accepta form A + F. 
accepta form С + F. aquatica cf. form A and 
F. aquatica form A + F. aquatica form В are 
significantly different subsets. Only F aquat- 
ica form A shows significant sexual dimor- 
phism for this character. 

AH — Combined sexes: significantly differ- 
ent for all forms of both species. Separate 
sexes: five subsets are discriminated; F ac- 
cepta form В, F. accepta form A + F accepta 
form С, F. aquatica cf. form A + F. aquatica 
form В, F. aquatica form А male and female. 
Sexual dimorphism is apparent in only F. 
aquatica form A. 

TW — Combined sexes: significantly differ- 
ent for the two forms of F aquatica, the forms 
of F accepta overlapping but, together, being 
significantly different from F aquatica. Sepa- 
rate sexes: two groups of overlapping subsets 
are discriminated; one with F accepta (all 
forms) + F aquatica cf. form A, the other with 
F aquatica form A + F aquatica form B. This 
character does not significantly differ between 
males and females. 

OL — Sexes combined: same result as for 
SH. Separate sexes: two groups of overlap- 
ping subsets are discriminated that corre- 
spond to the same groups as for the last vari- 
able (TW). There was no significantly different 
sexual dimorphism. 

PH — Combined sexes: significantly differ- 
ent for F. aquatica form A, F. aquatica form В 
+ F. accepta form В and F accepta form A + 
F. accepta form С. Separate sexes: all form 
overlapping subsets except F. aquatica cf. 
form A. None show significant differences be- 
tween sexes in this character. 

PC — Combined sexes: means significantly 
different for the two forms of F aquatica and 
these both separate from F accepta, the 
forms of that species not being dischminaJed. 




FIG. 19. Shells of species of Fonscoclilea. 

a. Fonscochlea variabilis form A, holotype. Blanche 
Cup Spring (009). 

b. Fonscochlea variabilis form A, Bubbler Spring 
(013) (AMS, C.I 53001). 

с Fonscochlea variabilis form C, Freeling Springs 
(045) (AMS, C.I 52882). 

d. Fonscochlea billakalina, holotype. Old Billa Kal- 
ina Spring (027). 



Separate sexes: three groups are discrimi- 
nated, F aquatica cf. form A, F aquatica form 
В and, the third (intermediate) group with the 
rest. There is no sexual dimorphism in this 
character. 

PN — Combined sexes: all overlap except F. 
aquatica form A. Separate sexes: all overlap 
except F aquatica cf. form A. There is no 
significant sexual dimorphism in this charac- 
ter. 

It is clear from these results that the Jersey 
Springs-Kewson Hill form of F. aquatica is 
very distinct, as is also demonstrated with the 
palliai characters below. 

Palliai characters (combined sexes only 
given here): LC — F accepta form В + F. 
aquatica cf. form A + F accepta form A are 



AUSTRALIAN SPRING HYDROBIIDS 



39 




FIG. 20. Blanche Cup pool and upper outflow (Stn. 
739), showing location of the 1 1 sampling sites for 
study of size-variation in Fonscochlea variabilis. 



not separated but F. acuática form В is sig- 
nificantly different from that subset and from a 
subset formed by F. accepta form С and F. 
aquatica form A. 

WC — F aquatica cf. form A is significantly 
different from all other forms, which form 
overlapping subsets. 

FC— F aquatica form A is significantly dif- 
ferent from all others, which form overlapping 
subsets. 

AC — There are no significant differences 
between any two forms. 

HC — Three subsets are separated, one 
with F aquatica cf. form A + F accepta form 
В, another with F accepta form С and the 
third (intermediate in size) with the three re- 
maining forms. 

LO — There are no significant differences 
between any two of the forms. 

WO — All forms contained in overlapping 
subsets. 

DO — Three different subsets are discrimi- 
nated, one with the forms of F accepta, the 
intermediate one with F aquatica cf. form A 
+ F aquatica form В and the third with F 
aquatica form A. 

Group 2: the small aquatic species. 

Fonscochlea variabilis n.sp. 

Derivation: a reference to the variable shell 
of this species. 
Diagnosis: Shell small (up to 3.5 mm long), 



conical, with 2-3.4 weakly to moderately con- 
vex (convexity ratio 0.05-0.30) teleoconch 
whorls. Aperture expanded in some popula- 
tions, not in others. Inner lip narrow and 
loosely attached to parietal wall or separated 
from it. Colour pale to dark brown. Operculum 
with 1-7 strong pegs, peg height 0.06- 
0.2 mm. 

Shell (Figs. 7e, 19a-c, 22a,c, 23d-f,h,i, 
25b) see diagnosis. 

Operculum (Fig. 8e) with strong pegs. 

Radula (Fig. 10e) as for genus. Inner mar- 
ginal with 9-15 cusps. See Table 3 for other 
details. 

Head-foot (Fig. 24a,b,d,e) variably pig- 
mented; cephalic tentacles with unpigmented 
narrow, dorsal stripe margined with pale grey 
to black lines. Cephalic tentacles and snout 
very pale grey to black, black around eyes or 
just behind eyes. 

Anatomy (Fig. 27a,d, female genitalia) sim- 
ilar to other species in subgenus. No consis- 
tent significant anatomical differences be- 
tween this species and F cónica noted, 
although data limited. 

The typical form of this species is described 
below as "form A" where a holotype is des- 
ignated for the species. 

Localities: Middle, Northern and Freeling 
Springs. 

Remarks: This species and two others, F 
cónica and F. billakalina, comprise the small 
aquatic group. They tend to prefer the upper 
outflow and spring head (Fig. 54) and to at- 
tach themselves to the undersides of hard ob- 
jects (stones, wood, bones, etc.). 

Some populations of this species show 
considerable variation, sometimes a dimor- 
phism, in size that does not seem to be sex- 
ually based. See the remarks on form A of this 
taxon for a detailed analysis and discussion of 
one of these populations. 

Apart from size-related differences, the 
three "small aquatic" species differ from F 
aquatica and F accepta in having the seminal 
receptacle displaced more posteriorly relative 
to the coiled oviduct (compare Figs. 12, 27). 

Fonscochlea variabilis form A. 

(Figs. 19a,b, 23d-f, shell; 24a,b,d,e, head- 
foot; 27a,d, female genitalia) 

Shell 1 .8-2.8 mm (mean 2.28, males; 2.42, 
females) in length, width/length ratio 0.58- 
0.65, with 2.00-3.38 moderately convex te- 
leoconch whorls (convexity ratio 0.05-0.30, 



40 



PONDER, HERSHLER & JENKINS 



TABLE 5. Descriptions of 1 1 stations in the Blanche Cup pool and upper outflow (Stn. 739) sampled for 
the study of shell variation in Fonscochlea variabilis. 



Station 



Distance from 
edge of pool 



Water 
depth 



Comments 



1 


2-5 cm 


1-2 cm 


2 


20 cm 


3 cm 


3 


1.8 m 


— 


4 


2.8 m 


15 cm 


5 


2.8 m 


15 cm 


6 


5 m 


30 cm 


7 


2m 


10 cm 


8 


2m 


10 cm 


9 


4.5 m 


1 m 


10 


— 


<2 cm 


11 


— 


<2 cm 



30-60% covered by short sedge, sandy bottom. 

mat of dead sedge on its side, mud bottom. 

mat of filamentous algae lying between sedges. 

bottom sample, 5% algal cover, some dead sedge. 

sample from sedges (20-30% cover). 

beyond edge of dense sedge mats, bottom consisting of 

dead, algal-covered sedge and water weed, 
middle of sedge zone, sparse (30%) cover, 
as in (7), but in densely covered (60%) area, 
fine sand bottom, 
outflow, under stones, 
outflow, filamentous algae. 



глеап 0.18) and aperture not markedly ex- 
panded. Operculum with strong pegs. 

Shell (Figs. 19a,b, 23d-f), see diagnosis. 
See Table 19A for measurements. 

Operculum with 1-5 (mean 2.96, males; 
3.14, females) strong pegs 0.09-0.2 mm 
(mean 0.14 mm, males; 0.15 mm, females) in 
height, calcareous area 0.16-0.34 mm 
(mean 0.24 mm) long. See Table 19A for 
measurements. 

Radula as for species. See Table 3 for de- 
tails. 

Head-foot (Fig. 24a,b,d,e) typically darkly 
pigmented with distinctive, triangular patch of 
black pigment behind eyes and patch of 
dense white granules anterior to, and on inner 
side of eyes. Small form of F. variabilis occur- 
ring at Blanche Cup Spring (see below) paler 
than large form (compare Fig. 24a,d), with 
pale grey snout and unpigmented tentacles. 

Anatomy (Fig. 27a,d, female genitalia) as 
for species. See Tables 19B-C for measure- 
ments. 

Type material: holotype (Fig. 19a) (SAM, 
D.I 6275, stn 009); and paratypes (008, SAM, 
D.3208, 74, AMS, С 152873, 1; 009, AMS, 
0.152871, many; 010, AMS, 0.1 52874, 50; 
Oil, AMS, 0.152875, 30; 739, AMS, 
0.152931, 5). 

Dimensions of holotype: length 2.45 mm, 
width 1.47 mm, length of aperture 1.07 mm. 

Localities: Middle Springs: Blanche Oup 
Spring (008-012, 685, 739), Little Bubbler 
Spring (744A-0), Bubbler Spring (013-017), 
unnamed spring in Blanche Cup Group (786), 
Coward Springs Railway Bore (018, 684, 743) 
(Fig. 26). 

Remarks: This form of F. variabilis anä F. 
cónica are found in the Blanche Cup Group 



although not in the same springs. Fonscoch- 
lea variabilis is found in the larger springs, 
whereas F. cónica is restricted to the small 
springs. This is the only detected example of 
parapatry of any taxa in the two species 
groups of Fonscochlea. 

Collections from Blanche Cup (Stn 739) 
contained not only typical Fonscochlea vari- 
abilis form A (SH, 2.0-2.7 mm), but also a 
smaller, adult (SH, <1.8 mm) "form," with a 
complete and thickened aperture. Possible 
explanations for the presence of these two 
phenotypes include sexual dimorphism, sym- 
patry of congeners (the second species being 
Fonscochlea cónica or another unnamed 
species), seasonal classes of F variabilis 
form A that attained different sizes at maturity, 
and distinct ecomorphs of F v.variabilis. In an 
effort to determine the significance and nature 
of this apparent size bimodality, the following 
data were gathered and analyzed. 

Samples were taken from 1 1 stations in the 
pool and upper outflow of Blanche Cup (Fig. 
20, Table 5), encompassing a range of micro- 
habitats and including samples along a 
transect from the edge to the center of the 
pool. Stations 1-9 were sampled on 31/8/83 
while Stations 10 and 11 were sampled on 
27/1 1/83. A fine sieve having a mesh size of 
1 mm was used to sample soft sediment and 
aquatic vegetation. At Station 10 snails were 
collected by washing them from the under- 
sides of stones into a container. A maximum 
of five minutes of sampling was done at each 
station and the snails were preserved in for- 
malin for later study. No snails were found at 
Stations 3, 6, and 9. 

From each sample, 50 mature small 
aquatic Fonscochlea having a "mature" aper- 



AUSTRALIAN SPRING HYDROBIIDS 



41 



TABLE 6. Shell height statistics for Fonscochlea variabilis from 8 stations at Blanche Cup (Stn. 739). 









Shell 


Height (mm) 










Males 






Females 




Station 


X 


SD 


N 


X 


SD 


N 


1 


2.13 


0.217 


23 


2.26 


0.272 


27 


2 


2.19 


0.208 


32 


2.20 


0.235 


18 


4 


2.26 


0.205 


29 


2.38 


0.142 


21 


5 


2.25 


0.15 


32 


2.36 


0.199 


18 


7 


2.26 


0.257 


24 


2.31 


0.222 


26 


8 


2.19 


0.259 


23 


2.24 


0.193 


27 


10 


1.78 


0.33 


25 


1.87 


0.444 


25 


11 


1.77 


0.315 


24 


2.19 


0.333 


26 



ture were selected at randonn and their shell 
heights were measured with the digitizing 
pad, for size-frequency analysis. The shells 
were then cracked and the snails sexed. 

The small aquatic snails from a large sam- 
ple obtained by general collecting at Blanche 
Cup on 29/8/83 were roughly sorted into typ- 
ical F. variabilis form A and the small "form." 
Fifty-seven of the former and 55 of the latter 
were selected at random and all shell param- 
eters were measured with the digitizing pad. 
The shells were then cracked, the snails 
sexed and the opercular data were obtained. 

Size-frequency histograms, sexes sepa- 
rate, for the shells measured from the various 
stations are given in Fig. 21 and appropriate 
statistics are shown in Table 6. The small 
"form" was almost totally absent from the 
pool samples. Noteworthy is the lack of bimo- 
dality and paucity of snails of SH less than 
1.87 mm in these samples (Fig. 21). The two 
samples from the outflow (10, 11) had large 
numbers of the small "form" (SH <1.69 mm) 
as well as typical F. variabilis form A. The 
results of a pairwise comparison, sexes sep- 
arate, of shell height among all stations (SNK 
Test, null hypotheses of equality of shell 
height rejected at P < 0.01) are given in 
Table 7. There is little difference in shell 
height among the 6 stations in the pool, with 
only 4 of 30 possible comparisons having a 
significant difference. However, the two out- 
flow samples (Stations 10, 11) do differ sig- 
nificantly in shell height for most painwise 
comparisons with the pool samples: for Sta- 
tion 10, all possible comparisons (12 of 12) 
are significantly different; for Station 11, 
seven of 12 comparisons are significantly dif- 
ferent. Note that shell height for females from 
Station 11 generally does not differ signifi- 
cantly from that of the pool samples. 

While the histograms for the samples from 



the outflow suggest bimodality in size within 
sexes, the sample sizes are too small to pro- 
vide statistically significant evidence of such 
bimodality. It is evident from the histograms 
that the apparent size bimodality is not due 
simply to sexual dimorphism: while females 
are generally larger than males, the outflow 
samples include both male and female snails 
assignable to the small "form", as well as nor- 
mal-sized males and females. 

Typical individuals of both sexes of F. vari- 
abilis form A and the small "form" were found 
to differ significantly (LSD Test, null hypothe- 
ses rejected at P < 0.01) in all shell and oper- 
cular parameters, excluding convexity, as well 
as the following ratios: PD/SH, SW/SH, AH/ 
SH, and PC/OL. While these data suggest that 
two distinct phenotypes are indeed present in 
Blanche Cup, we do not have sufficient evi- 
dence at this point to separate them as spe- 
cies, or to determine whether they represent 
ecomorphs, seasonal classes, or different 
species. At this point, we consider them, ten- 
tatively, as forms of F variabilis form A. The 
measurement data for the small form are not 
included in the summary of measurement data 
of F variabilis, but are shown as separate data 
in Table 19. The small form is also treated 
individually in the discriminate analysis and it 
groups separately from typical F variabilis 
form A and F cónica (Figs. 28-30; Table 8). 

Using discriminate analysis on shell and 
opercular measurements this form, excluding 
the small Blanche Cup form, separated rather 
well from the other small aquatic taxa of Fons- 
cochlea (Figs. 28-30; Table 8), although with 
a small amount of overlap with F cónica. 

Fonscochlea variabilis form B. 

(Figs. 7e, 22c, 23h,i, 25b, shell; 8e, opercu- 
lum; lOe, radula) 
Diagnosis; Shell 2.09-3.48 mm (mean 



42 



PONDER, HERSHLER & JENKINS 



Д 



Ú 



I 




115-132 133-150 161 U 



169186 187-204 205-222 2 23-240 241-258 259-276 277-294 

SHELL LENGTH (mm) 



FIG. 21 . Size-frequency histograms for Fonscochlea variabilis frorn eight stations at Blanche Cup (Stn. 739). 
Darkened columns, males; white columns, females. 

A. Pool stations. A, stn 2; B, stn 1 ; C, stn 8; D, stn 5. 

B. Pool and outflow stations. A, stn 1 1 ; B, stn 10; C, stn 4; D, stn 7. 



AUSTRALIAN SPRING HYDROBIIDS 



43 




mJLI 




Д 



Q 

> 10 

Q 

Z 



i 



t 



I 



1 




115-132 1.33-150 151-168 169-186 187-204 205-222 223-240 241-2.58 2.59-2.76 

В SHELL LENGTH (mm) 



44 



PONDER, HERSHLER & JENKINS 



TABLE 7. Significant differences (SNK Test, P < 0.01 ) in shell height of Fonscochlea variabilis annong 
stations at Blanche Cup (Stn. 739). Empty boxes indicate shell height for males or females does not 
differ significantly between that pair of stations. 



Station 



Station 



10 



11 



1 


— 












2 




— 










4 


F 


F 


— 








5 


M 






— 






7 


M 








— 




8 












— 


10 


M, F 


f^,F 


M, F 


M,F 


M, F 


M,F 


11 


M 


M 


M, F 


M 


M 


M 




FIG. 22. Shells of species of Fonscochlea. 

a. Fonscochlea variabilis form С Freeling Springs (664) (SAM, D.I 7913). 

b. Fonscochlea cónica, holotype. Welcome Springs (003). 

с Fonscochlea variabilis iorm B. Twelve Mile Spring (036) (SAM, D.I 791 2). 
Scale: 1mm. Scale A: a,c,; scale B: b. 



2.58 mm, males; 2.79 mm, females) in length 
with width/length ratio of 0.57-0.62, thus gen- 
erally narrower than form A, but not consis- 
tently so. Teleoconch whorls 2.38-3.5 (mean 
2.89, males; 2.98, females), convex (con- 
vexity ratio 0.10-0.25; mean 0.16, males; 
0.19, females) and aperture noticeably ex- 
panded. Operculum with well-developed 
pegs. 

Shell (Figs. 7e, 22c, 23h,i, 25b), see diag- 
nosis. See Table 19A for measurements. 

Operculum (Fig. 8e) with 2-7 (mean 3.88, 
males; 4.85, females) well-developed pegs 
0.06-0.16 mm (mean 0.10 mm, males; 
0.11mm, females) long, calcareous area 
0.16-0.47 mm (mean 0.28 mm, males; 
0.31 mm, females) long. Calcareous area 
longer and PH/OL smaller than in most spec- 



imens of F. variabilis form A. See Table 19A 
for measurement details. 

Radula (Fig. lOe) as for species. See Table 
3 for data. 

Head-foot not observed in living material 
but generally similar to form A except median 
dorsal unpigmented band on tentacles usu- 
ally very narrow or absent but black lines usu- 
ally present. Background pigmentation dark 
grey to black. 

Anatomy as for species. See Tables 1 9B-C 
for measurements. 

Voucher material: primary voucher speci- 
men (Fig. 22c) (SAM, D.17912, stn 036); ad- 
ditional material from same station (SAM, 
D.2031, 9; 037, AMS, C.1 52876, many; 036, 
AMS, С 152877, many; 1003A, AMS, 



AUSTRALIAN SPRING HYDROBIIDS 



45 



TABLE 8. Summary of results of discriminate analysis of sfiell + opercular (right side) and palliai 
characters (left side) of small aquatic species of Fonscochlea. The numbers are the Euclidean (taxonomic) 
distances between the groups. 



Right side: 





F.va.A 


f.va{sma\\) 


F. l'a. В 


F.va.C 


F. cónica 


F bill. 


F. variabilis form A 


X 


2.268 
1.792 


1.780 
1.003 


1.444 
1.478 


0.724 
0.697 


1.584 
1.613 


F. variabilis (small form) 


3.421 


X 


3.953 
2.727 


3.662 
3.139 


1.577 
1.127 


1.684 
1.714 


F variabilis form В 


1.480 


3.382 


X 


0.446 
0.502 


2.440 
1.661 


3.283 
2.492 


F. variabilis form С 


1.506 


3.421 


0.162 


X 


2.139 
2.978 


2.899 
2.978 


F. cónica 


0.660 


1.376 


2.073 


2.115 


X 


1.283 
1.395 


F. billakalina 


1.524 


1.683 


2.906 


2.908 


1.311 


X 



Left side: Combined sexes. 



C.1 52878, many; 1003B, AMS, C.I 52879, 
many; 1003C, AMS, C.I 52880, many; 
1003D, AMS, C.I 52881, 20; 037, AMS, 
C.1 52929, 3). 

Dimensions of primary voucher specimen: 
length 2.95 mm, width 1 .58 mm, length of ap- 
erture 1.21 mm. 

Localities: Northern Springs: Hawker 
Springs (670A-C, 672A,B,D, 673), Fountain 
Spring (031-032), Twelve Mile Spring (035- 
037, 1003A-D), Big Perry Springs (034), Out- 
side Springs (038, 040) (Fig. 26). 

Remarks: This form is not readily separable 
from F. variabilis form A quantitatively on any 
single character. Shells are generally separa- 
ble on the characters given in the diagnosis, 
although there is considerable overlap. Using 
discriminate analysis on a subset of shell 
measurements and opercular measurements, 
F. variabilis form В separated rather well from 
F. variabilis form A and F. cónica (Figs. 28- 
30; Table 8). 

Fonscochlea variabilis form С 

(Figs. 19c, 22a, shell) 

Diagnosis: Shell similar to F. variabilis form 
В but typically relatively broader than most 
populations of that form (width/length ratio 
0.60-0.62), thicker (I.e. more solid) and 
sometimes larger (length 2.31-3.48 mm; 
mean 2.60 mm, males; 2.84, females). Oper- 
culum with well-developed pegs and long cal- 
careous smear. 

Shell (Figs. 19c, 22a) with 2.25-3.25 



(mean 2.84, males; 2.96, females) teleoconch 
whorls, convexity raiio 0.16-0.25 (mean 
0.23, males; 0.20, females), see diagnosis for 
other details. See Table 19A for measure- 
ments. Colour brown to reddish-brown. 

Operculum with 3-6 (mean 4.36, males; 
4.44, females) well-developed pegs 0.11- 
0.17 mm (mean 0.13 mm, males; 0.16 mm, 
females) in height, calcareous smear 0.31- 
0.50 mm (mean 0.37, males; 0.43, females), 
generally longer than in other forms of this 
species (but close to F. variabilis form B) and 
therefore PC/OL ratio significantly different. 
See Table 19A for measurement details. 

Radula as for species. See Table 3 for data. 

Head-foot as for species. Not examined in 
living material. 

Anatomy as for species. See Tables 1 7B-C 
for measurements. 

Voucher material: primary voucher speci- 
men (Fig. 22a) (SAM, D. 17913, stn 664B); 
additional material from same station (045, 
AMS, C.1 52882, 1, figured; 045, AMS, 
С 152883, many; 664A2, AMS, С 152884, 
many; 664A1, AMS, C.1 52889, 16; 664B, 
AMS, C.1 52885, many); 665A, AMS, 
С 152886, many; 665B, AMS, С 152887, 
many; 665C, AMS, C.1 52888, 50; 046, AMS, 
С 152890, 5. 

Dimensions of primary voucher specimen: 
length 3.25 mm, width 2.00 mm, length of ap- 
erture 1.50 mm. 

Localities: Freeling Springs (042-043, 
045-046, 663, 664A,B, 665A-C) (Fig. 26). 

Remarks: Specimens of this form are 



46 



PONDER, HERSHLER & JENKINS 




Imnn 



FIG. 23. Shells of Fonscochlea billakalina and F. variabilis. 

a-c. Fonscochlea billakalina. Strangways Springs (679), showing size variation (AMS, C.1 52967). 

d-e. Fonscochlea variabilis form A, Blanche Cup Spring (739), showing size variation (paratypes, AMS, 

C.152931). 

f. Fonscochlea variabilis, зглаН form from Blanche Cup Spring (739) (AMS, С 155863). 

g. Fonscochlea billakalina, Strangways Sphngs (678) (AMS, C.1 52969). 

h,i. Fonscochlea variabilis form B. Twelve Mile Spring (037), showing size variation (AMS, C.1 52967). 



readily distinguished from other populations 
of F. variabilis on shell characters despite a 



small number of quantifiable differences. Dis- 
criminate analysis separated the single mea- 



AUSTRALIAN SPRING HYDROBIIDS 



47 



sured population of this form from the rest of 
the small aquatics (Figs. 28-30; Table 8). 
This is one of four "taxa" endemic to Freeling 
Springs. 

Fonscochlea cónica n.sp. 

Derivation: a reference to the conical shape 
of the shell. 

(Figs. 22b, 53b,f, shell; 9f, protoconch; 24f,g, 
head-foot; 27b, female genitalia). 

Diagnosis: Shell small (1 .41-2.83 mm long; 
mean 1.94 mm, males; 2.07 mm, females), 
conical, with 2.0-3.2 (mean 2.57, males; 
2.67, females) weakly to moderately convex 
(convexity ratio 0.04-0.24; mean 0.13, 
males; 0.16, females) teleoconch whorls. Ap- 
erture not expanded; inner lip narrow, usually 
attached to parietal wall; outer lip slightly 
prosocline. Colour of shell ranges from yel- 
lowish brown to dark brown or orange-brown. 
Operculum with strong pegs. Head-foot lightly 
pigmented except for black triangle behind 
eyes. 

Shell (Figs. 22b, 53b, f; 9f, protoconch), see 
diagnosis. Measurement data in Table 19A. 

Operculum with 1-5 (mean 2.48, males; 
2.64, females) strong pegs 0.05-0.17 mm 
(mean 0.10 mm, males; 0.1 1 mm, females) in 
height, calcareous area 0.08-0.29 mm 
(mean 0.17 mm, males; 0.18 mm, females) 
long. See Table 19A for measurement data. 

Radula as for genus. Inner marginal teeth 
with 14-18 cusps. See Table 3 for other de- 
tails. 

Head-foot (Fig. 24f,g) is lightly pigmented 
with grey or pale grey, snout and cephalic ten- 
tacles very pale grey or unpigmented. Con- 
spicuous black triangle behind eyes. Cephalic 
tentacles with inconspicuous pale dorsal line 
in posterior quarter to half. 

Anatomy (Fig. 27b, female genitalia) very 
similar to that of F. variabilis except in size- 
related characters. See Tables 19B-E for 
measurement data. 

Type material holotype (Fig. 22b) (SAM, 
D.I 791 4, stn 003); and paratypes (003, AMS, 
C.I 52895, many; 756A, AMS, C.I 52896, 6; 
756B, AMS, C.I 52897, many; 756C, AMS, 
C.I 52898, many). 

Dimensions of holotype: length 2.15 mm, 
width 1.16 mm, length of aperture 0.90 mm. 

Localities: Southern Springs: Welcome 
Springs (003, 755A,B,D, 756A-C), Davenport 
Springs (004, 005, 753A,B), Old Woman 
Spring (733B). Shells have been found at Fin- 



niss Swamp West (690), Venable Spring 
(687) and Priscilla Spring (686). 

Middle Springs: Horse Springs East 
(747A,B, 748A-C), Horse Springs West 
(746), Strangways Spring (007, 745A), an un- 
named spring in Blanche Cup Group (739, 
785, 787), Coward Springs (019-022, 023, 
764A-C), Kewson Hill (741, 742A, 765), Julie 
Springs (772A,B,D, 773A-C), Elizabeth 
Springs (024, 766A,C-E, 771 A-C), Jersey 
Springs (025, 683A,B, 768A,B, 769A,B, 
770A,B), Warburton Spring (681 A-C, 682). 
Beresford Spring (028) (Fig. 26). 

Remarks: The shells of the specimens as- 
signed to this species are smaller, more com- 
pact and more solid than are those of most 
specimens of F. variabilis. These two species 
do not occupy the same spring groups, ex- 
cept in the Blanche Cup Group in which F. 
cónica is found in small springs and F. vari- 
abilis form A in the larger springs. 

The smaller, more conical shells and pale 
head-foot serve to distinguish this species 
from F variabilis in the Blanche Cup Group 
and elsewhere. Because the protoconchs in 
both species have a similar diameter, the PD/ 
SH ratio is significantly larger in nearly all 
populations of F cónica compared with F 
variabilis, reflecting the generally larger shell 
of F variabilis. The radulae also differ in the 
two species, F cónica having more cusps on 
the inner lateral teeth than do most speci- 
mens examined in the F variabilis complex. 

Discriminate analysis using shell and oper- 
cular measurements separated the popula- 
tions of F cónica and F variabilis well, al- 
though there is minor overlap with F variabilis 
form A in the plot using the first and second 
axes. F variabilis form В is well separated 
except in the plot using the second and third 
axes. 

Despite the lack of any single character that 
consistently and significantly separates all in- 
dividuals of F cónica from all individuals of F 
variabilis, they are recognised as distinct spe- 
cies because of their virtually sympatric asso- 
ciation in the Blanche Cup Group. The differ- 
ences in radulae and in the pigmentation of 
the head-foot noted above reinforce the re- 
sults of the discriminate analysis using the 
quantifiable shell and opercular differences. It 
is, however, freely admitted that the relation- 
ships of all of the small Fonscochlea are by no 
means clear and further analysis using elec- 
trophoretic methods is required to resolve the 
somewhat tentative arrangement proposed 
here. 



48 



PONDER, HERSHLER & JENKINS 




FIG. 24. Dorsal views of heads of species of Fonscochlea and Trochidrobia punicea. All figures except i from 
living material. 

a, d. Fonscochlea variabilis, Blanche Cup Spring, a, form A, typical; d, small form. 

b. Fonscochlea variabilis form A, Bubbler Spring, right tentacle only, showing the unpigmented stripe on the 
tentacle in this population. The remainder of the head is similar to that in a. 

с Fonscochlea billakalina, Old Billa Kalina Spring, 
e. Fonscochlea vahabilis form A, Coward Springs Railway Bore. 
f,g. Fonscochlea cónica; f, Welcome Springs; g, Elizabeth Springs, 
h. Trochidrobia punicea, Blanche Cup Spring. 

i. Fonscochlea aquatica cf. form A, Elizabeth Springs, showing abnormal tentacle development (from pre- 
served specimen). 
Scale: 0.2mm. 



Fonscochlea billakalina n.sp. 

Derivation: refers to Billakalina Station on 
which many of the springs containing this 
species are found. 
(Figs. 7d, 19d, 23a-c,g, 25a,c-g, shell; 8d, 



operculum; 24c, head-foot; 27c, female gen- 
italia). 

Diagnosis: Shell similar to F. variabilis and 
F. cónica but operculum differs markedly in 
having very weak to absent pegs. 

Shell (Figs. 7d, 19d, 23a-c,g, 25a,c-g) with 



AUSTRALIAN SPRING HYDROBIIDS 



49 




FIG. 25. Shells of Fonscochlea billakalina and F. variabilis form B. 

Fonscochlea billakalina: 

a,g. Billa Kalina springs, a, (759) (paratypes, AMS, C.1 52930); g, (763) (AMS, C.I 52964). 

с Old Billa Kalina Spring (027) (paratype, AMS, C.1 52963). 

d,e. Francis Swamp, d, (721) (AMS, C.1 52966); e, (720) (AMS, C.1 52968). 

f. Fenced Spring, Billa Kalina (723) (AMS, С 152965). 

Fonscochlea variabilis form B: 

b. Hawker Springs (673) (AMS, C.1 52970). 



two forms present. One form (Figs. 23a-c,g, 
25a,d,f,g) with shell similar to that of F. vari- 
abilis form В and 1 .9-2.4mm in length; other 
form (Figs. 7d, 25c), restricted to spring at Old 
Billa Kalina Homestead ruin (027, 759), is 
similar to F. variabilis form A but is larger 
(2.8-3.2 mm long compared with 1.8- 
2.8 mm). Overall mean shell length 2.60 mm 
(males) and 2.64 mm (females). Teleoconch 
whorls 2.30-3.38 (mean 2.75, male; 2.77, 
female), convexity ratio 0.03-0.24 (mean 
0.15, male; 0.14, females). Measurement 
data in Table 19A. 
Operculum (Fig. 8d) with 0-5 (mean 1.40, 



males; 1.59, females) small pegs 0.02-0.14 
mm (mean 0.07 mm) in height; calcareous 
area 0-0.33 mm (mean 0.13 mm) long. See 
Table 19A for measurement data. 

Radula, see Table 3 for data. 

Head-foot (Fig. 24c) as for species. Back- 
ground pigmentation of snout and tentacles 
dark grey to black. 

Anatomy (Fig. 27c, female genitalia) as for 
species. See Tables 19B-E for measurement 
data. 

Type material: holotype (Fig. 19d) (SAM, 
D.I 7911, stn 027); and paratypes (SAM, 
D.2034, 30; SAM, D.2035, 32; 759B, AMS, 



50 



PONDER, HERSHLER & JENKINS 




AUSTRALIAN SPRING HYDROBIIDS 



51 







FIG. 27. Female genitalia of species of Fonscochlea. 

a,d. Fonscochlea variabilis form A. The Bubbler Spring, d, detail of sperm sacs. 

b. Fonscochlea cónica, Horse Spring East. 

с Fonscochlea billakalina, Old Billa Kalina Spring. 

ag, albumen gland; be, bursa copulathx; eg, capsule gland; cv, coiled oviduct; go, oviduct opening; mcp, 

posterior limit of palliai cavity; sr, seminal receptacle; vc, ventral channel; vcp, posterior extension of ventral 

channel. 

Scale: 0.25mm. 



C.1 52891, 18; 759B, AMS, C.I 52892, many; 
026, AMS, C.I 52893, many, С 152995, 1, 
figured; 027, AMS, C.I 52894, many, 
C.I 52963, 1, figured; 759, AMS, C.I 52930, 1, 
figured). 

Dimensions of holotype: length 2.78 mm, 
width 1.68 mm, length of aperture 1.33 mm. 

Localities: South Western Springs: Billa 
Kalina Springs (026-027, 723A-D, 758C, 
759A-C, 760, 761, 763A,B), Francis Sv\/amp 
(717A,B, 720A-C, 721 A-C), Strangways 
Springs (029, 030, 678A,B, 679A-C, 680). 
Shells from Welcome Bore/Spring (758) and 
Margaret Spring (722) might belong to this 
form (Fig. 26). 

Remarks: The two shell forms seen in pop- 
ulations included in this taxon are, when ex- 
tremes are examined, readily distinguished. 
Intermediate specimens, however, do occur 
in some populations. 

The shell characters are virtually identical, 



in most populations, with those of F. variabilis 
form В but that taxon can be readily distin- 
guished by its strong opercular pegs. With 
discriminate analysis, using shell and opercu- 
lar measurements, this species is clearly dif- 
ferentiated from the other small aquatic taxa 
(Figs. 28, 29, 30; Table 8). 

This taxon is recognised as a species be- 
cause of the considerable differences be- 
tween its operculum and those of the other 
small aquatic taxa. The lack of obvious corre- 
lated shell or anatomical characters is, in this 
case, judged to be outweighed by the strongly 
diagnostic opercular characters. 

Discrimination of the small aquatic taxa (in- 
cluding the geographic forms) of Fonscochlea 
was tested using discriminate analysis on 
shell and opercular measurements. The re- 
sults showed that all groups could be discrim- 
inated using these data, 87% of the measured 
specimens (n = 617) being correctly classi- 



52 



PONDER, HERSHLER & JENKINS 




FIG. 28. Plot of group centroids, using first two canonical axes, obtained from discriminate analysis of 
populations of small aquatic species and forms of Fonscochlea using shell and opercular measurements. 
Males and females of each population are, for the purposes of this analysis, treated as distinct populations. 
The axes contain the following percentages of the variance of the variables used: first (horizontal) axis: SH, 
27.96%; SW, 3.82%; AH, 59.38%; TW, 8.45%; OL, 75.41%; PH, 57.85%; PC, 5.26%; PN, 1.00%. Second 
(vertical) axis: SH, 18.18%; SW, 42.38%; AH, 5.27%; TW, 48.72%; OL, 14.38%; PH, 23.02%; PC, 72.13%; 
PN, 54.37%. b, F. blllakalina; с, F. cónica; e, F. variabilis form В; g, F. variabilis form С; v, F variabilis form 
A; s, F variabilis, small Blanche Cup form. 



AUSTRALIAN SPRING HYDROBIIDS 



53 




FIG. 29. Plot of group centroids, using first and third canonical axes, obtained from discriminate analysis of 
populations of small aquatic species and forms of Fonscochlea using sfiell and opercular measurements. 
Males and females of each population are, for the purposes of this analysis, treated as distinct populations. 
The axes contain the following percentages of the variance of the variables used: first (horizontal) axis: SH, 
27.96%; SW, 3.82%; AH, 59.38%; TW, 8.45%; OL, 75.41%; PH, 57.85%; PC, 5.26%; PN, 1.00%. Third 
(vertical) axis: SH, 14.76%; SW, 3.03%; AH, 13.22%; TW, 5.61%; OL, 1.30%; PH, 9.07%; PC, 0.39%; PN, 
1 .98%. b, F. billakalina; с, F. cónica: e, F. variabilis form В; g, F. variabilis form С; v, F. variabilis form A; s, 
F. variabilis, small Blanche Cup form. 



tied. The Euclidian (taxonomic) distances be- 
tween the groups are given in Table 8. The 
greatest distance score achieved between all 
pairwise comparisons was 3.95, between the 
snnall Blanche Cup form of F. variabilis and F. 
variabilis form B. Differences between the 



species was >1 in all cases except between 
F. variabilis form A and F. cónica (score 
>0.69). F. cónica separated from the other 
forms of F. variabilis with scores >1 .1 . F. bil- 
lakalina had a distance score of > 1 .58 when 
compared with all other groups. Within F. vari- 



54 



PONDER, HERSHLER & JENKINS 




FIG. 30. Plot of group centroids, using second and third canonical axes, obtained from discriminate analysis 
of populations of small aquatic species and forms of Fonscochlea using shell and opercular measurements. 
Males and females of each population are, for the purposes of this analysis, treated as distinct populations. 
The axes contain the following percentages of the variance of the variables used: second (horizontal) axis: 
SH, 18.18%; SW, 42.38%; AH, 5.27%; TW, 48.72%; OL, 14.38%; PH, 23.02%; PC, 72.13%; PN, 54.37%. 
Third (vertical) axis: SH, 14.76%; SW, 3.03%; AH, 13.22%; TW, 5.61%; OL, 1.30%; PH, 9.07%; PC, 0.39%; 
PN, 1 .98%. b, F. billakalina; с, F. cónica; e, F. variabilis form В; g, F variabilis form С; v, F. variabilis form 
A; s, F. variabilis, small Blanche Cup form. 



abilis the scores separating the fornns were 
>0.44, the lowest scores being achieved be- 
tween forms В and С (0.44 females, 0.50 
males), the comparisons between the other 
forms being > 1 . 



SNK tests (5% level) using pooled data, 
combined and separate sexes, for each vari- 
able used in the discriminate analyses gave 
the following results: 

SH — Combined sexes: significantly differ- 



AUSTRALIAN SPRING HYDROBIIDS 



55 



ent for all except F. variabilis form С and F. 
variabilis form B. Separate sexes: only the 
small Blanche Cup form of F. variabilis and F. 
cónica were clearly distinct, the others form- 
ing overlapping subsets. The means for this 
character were significantly different between 
males and females for all species and forms 
(females larger) except the small Blanche 
Cup form and F. billakalina. 

SW — Combined sexes: all means signifi- 
cantly different. Separate sexes: the small 
Blanche Cup form, F. cónica and F. variabilis 
form A all form separate subgroups but the 
others are included in overlapping subgroups. 
All except the small Blanche Cup form are 
sexually dimorphic (females larger), the 
means in all cases being significantly differ- 
ent. 

AH — Combined sexes: significantly differ- 
ent for all species and forms except F vari- 
abilis form В and F billakalina. Separate 
sexes: same results as for SH. 

TW — Combined sexes: significantly differ- 
ent for all except F variabilis form A + F 
billakalina and F variabilis form С + F vari- 
abilis form B. Separate sexes: the small 
Blanche Cup form is distinctly different but all 
other groups, except males of F cónica, form 
overlapping subsets. This character does not 
significantly differ between males and fe- 
males except in F cónica. 

OL — Sexes combined: same result as AH. 
Separate sexes: the small Blanche Cup form 
and F cónica formed distinct groups as did 
males of F variabilis form A and females of F 
variabilis form С All other groups formed 
overlapping subsets. Differences between the 
sexes in this character were statistically sig- 
nificant in F cónica, F. variabilis form A, F 
variabilis form С and F variabilis form B. 

PH — Combined sexes: significantly differ- 
ent for all except F cónica and F. variabilis 
form B, and F variabilis form A and F vari- 
abilis form С Separate sexes: all form over- 
lapping subsets except males of F variabilis 
form С and F variabilis form A which form 
their own group, as do the females of these 
two forms, these also being the only two 
groups to show significant sexual dimorphism 
in this character. 

PC — Combined sexes: all means signifi- 
cantly different. Separate sexes: the small 
Blanche Cup form and F billakalina are not 
significantly different but all others are. Sex- 
ual dimorphism is exhibited in F variabilis 
form В and F variabilis form С 

PN — Combined sexes: significantly differ- 



ent for all except the small Blanche Cup form 
and F billakalina. Separate sexes: the same 
result but with F variabilis form В and F vari- 
abilis form С not discriminated. Only F vari- 
abilis form В shows significant sexual dimor- 
phism in this character. 

Subgenus Wolfgangia n.subgen. 

Derivation: named for Wolfgang Zeidler 
(Fem.). 

Type species: F (W.) zeidleri n.sp. 

Diagnosis: Shell (Figs. 6e-h, 7a,b, 14a,c, 
53a,d) as for genus; differs from Fonscochlea 
s.S. in being rather thick-shelled, aperture 
with thickened peristome and protoconch mi- 
crosculpture consisting of spiral lines (Fig. 
9c,d). 

Operculum (Fig. 8a,b) with prominent pegs. 

Radula (Fig. 10a,b) as for genus. Central 
teeth always with two pairs of basal cusps. 

Head-foot (Fig. 11 a, b) with cephalic tenta- 
cles about same length as snout or slightly 
shorter. 

Anatomy: Female genital system (Figs. 
12a,b, 47) as for genus except oviduct be- 
tween capsule gland and bursal duct always 
straight and sperm sacs lie dorsal to muscular 
oviduct. Ducts of sperm sacs ventral to sacs. 
Male (Fig. 46b, penis) system as for genus. 

Remarks: The species included in this sub- 
genus can be divided into two morphologi- 
cally similar forms, one of which is amphibi- 
ous and the other aquatic. The amphibious 
form is the most widely distributed of the 
mound-spring snails; the other, one of the 
most restricted, is confined to a single spring. 

The differences in the protoconch micro- 
sculpture, and in the female genital tract, to- 
gether with the relatively larger snout and 
shorter tentacles possessed by F (W.) zeid- 
leri, are characters that separate this species 
from the remainder of those in the genus. This 
species does, however, possess several key 
features in common with species of Fons- 
cochlea s.S., the equal-sized sperm sacs be- 
ing the most outstanding. For this reason, and 
because there do not appear to be any inter- 
grading states represented in any of the 
known species, F (W.) zeidleri is judged to be 
subgenerically separable from Fonscochlea. 

This subgenus and its type species are 
named for Wolfgang Zeidler of the South Aus- 
tralian Museum, Adelaide, who first intro- 
duced the senior author to the mound springs 
and since then has assisted with this project 
in many ways. 



56 



PONDER, HERSHLER & JENKINS 



Fonscochlea (Wolfgangia) zeidleri n.sp. 

Diagnosis: As for subgenus description. 

The typical form of this species is described 
below as "form A" where a holotype is des- 
ignated for the species. 

Localities: Oodnadatta Complex, Northern, 
Middle, Western and Southern Springs (Fig. 
31). 

Remarks: The characters separating the 
subgenus Wolfgangia from species of Fons- 
cochlea s.S. also serve to separate this spe- 
cies. The shell of this species is similar to that 
of the two large aquatic species of Fonscoch- 
lea, F. accepta and F. aquatica, in size and 
shape but can be distinguished by its thicker 
peristome, with the inner lip separated from 
the parietal wall, and its more convex whorls. 
Two geographic forms are recognised and 
additional details are given under the descrip- 
tions of each of them. 

Fonscochlea (Wolfgangia) zeidleri form A. 

(Figs. 6e-h, 7a, 14a, 53a, d, shell; 9c, d, pro- 
toconch; 8b, operculum; 10b, radula; 11a,b, 
head-foot; 12b, 47, female genitalia; 46b, pe- 
nis; 45, stomach) 

Diagnosis: Shell large for genus, up to 
about 5.3 mm long, solid, width/length ratio 
0.55-0.7 (usually 0.6-0.65) with 3-4.4 con- 
vex (convexity ratio 0.04-0.26; mean 0.16, 
males; 0.18, females) teleoconch whorls 
sculptured with distinct growth lines and, in 
some specimens, faint spiral scratches. Pro- 
toconch microsculpture (Fig. 9c, d) of fine, 
closely-spaced, irregular spiral lines. Aperture 
with thickened peristome, inner lip thickened 
and separated from parietal wall; outer lip or- 
thocline to opisthocline, edge blunt. Colour 
yellowish brown to purplish brown. Opercu- 
lum thick, with prominent pegs. 

Shell (Figs. 6e-h, 7a, 14a, 53a,d; 9c,d, pro- 
toconch microsculpture), see diagnosis. See 
Table 20A for measurement data. 

Operculum (Fig. 8b) thick, with 2-6 (mean 
4.02) heavy opercular pegs. See Table 20A 
for measurement data. 

Radula (Fig. 1 0b) as for subgenus. See Ta- 
ble 3 for data. 

Head-foot (Fig. 1 1 a,b) variable in degree of 
pigmentation; snout long and mobile, with 
well-developed concentric ridges. Cephalic 
tentacles tapering, about same length as 
snout or slightly shorter. Usually an unpig- 
mented area around eyes; tentacles, in some 



populations, very pale and, in others, dark 
grey or black. 

Anatomy (Fig. 12b, 47, female genitalia; 
46b, penis; 45, stomach) as described for 
subgenus. See Tables 20B-E for measure- 
ments. 

Type material: holotype (Fig. 14a) (SAM, 
D.I 791 5, stn 764C); and paratypes (SAM, 
D.3206, 61 ; 764A, AMS, C.I 52889, 13; 764C, 
AMS, C.I 52900, many; 020, AMS, C.I 52901, 
20; 021, AMS, C.I 52902, many; 022, AMS, 
C.I 52903, many; 023, AMS, C.1 52904, 
many; 019, AMS, C.I 52928, 6). 

Dimensions of holotype: length 4.82 mm, 
width 2.87 mm, length of aperture 1.85 mm. 

Localities: Southern Springs: Welcome 
Springs (754A, 755A,B,D; 756, shells only), 
Hermit Hill Springs (71 1 A,B, 712), Old Finniss 
Springs (693A, 694B,C), Old Woman Springs 
(732, 733), Finniss Swamp West (690A,C). 
Shells have been collected from Phscilla 
Spring (686), Venable Spring (687) and an 
unnamed spring in Lake Eyre South (702). 

Middle Springs: Horse Springs West 
(746A,B), Horse Springs East (748B,C), Mt. 
Hamilton Homestead ruins (006, 749), 
Strangways Spring (007, 745A), Blanche Cup 
Spring (008-01 2, 685), Bubbler Spring (01 3- 
017), Little Bubbler Spring (744), unnamed 
springs, Blanche Cup Group (785, 786, 787), 
Coward Springs (019-023, 764A-C), Cow- 
ard Springs Railway Bore (018, 684, 743), 
Kewson Hill Springs (740A, 741, 742A), Julie 
Springs (772A-D, 773A-C), Elizabeth 
Springs (024, 766A-G, 767, 771 A-C), Jersey 
Springs (025, 683A,B; 768, shells only; 
769A,B, 770A-C), Warburton Springs (681 A- 
C, 682), Beresford Spring (028). Fossil shells 
have been collected from the top of Hamilton 
Hill. 

South Western Springs: Billa Kalina 
Springs (026, 027, 723A,C,D; 759, shells 
only; 760, 763A,B), Francis Swamp (71 7B, 
720A,B, 721 B,C), Strangways Springs (029- 
030, 678A,B, 679A-C, 680). Shells from Mar- 
garet Spring (722) and Welcome Bore (758). 

Northern Springs: Brinkley Spring (677), 
Hawker Springs (670A-C, 671 , 672C,D, 673), 
Big Perry Springs (034), Twelve Mile Spring 
(036, 037), Outside Springs (039). Shells 
from Spring Hill Springs (674). 

Freeling Springs (043, 046, 664A-C, 
665A-C). 

Remarks: This form, the most widely dis- 
tributed of the mound-spring snails, is of spe- 
cial interest because of its amphibious habit. 
It lives, in most springs, along the edges of 



AUSTRALIAN SPRING HYDROBIIDS 



57 




58 



PONDER, HERSHLER & JENKINS 



the outflows where it is either exposed, as on 
the hard substrates found on the calcareous 
nnounds, or partly or completely buried in the 
sediment. The preference for burrowing in the 
substrate appears to differ between spring 
groups and might not be due entirely to sub- 
strate differences. For example the popula- 
tions of this species at Hermit Hill are ex- 
tremely cryptic, mainly because of this habit, 
whereas at Welcome Springs, with similar 
substrate available, they are much more con- 
spicuous, large numbers being present on the 
surface. 

Populations at Kewson Hill and Elizabeth 
Springs have two recognisable phenotypes. 
One is the typical shell form (Fig. 53d) indis- 
tinguishable from specimens found else- 
where. Another form (Fig. 53a) is shorter, 
darker, relatively broader, and with a relatively 
larger aperture than the typical form. These 
two forms have been found living together but 
usually occupying different microhabitats. The 
typical form is found along the edges of the 
outflow and around the head of the spring or 
seepage, the normal habitat for this species, 
whereas the squat form is invariably found in 
the outflows where it lives attached to any 
available emergent substrate, usually in very 
large numbers. Some individuals are found in 
the water but most are out of it. Some other 
populations (e.g., Blanche Cup and Horse 
Springs East) contain many intermediates be- 
tween these two types (Fig. 6e-h). 

Fonscochlea (Wolfgangia) zeidleri form B. 

(Figs. 7b, 14c, shell; 8a, operculum; 10a, rad- 
ula; 12a, female genitalia) 

Diagnosis: Shell smaller than typical spec- 
imens of F. (W.) zeidleri form A (up to 4.06 
mm long) and with relatively broader (shell 
width/shell length 0.63-0.65) than many pop- 
ulations of F. (W.) zeidleri form A. 2.9-3.5 
convex (convexity ratio 0.14-0.22) teleo- 
conch whorls. Aperture with orthocline outer 
lip. Value of aperture length/shell length sig- 
nificantly larger than in most populations of F. 
(W.) zeidleri form A. Colour dark brown. 

Shell (Figs. 7b, 1 4c), see diagnosis. See Ta- 
ble 20A for measurement data. 

Operculum (Fig. 8a) with 2-6 (mean 3.85, 
males; 3.4, females) prominent opercular 
pegs. See Table 20A for measurement data. 

Radula (Fig. 1 0a) as for subgenus. See Ta- 
ble 3 for data. 



Head-foot similar to that of F (W.) zeidleri 
form A but, in most specimens, weakly pig- 
mented except for large patch of black pig- 
ment behind eyes. Snout and cephalic tenta- 
cles lack pigment in some specimens but in a 
few are darkly pigmented. 

Anatomy (Fig. 12a, female genitalia) as de- 
scribed for subgenus. See Tables 20B-E for 
measurements. 

Voucher material: primary voucher speci- 
men (Fig. 14c) (SAM, D.17916, stn 661); and 
material from the same population (661, 
SAM, D.I 7945, many; AMS, С 152905, 
many, C.I 52993, 1, figured). 

Dimensions of primary voucher specimen: 
length 4.04 mm, width 2.56 mm, length of ap- 
erture 1.72 mm. This is one of the largest 
specimens of this form. 

Locality: Oodnadatta Spring Complex: Big 
Cadnaowie Spring (661). 

Remarks: This population is distinguished 
as a separate form, despite few morphologi- 
cal differences, because it is considerably 
geographically isolated, has a distinctive shell 
shape (although duplicated in a few examples 
of F (W.) zeidleri form A) and its fully aquatic 
habit is a considerable departure from the 
amphibious habit of the typical form. The lack 
of significant morphological differentiation 
suggests that it is probably only recently de- 
rived from F (W.) zeidleri form A. 

The populations of F (W.) zeidleri form A 
that develop squat shells with width/length ra- 
tios similar to those of F (W.) zeidleri form В 
are virtually all associated with harsh environ- 
ments, e.g., the Kewson Hill Springs (Fig. 
53a). The conditions that appear to bring 
about the shortening of the shell in F (W.) 
zeidleri form A, small, shallow outflows and 
hard substrate, are not those in which F (W.) 
zeidleri form В is found. This form lives in a 
large, degraded spring in a few metres of 
sedges in a narrow, outflow with a significant 
flow of water. It is completely aquatic and very 
abundant in this part of the habitat. A very few 
individuals were found in the remainder of the 
spring, which has been severely damaged by 
livestock. This spring has since been fenced 
as part of the mound-spring fencing pro- 
gramme, mainly because of the reported ex- 
istence of this unusual population (Ponder & 
Hershler, 1984). 

Discrimination of the two forms of Fonscoc/7- 
lea zeidleri was tested using discriminate 
analysis on a subset of shell measurements. 



AUSTRALIAN SPRING HYDROBIIDS 



59 




FIG. 32. Plot of group centroids, using first two canonical axes, obtained fronn discriminate analysis of 
populations of Fonscochlea zeidleri using shell nneasurements. Males and fernales of each population are, 
for the purposes of this analysis, treated as distinct populations. The axes contain the following percentages 
of the variance of the variables used: first (horizontal) axis: SH, 82.65%; SW, 0.29%; AH, 69.38%; TW, 
18.79%. Second (vertical) axis: SH, 2.19%; SW, 55.66%; AH, 21.12%; TW, 47.45%. 
Closed circles, F. zeidleri form A; open circles, F. zeidleri form B. 



The results (Fig. 32) showed that both groups 
could be discriminated using these data, 92% 
of the measured specimens (n = 284) being 
correctly classified. 

SNK tests (5% level) using pooled data, 
combined and separate sexes, for each vari- 
able used in the discriminate analyses gave 
the following results: 

SH, AH and TW were significantly different 
for combined sexes of both forms. No char- 
acters separated the two forms using separate 
male and female data. Sexual dimorphism 
was apparent only in TW for both forms. 

Genus Trochidrobia n.gen. 

Derivation: Trochi (Latin), a child's hoop, 
and used for a genus of gastropods (Tro- 



chus), pertaining to the shape of the shell; 
drobia, from Hydrobia, the type genus of Hy- 
drobiidae (fem.). 

Type species: Trochidrobia punicea n.sp. 

Distribution: Artesian springs between Mar- 
rée and Oodnadatta, northern South Austra- 
lia. 

Diagnosis: Shell (Figs. 33, 37) of known 
species small (as much as 2mm in dia- 
meter), trochiform to depressed-trochiform, 
umbilicate, smooth, with only sculpture weak 
axial growth lines. Protoconch (Fig. 34) of 
about one and one-half whorls, sculptured 
with irregular minute pits, or pits and spiral 
threads. Aperture oval, peristome thin, no ex- 
ternal varix; outer lip simple, not expanded or 
flared, with thin edge. Periostracum smooth, 
thin. 



60 



PONDER, HERSHLER & JENKINS 






1mm 




FIG. 33. Shells of Trochidrobia. 

a-c. Trochidrobia punicea, holotype. Blanche Cup Spring (009). 

d-f. Trochidrobia smithi, holotype. Twelve Mile Spring (036). 



Operculum (Fig. 35a,c,e,f) corneous, oval, 
nucleus subcentral, thin, simple. 
Radula (Fig. 35b,d) with central teeth 



formula 



4-8+1 +4-£ 



lateral teeth 3-6 



1-2 1-2 

1+4-7, inner marginal teeth with 18-31 



cusps, outer marginal teeth with many small 
cusps. 

Head-foot (Fig. 24h) with cephalic tenta- 
cles longer than snout, parallel sided, in- 
conspicuously ciliated ventrally. Pigmenta- 
tion usually dense, pigment granules black 



AUSTRALIAN SPRING HYDROBIIDS 



61 




FIG. 34. Protoconchs of species of Trochidrobia. 
a,b. Trochidrobia punicea, Coward Springs (020). 
c,d. Trochidrobia smithi, The Fountain Spring (032). 
e,g. Trochidrobia minuta, Freeling Springs (045). 
f. Trochidrobia inflata, Freeling Springs (043). 
Scale: a,c,f,g = 0.1mm; b,d,e = 0.01mm. 



and white. General head-foot typical of 
family. 

Anatomy: palliai cavity (Fig. 48) with well- 
developed ctenidium; osphradium oval, about 
2-4 times as long as broad and about one- 



half to one-third length of ctenidium, its pos- 
terior extremity situated near posterior end of 
ctenidium. 

Female reproductive system (Figs. 36, 38) 
with single sperm sac and coiled oviduct lying 



62 



PONDER, HERSHLER & JENKINS 




FIG. 35. Radulae and opercula of Trochidrobia. 

a. Operculum of Trochidrobia punicea, Blanche Cup Spring (008). 

b. Radula of Trochidrobia punicea, Welcome Springs (002). 
с Operculum of Trochidrobia inflate, Freeling Springs (043). 

d. Radula of Trochidrobia smithi. Old Billa Kalina Spring (027). 

e. Operculum of Trochidrobia smithi. Old Billa Kalina Spring (027). 

f. Operculum of Trochidrobia minuta, Freeling Springs (045). 
Scale: a,c,d,e = 0.1mm; b,d = 0.01mm. 



AUSTRALIAN SPRING HYDROBIIDS 



63 




FIG. 36. Female genitalia of species of Trochidrobia. 

a. Trochidrobia smithi. Outside Springs (039). 

b. Trochidrobia punicea. Strangways Spring, E. of Blanche Cup (007). 

ag, albumen gland; be, bursa copulatrix; eg, capsule gland; со, coiled part of oviduct; go, oviduct opening; 
int, intestine; mcp, posterior limit of palliai cavity; r, rectum; st, tissue connection between oviduct and 
pericardium; uo, upper oviduct; vc, ventral channel. 
Scale: 0.1mm. 



64 



PONDER, HERSHLER & JENKINS 



on inner (left) side of albumen gland or mainly 
situated behind this gland. Coiled part of ovi- 
duct an unpigmented tube lying largely in 
front of large bursa copulatrix. Bursa copula- 
trix about one-third to one-half of length of 
albumen gland, its narrow duct opens to ovi- 
duct in different locations depending on spe- 
cies. Gonopericardial duct absent but repre- 
sented by strand of tissue. Oviduct straight 
anterior to point of opening of bursal duct. Ac- 
cessory sperm storage occurs in swollen part 
of posterior ventral channel of capsule gland 
or in coiled oviduct. Capsule gland about 
same length as albumen gland to about half 
its length, with a well-developed ventral chan- 
nel containing ciliated lateral fold. Genital 
pore terminal, subterminal or placed at about 
one-third of distance along capsule gland. 
Egg capsules spherical, cemented in umbili- 
cus of shell with mucus (known only in T. pu- 
nicea). 

Male reproductive system with vas defer- 
ens complexly coiled beneath anterior part of 
testis. Palliai and visceral vas deferens enter 
and leave prostate gland in middle section. 
Prostate gland extends into palliai wall one- 
third to one-half of its length. Palliai vas def- 
erens a narrow, straight, ciliated tube lying 
just beneath epithelium on right side of palliai 
floor but undulates as it passes up right side 
of neck to enter base of penis. Penis (Fig. 49) 
with swollen basal portion and tapering distal 
portion. Basal part unpigmented, concentri- 
cally creased and narrow penial duct undu- 
lates within it. Distal portion smooth, usually 
pigmented, coiled anti-clockwise when at 
rest, penial duct straight within it, emerging at 
pointed distal extremity. 

Alimentary canal typical of family; buccal 
mass well developed with U-shaped radular 
sac protruding behind. Salivary glands sim- 
ple, tubular. Stomach (Fig. 44a) with distinct 
anterior and posterior chambers, anterior one 
larger, lacks caecal appendage. Style sac 
contains crystalline style, comprises about 
one-third to one-half of total length of stom- 
ach. Single digestive gland opening immedi- 
ately posterior to oesophageal opening. Di- 
gestive gland covers inside of right side of 
stomach to about halfway across anterior 
chamber. Intestine makes U-shaped fold on 
palliai roof in one species. 

Nervous system with left pleural and sub- 
oesophageal ganglia abutting and hght pleu- 
ral and supra-oesophageal ganglion sepa- 
rated by long connective. 



See anatomical account for further detail. 

Remarks: The species contained in Tro- 
chidrobia are similar in shell and opercular 
characters to those in the European Horatia- 
Pseudamnicola complex but differ in several 
important character states. These include the 
lack of a seminal receptacle (not one or two); 
a longer, coiled oviduct; two pairs of basal 
cusps on the central teeth of the radula (not a 
single pair); a penis having a slender, simple 
distal portion longer than the basal part (not 
shorter than the base); and the left pleural 
and suboesophageal ganglia abutting (not 
separated by a connective) (see Radoman, 
1966, 1983, for further detail regarding the 
European taxa). 

Some species in the Beddomeia complex 
in Tasmania, particularly Valvatasma tasma- 
nica (T. Woods, 1876), are similar to species 
of Trochidrobia in shell form. They differ, how- 
ever, in having an operculum with an eccen- 
tric nucleus and a radula with a single pair of 
basal cusps. All of the species in the Bed- 
domeia complex have a seminal receptacle. 
Another species similar to the Beddomeia 
group is Jardaniella ttiaanumi {PWsbry, 1900), 
from north Queensland. This species has two 
pairs of basal cusps on the radula, an eccen- 
tric opercular nucleus and a seminal recepta- 
cle (all data on Beddomeia group from Pon- 
der, unpublished). 

IHeterocyclus petiti (Crosse, 1872) from 
New Caledonia has a depressed, umbilicate 
shell but the outer lip is flared and the calcar- 
eous, multispiral operculum is of different 
construction (StarmCilner, 1970). It is unlikely 
that this species is even remotely related. 

The only other Australian genus of de- 
pressed shell form is Posticobia, which is re- 
lated to Hemistomia (see Ponder, 1981). i4or- 
atia nelsonensis Climo, 1977, from Nelson, 
New Zealand, is known only from shells but it 
is probable that this species is a depressed 
form of Opacinacola, a New Zealand genus 
normally having higher-spired shells. 

Trochidrobia punicea n.sp. 

Derivation: puniceus (Latin) purple, red. A 
reference to the dark purple-red colour of the 
shell of living specimens. 

(Figs. 33a-c, shell; 34a, b, protoconch; 35a, 
operculum; 35b, radula; 24h, head-foot; 48, 
palliai cavity; 44a, stomach; 36b, female gen- 
ital system; 49a, penis) 



AUSTRALIAN SPRING HYDROBIIDS 



65 



Diagnosis: Shell up to 2.22 mm in diameter, 
depressed (width/height ratio 1.1-1.3), with 
1.50-2.25 convex whorls and widely umbili- 
cate. Protoconch microsculpture (Fig. 34a, b) 
of close spiral ridges with irregular surface pit- 
ting over the entire surface. Aperture some- 
times separated from parietal wall. Colour yel- 
lowish brown to dark orange-brown. Female 
genitalia with very much thickened coiled ovi- 
duct, long bursal duct and simple ventral 
channel. Rectal arch absent in male (rectum 
lies alongside prostate gland). 

Shell (Fig. 33a-c), see diagnosis. See Ta- 
ble 21 A for measurement data. 

Operculum (Fig. 35a) as for genus. 

Radula (Fig. 35b) as for genus. See Table 3 
for data. 

Head-foot (Fig. 24h) variably pigmented, 
dark pigmentation common, usually with nar- 
row dorsal unpigmented stripe on proximal 
half of tentacles continuous with unpigmented 
zone around eyes. 

Anatomy (Figs. 48, palliai cavity; 44a, stom- 
ach; 36b, female genital system; 49a, penis), 
see anatomical section below for full descrip- 
tion. See Tables 21B-C for measurement 
data. 

Type material: holotype (Fig. 33a-c) (SAM, 
D.I 7922, stn 009); and paratypes (008, SAM, 
D.3208, 58; SAM, D.2030, 60; 739, AMS, 
С 152906, many; 009, AMS, С 152907, 
many; 008, AMS, C.I 52908, many; 010, 
AMS, C.I 52909, many; 01 1 , AMS, C.I 5291 0, 
20; 012, AMS, C.152911, 10). 

Dimensions of holotype: length 1.62 mm, 
width 2.08 mm, length of aperture 1.08 mm. 

Localities: Southern Springs: Welcome 
Springs (002, 003, 754A-D, 755A-D, 756A- 
C), Davenport Springs (005, 752A-C, 
753A,B), Hermit Hill Springs (712), Dead Boy 
Springs (689), Finniss Swamp West (690A- 
C, 691), Bopeechee Springs (692A,B), Old 
Finniss Springs (693A-C, 694A-C, 710), Old 
Woman Spring (733A-E), Sulphuric Springs 
(735, 737). Shells from Priscilla Spring (686), 
Venable Spring (687). 

Middle Springs: Horse Springs East 
(747A,B, 748A-C), Horse Springs West 
(746A), Mt. Hamilton Homestead ruins (749), 
Strangways Springs (007,745A,B), Blanche 
Cup Spring (008-012, 739), Bubbler Spring 
(013-017), Little Bubbler Spring (744A-C), 
an unnamed spring, Blanche Cup Group 
(785, 786, 787), Coward Springs (019-022, 
023, 764A-C), Kewson Hill Springs (741, 
742B, 765), Julie Springs (772A-D, 773A-C), 



Jersey Springs (025, 768A, 769A,B, 770A,B), 
Elizabeth Springs (024, 766A-E, 767A,B, 
771A,B) (Fig. 39). 

Fossil shells similar to this species have 
been collected from travertine on the top of 
Hamilton Hill. 

Remarks: The shell of this species is virtu- 
ally identical to that of T. smithi described be- 
low, the only characters, apart from proto- 
conch microsculpture (which has been 
examined in only a few specimens), distin- 
guishing these two species being anatomical 
ones. See under T. smithi for details. 

Both of these species are extremely abun- 
dant in most of the springs in which they oc- 
cur. They live in a variety of microhabitats and 
appear to be particularly abundant in shallow, 
firm-bottomed outflows. They are positively 
phototropic, living fully exposed in the out- 
flows. See physiology section below for more 
details. 

Trochidrobia smithi n.sp. 

Derivation: named for Dr. B.J. Smith. 
(Figs. 33d-f, shell; 34c, d, protoconch; 35e, 
operculum; 35d, radula; 36a, female genitalia; 
49b, penis) 

Diagnosis: Shell and head-foot virtually 
identical to those of T. punicea, maximum 
width of shell 2.13 mm, with 1.63-2.13 (mean 
1.92) teleoconch whorls. Protoconch micro- 
sculpture (Fig. 34d) of spirally arranged wrin- 
kles, weaker than spiral sculpture of T. puni- 
cea. Female genitalia with narrow coiled 
oviduct and expanded posterior part of ventral 
channel (Fig. 36a). Rectal arch present in 
male (rectum separated from prostate gland). 

Shell (Figs. 33d-f; 34c, d, protoconch, see 
diagnosis. See Table 21 A for measurements. 

Operculum (Fig. 35e) as for genus. 

Radula (Fig. 35d) as for genus. See Table 3 
for data. 

Head-foot very similar to that of T. punicea, 
variably pigmented, uniformly dark pigmenta- 
tion being common. 

Anatomy (Figs. 36a, female genitalia; 49b, 
penis) very similar to that of T. punicea; see 
diagnosis for differentiating characters. See 
Tables 21B-C for measurements. 

Type material: holotype (Fig. 33d-f) (SAM, 
D.I 7923, stn 036); and paratypes (SAM, 
D.2028, 5; 037, AMS, C.152912, many; 036, 
AMS, C.I 52913, many; 1003B, AMS, 
C.152915, many; 1003C, AMS, C.152916, 
many; 1003D, AMS, C.I 5291 7, many). 



66 



PONDER, HERSHLER & JENKINS 







1mm 



FIG. 37. Shells of species of Trochidrobia. 

a-c. Trochidrobia inflata, holotype. Freeling Springs (042). 

d-f. Trochidrobia minuta, holotype. Freeling Springs (046). 



Dimensions of holotype: length 1.31 mm, 
width 1.66 mm, length of aperture 0.78 mm. 

Localities: Middle Springs: Warburton 
Spring (681 A-C, 682), Beresford Spring 
(028). 

South Western Springs: Billa Kalina 
Springs (026-027, 723A-D, 758C, 759A; 
760, shells only; 761 B, 762A,B, 763A,B), 
Francis Swamp (717A-C, 720A-B, 721 A-C), 
Margaret Spring (722, shells only), Strang- 
ways Springs (029-030, 678A,B, 679A-C). 

Northern Springs: Bhnkley Springs (677), 
Hawker Springs (670A-C, 671, 672A-D, 
673), Fountain Spring (031-033), Twelve 
Mile Spring (035-037, 1003B-C) Outside 
Springs (038-040), Big Perry Spring (034) 
(Fig. 39). 

Remarks: Although this species is virtually 
identical to T. punicea in shell characters, it 
can be immediately recognised on dissection, 
the female genitalia being readily distin- 
guished from those of T. punicea in having a 
markedly narrower coiled oviduct and in the 
posterior part of the ventral channel being ex- 
panded and, in males, in having a prominent 
rectal arch. The ecology of this species ap- 



pears to be very similar to that of Г. punicea. 

Discriminate analysis, using only shell 
measurements, achieved some separation of 
T. punicea and T. smithi (Figs. 40, 41 ; Table 
9). A clear separation was achieved with fe- 
male genital measurements (Fig. 42; Table 
9). 

This species is named for Dr. B. J. Smith, 
formerly of the Museum of Victoria, Mel- 
bourne, as a small mark of appreciation of his 
contributions to the study of Australian non- 
marine molluscs. 

Trochidrobia minuta n.sp. 

Derivation: a reference to the small size of 
this species. 

(Figs. 37d-f, shell; 34e,g, protoconch; 35f, 
operculum; 38b, female genitalia) 

Diagnosis: Shell very small (up to about 
1.2 mm in diameter), very depressed (width/ 
height ratio 1.5-1.6), with 1.25-1.5 (mean 
1.47, males; 1.43, females) weakly convex 
whorls and widely umbilicate. Protoconch 
sculptured with irregular wrinkles and pits not 
arranged spirally (Fig. 34e,g). Colour yellow- 
ish white to pale brown. Head-foot darkly pig- 



AUSTRALIAN SPRING HYDROBIIDS 



67 




FIG. 38. Female genitalia of species of Trochidrobia. 

a. Trochidrobia inflata, Freeling Springs. 

b. Trochidrobia minuta, Freeling Springs. 

ag, albumen gland; be bursa copulatrix; eg, capsule gland; со, coiled part of oviduct; go, oviduct opening; int, 
intestine; mcp, posterior limit of palliai cavity; r, rectum; vc, ventral channel; vcp, posterior extension of 
ventral channel. 
Scale: 0. 1mm 



mented. Female genitalia v\/ith bursa 
copulatrix placed largely behind albumen 
gland (in other species it lies alongside albu- 
men gland). Coiled oviduct narrow, short, and 
ventral channel simple. 

Shell (Figs. 37d-f; 34e,g, protoconch), see 
diagnosis. See Table 21 A for measurement 
data. 



Operculum (Fig. 35f) as for genus. 

Radula as for genus. See Table 3 for data. 

Head-foot with darkly pigmented snout and 
grey triangular zone posterior to eyes. Very 
narrow unpigmented zone around eyes. 
Cephalic tentacles pale grey, unpigmented 
distally, without median line. 

Anatomy (Fig. 38b, female genitalia), as for 



68 



PONDER, HERSHLER & JENKINS 



genus. See diagnosis for differentiating char- 
acters. See Table 21B-C for measurement 
data. 

Type material: holotype (Fig. 37d-f) (SAM, 
D. 17924, stn 046); and paratypes (045, AMS, 
C.152918, many; 664A1, AMS, C.152919, 2; 
664A2, AMS, С 152920, 29; 046, AMS, 
C.1 52921, many). 

Dimensions of holotype: length 0.72 mm, 
width 1.11 mm, length of aperture 0.50 mm. 

Localities (Fig. 39): Northern Springs: 
Fountain Spring (031-032, 1002), Big Perry 
Springs (034,1001), Outside Springs (1006), 
Twelve Mile Spring (1003). 

Freeling Springs (043, 045, 046, 663, 664), 
unnamed spring north of Freeling Springs 
(666). 

Remarks: This minute species is very dis- 
tinctive and is readily separable on shell char- 
acters from T. punicea and T. smithi, although 
small individuals of those species approach it 
in size. Apart from most shell dimensions, the 
shell ratios PD/SH and SW/SH are signifi- 
cantly different in populations of 7. minuta 
when compared with T. smithi and T. punicea. 
The flat spire and pale colour are particularly 
characteristic. Discriminate analysis (Figs. 
40, 41, shell; 42, female genital anatomy; Ta- 
ble 9) readily distinguished this species from 
congeners. 

This species is abundant in the upper and 
middle parts of the spring outflows at Freeling 
Springs, but appears to be less common in the 
Northern Springs. The occurrence of this spe- 
cies together with 7. smithi in some of the 
Northern Springs is of interest because the 
size difference between these species is not 
so marked as it is between all other sympatric 
congeners in the mound springs. It would be of 
interest to compare the interactions between 
these two species with those between 7 
minuta and 7 inflata, which show greater size 
differences. 



Trochidrobia inflata n.sp. 

Derivation: a reference to the inflated shell 
of this species. 

(Figs. 37a-c, shell; 34f, protoconch; 35c, 
operculum; 38a, female genitalia) 

Diagnosis: Shell up to 1 .72 mm in diameter, 
with rather high spire (width/height ratio about 
1), 1.38-2.13 (mean 1.94, males; 1.95, fe- 
males) convex whorls, and narrowly umbili- 
cate. Protoconch microsculpture (Fig. 34f) of 



spirally arranged pits and wrinkles. Colour 
brown. Female genitalia similar to those of 7 
smithi but lacking expansion of ventral chan- 
nel. 

Shell (Fig. 37a-c), see diagnosis. See Ta- 
ble 21 A for measurement data. 

Operculum (Fig. 35c) as for genus. 

Radula as for genus. See Table 3 for data. 

Head-foot darkly pigmented, dark grey to 
black, with rather narrow unpigmented zone 
around eyes and very narrow median unpig- 
mented line on cephalic tentacles in some 
specimens, sometimes margined with black 
lines. 

Anatomy (Fig. 38a, female genitalia) as for 
genus. See diagnosis for differentiating char- 
acters. See Table 21B-C for measure- 
ments. 

Type material: holotype (Fig. 37a-c) (SAM, 
D.I 7925, stn 042); and paratypes (042, AMS, 
C.I 52922, many; 043, AMS, C.I 52923, 
many; 044, AMS, С 152924, many; 663, 
AMS, C.I 52925, 4). 

Dimensions of holotype: length 1 .58 mm, 
width 1.61 mm, length of aperture 0.88 mm. 

Localities: Freeling Springs (042-046, 663, 
664B,C, 665A-C) (Fig. 39). 

Remarks: The small umbilicus and relatively 
high spire enable this species to be readily 
distinguished. It is particularly abundant in the 
lower parts of the spring outflows and is sym- 
patric with 7 minuta. These two species differ 
significantly in size and in the values of shell 
ratios PD/SH, SW/SH and AH/SH. 

Discrimination of all of the taxa of Tro- 
chidrobia was tested using discriminate ana- 
lysis of measurements of shell and female 
genitalia. With the shell measurements 76% 
of the measured individuals (n = 219) were 
correctly classified (combined sexes) (Figs. 
40, 41). With female genital measurements 
(Fig. 42) 88% of all measured individuals 
(n = 26) were correctly classified. The gener- 
alized (taxonomic) distances between the 
groups are given in Table 9. Using shell mea- 
surements the greatest distance score 
achieved with painwise comparisons between 
the species was 1 .4 (the comparison between 
7 minuta and 7 smithi; males 1.41, females 
1.46), the lowest 0.12 (between females of 7 
smithi and 7 punicea). With female genitalia 
the highest score (4.3) was achieved between 
7 minuta and 7 punicea, with the compari- 
son between 7 punicea and 7 smithi being 
2.64. The lowest score (0.74) was between 7 
smithi and 7 inflata. 



AUSTRALIAN SPRING HYDROBIIDS 



69 




70 
3 



PONDER, HERSHLER & JENKINS 




FIG. 40. Plot of group centroids, using first two canonical axes, obtained from discriminate analysis of 
populations of species of Trochidrobia using shell measurements. Males and females of eacfi population are, 
for the purpose of this analysis, treated as distinct populations. The axes contain the following percentages 
of the variance of the variables used: first (horizontal) axis: SH, 88.75%; SW, 59.20%; AH, 89.98%; AW, 
91 .18%, BW, 81.96%; TW, 0.56%; PD, 1 1.13%. Second (vertical) axis: SH, 2.72%; SW, 19.12%; AH, 2.21%; 
AW, 1.31%; BW, 9.23%; TW, 2.27%; PD, 75.38%. 
i, T. inflata: m, 7. minuta: p, 7. punicea; s, 7. smithi. 



SNK tests (5% level) using pooled data, 
confibined and separate sexes, for each vari- 



able used in the discriminate analyses gave 
the following results: 



AUSTRALIAN SPRING HYDROBIIDS 



71 




FIG. 41 . Plot of group centroids, using first and third canonical axes, obtained from discriminate analysis of 
populations of species of Trochidrobia using shell measurements. Males and females of each population are, 
for the purposes of this analysis, treated as distinct populations. The axes contain the following percentages 
of the variance of the variables used: first (horizontal) axis: SH, 88.75%; SW, 59.20%; AH, 89.98%; AW, 
91.18%, BW, 81.96%; TW, 0.56%; PD, 11.13%. Third (vertical) axis: SH, 0.11%; SW, 0.34%; AH, 5.02%; 
AW, 5.01%; BW, 0.88%; TW, 93.46%; PD 0.02%. 
i, T. inflata; m, T. minuta; p, T. punicea; s, T. smithi. 



Shell characters: 

SH — Combined and separate sexes: T. 
minuta significantly different from all other 
taxa, which form a single subgroup. There is 
no sexual dimorphism in this character. 

SW — Combined sexes: all means are sig- 
nificantly different. Separate sexes: three dis- 
crete subgroups are formed, T. minuta, T. in- 
flata + T. punicea male, and T. punicea 
female + T. smithi. T. punicea is the only 
species sexually dimorphic (females larger) in 
this character. 

AH — Combined and separated sexes: the 
only taxon significantly different from the oth- 
ers is T. minuta. There is no sexual dimor- 
phism in this character. 



AW — Combined sexes: T. minuta and 
T. smithi form two separate subgroups with 
an intermediate, separate group formed by 
the other two taxa. Separated sexes: dis- 
crete subsets are formed by T. minuta, T. 
punicea (male) + T. inflata (male), and T. 
punicea female + T. inflata female + T. 
smithi. Thus significant sexual dimorphism is 
apparent in T. punicea and T. inflata in this 
character. 

BW, TW — Combined and separate sexes: 
only T. minuta is separated as a distinct sub- 
group. There is no sexual dimorphism appar- 
ent in these characters. 

PD — Combined sexes: two separate sub- 
groups, T. minuta + T. punicea, and T. smithi 



72 



PONDER, HERSHLER & JENKINS 




FIG. 42. Plot of discriminate scores for individuals, using first two canonical axes, obtained from discriminate 
analysis of specimens of Trochidrobia using female genital measurements. The axes contain the following 
percentages of the variance of the variables used: first (horizontal) axis; GO, 51.32%; CG, 61.65%; AG, 
93.08%; ВС, 52.76%; WB, 98.57%; DB, 90.61%; CV, 98.19%; DV, 95.13%. Second (vertical) axis: GO, 
39.60%; CG, 37.93%; AG, 3.61%; ВС, 43.64%; WB, 0.15%; DB,3.02%; CV, 1.25%; DV, 3.71%. 
i, T. inflata; m, T. minuta; p, T. punicea; s, T. smithi. 



AUSTRALIAN SPRING HYDROBIIDS 



73 



TABLE 9. Summary of results of discriminate analysis of species of Trochidrobia. The numbers are the 
Euclidean (taxonomic) distances between the groups. 



T. 
punicea 



Left side: Combined sexes, shell 
Female, genital 



T. 
smithi 



T. 
minuta 



T. 
inflata 



T. punicea 


X 


0.128 
0.271 


1.383 
1.149 


0.239 
0.219 


T. smithi 


0.155 
2.646 


X 


1.462 
1.414 


0.274 
0.385 


T. minuta 


1.306 
4.301 


1.437 
1.675 


X 


1.356 
1.142 


T. inflata 


0.243 
3.377 


0.324 
0.744 


1.248 
0.999 


X 



Right side: 
Female, shell 
Male, shell 



and T. inflata. Separate sexes: these two 
groups are not discriminated, all means falling 
into overlapping subsets. 

Female genital characters: 

GO — T. minuta separated from the rest of 
the species. 

CG, AG — no distinct subgroups. 

ВС, WB, DB, DV— Г. punicea separated 
from the other species. 

CV — T. minuta and T. inflata form a sub- 
group and T. smittii and T. punicea both sig- 
nificantly different. 

Anatomy 

Anatomical description of Fonscocfilea ac- 
cepta: Head foot (Fig. 11d). The distally bi- 
lobed snout is slightly shorter than the narrow, 
parallel-sided tentacles. These tentacles 
move slowly up and down and are held at 
about 45° to the longitudinal axis of the snout. 
They are not ciliated dorsally and weakly cil- 
iated ventrally, the cilia beating backwards at 
right angles to the longitudinal axis of the ten- 
tacle. The tentacles have blunt, rounded ends 
and the conspicuous, black eyes are in 
bulges at their outer bases. The entire dorsal 
side of the snout and most of the head are 
black or grey, and the tentacles are usually 
grey with a narrow, pale, longitudinal mid- 
dorsal stripe. The eyes are surrounded by a 
rim of unpigmented epithelium and immedi- 
ately behind them is a triangular zone of black 
pigment. The inner sides of the proximal ends 
of the tentacles have scattered, minute, 
opaque white spots, and poorly developed 
subepithelial pigment gives this area a slight 
reddish-brown tinge. There is an unpig- 



mented or weakly pigmented, ciliated, narrow 
rejection tract running down each side of the 
head-foot, at the junction of the foot and the 
"neck", to the sides of the foot. The tract on 
the right is more strongly developed in fe- 
males than in males. Metapodial and palliai 
tentacles are absent. The mantle collar has 
numerous black and a few white subepithelial 
pigment cells giving it a greyish appearance. 
The head-foot, by way of contrast, is pig- 
mented by epithelial cells. 

The foot is slightly expanded anteriorly, 
rather short (about two-thirds the shell length), 
about two and one-fourth times as long as it is 
wide, and has a prominent slit along the an- 
terior edge. The anterior mucous gland opens 
by way of this slit and can be seen dorsally 
through the unpigmented propodium. It is 
roughly triangular and composed of about 18 
simple tubules that lie along the longitudinal 
axis of the foot. There is a slight lateral con- 
striction in the anterior third of the foot and it is 
rounded behind. The foot is pale grey to dark 
grey along the sides and posteriorly but the 
anterior end is unpigmented mid-dorsally. The 
sole is pale grey, this colour being imparted by 
scattered black pigment cells in the connective 
tissue in the pedal haemocoel. Subepithelial 
gland cells make up the sole gland. The sole 
is ciliated, the cilia beating in a posterior di- 
rection. Cilial currents around the edges of the 
foot pass particles posteriorly. 

f[/lantle cavity (Fig. 4F). The mantle cavity is 
longer than broad and contains a well-devel- 
oped ctenidium (CT) with triangular filaments 
(see Table 18B) for statistical details), which 
extends through almost the entire length of the 
mantle cavity and occupies about half of the 



74 



PONDER, HERSHLER & JENKINS 



ac 





FIG. 43. a. Stomach of Fonscochlea accepta form A, Welcome Springs, viewed from its inner (left) side, 
b. Circum-oesophageai ganglia of F. accepta form A, Welcome Springs viewed dorsally (pedal ganglia 
omitted), ac, anterior chamber of stomach; cgl, left cerebral ganglion; dgo, digestive gland opening; int, 
intestine; Ip, left pleural ganglion; os, oesophagus; pc, posterior chamber of stomach; ss, style sac; rp, right 
pleural ganglion; sbo, suboesophageal ganglion; spo, supra-oesophageal ganglion. 
Scale: 0.25mm. 



pallia! roof in the posterior section, but narrows 
considerably anteriorly. An oval, unpigmented 
osphradium (OS) lies to the left of the posterior 
end of the ctenidiunn. It is about one-third the 
length of the ctenidium and consists of a 
raised, unciliated central portion containing 
the osphradial ganglion bordered by a slightly 
lower, weakly ciliated region with longer epi- 
thelial cells. Part of this border is separated 
from the central area by a narrow groove, 
forming a weak encircling ndge. A very poorly 
developed hypobranchial gland lies over the 
posterior end of the rectum. The mantle collar 
is ciliated, the cilia driving particles outwards. 

Alimentary system. A small pair of jaws 
composed of chitinous rodlets lies in the an- 
terior end of the buccal tube. The buccal 
mass occupies the length of the snout and the 
radular sac protrudes behind it. The free por- 
tion of this sac is about twice as long as the 
buccal mass. Two simple, tubular salivary 
glands open to the buccal cavity and lie dorsal 
to the nerve ring. The oesophagus is simple, 
narrow and the anterior part (mid-oesopha- 
gus) contains long dorsal folds that coil in a 
dorsal direction. The dorsal folds are lined 
with low ciliated cells but the lateral walls are 
predominantly lined with dark-blue-staining 
short cells which appear to be glandular. 

The stomach (Figs. 43a, 44b; see also Fig. 



45, stomach of F. zeidleri) is typical of the 
family in having a style sac (ss), an anterior 
(ac) and a posterior chamber (pc), and a sin- 
gle, posterior, slit-like digestive gland opening 
(dgo). There is no caeca! appendage. The 
style sac occupies about 0.6 of the stomach 
length and contains a crystalline style; the in- 
testine (int) opens to it along about two-thirds 
of its length. Externally the anterior and pos- 
terior chambers are distinguishable only on 
the inner (ventral) side and the oesophagus 
(os) and digestive gland open on this side. 
Internally the major typhlosole (t1) runs to the 
posterior end of the stomach and is subdi- 
vided into two low, strongly ciliated ridges 
(t1a,t1b). The minor typhlosole (t2) is also 
subdivided by a deep groove and terminates 
immediately in front of the gastric shield (gs). 
Posterior to the gastric shield the posterior 
chamber is finely transversely ridged on both 
floor and roof and functions as a sorting area 
(sa). These narrow, ciliated ridges are in 
marked contrast to the broad, low ridges (cr), 
separated by narrow grooves, that cross the 
roof of the anterior two-thirds of the stomach. 
These ridges are cuticulahzed, presumably to 
protect the epithelium from the rotation of the 
crystalline style. This ridged area is incor- 
rectly referred to as the sorting area by Davis 
et. a/ (1982). Fig. 44b illustrates the major fea- 



AUSTRALIAN SPRING HYDROBIIDS 



75 




FIG. 44. Stomachs of Trochidrobia punicea (a) and F. accepta form A (b) opened from outer (right) sides. 
Arrows in b indicate directions of main ciliary currents; letters A-E correspond approximately to sections with 
same letters in Fig. 45. cr, chitin-lined ridges; dgo, digestive gland opening; gs, gastric shield; ing, intestinal 
groove; int, intestine; oso, oesophageal opening; sa, sorting area; ss, style sac; t1, major typhlosole; t1a, t1b, 
folds developed from major typhlosole; t2, minor typhlosole; t2a, fold developed from minor typhlosole. 
Scale: 0.25mm. 



tures of the stomach with the dorsal (outer) 
wall opened. The transverse sections of the 
stomach of F. zeidleri (Fig. 45) show the re- 
lationships of the typhlosoles to the rest of the 
stomach and the extent of the ciliated epithe- 
lium. 

The digestive gland opening (dgo) lies pos- 
terior to the oesophageal opening. The diges- 
tive gland overlies the posterior end of the 
inner wall of the stomach and occupies the 
remainder of the visceral coil. It is composed 
predominantly of digestive cells with smaller 
excretory cells, which contain occasional ex- 
cretory granules, in the creases of the tu- 
bules. 

The intestine passes around the style sac, 
loops towards the anterior chamber of the 
stomach alongside the style sac, and then 
runs more or less straight to the right side of 
the mantle cavity. The rectum (Fig. 4F,R) 
passes along the right side of the mantle cav- 
ity and opens a little behind the mantle edge. 
The proximal part of the intestine contains a 
large typhlosole but the remainder is simple. 

Renal organ and pericardium. The renal or- 
gan lies behind the posterior wall of the man- 
tle cavity on the right side and opens to it by 
way of a short, dorsoventrally orientated slit 



(Fig. 4F,R0). This slit is located in the middle 
of the posterior palliai wall and is rendered 
conspicuous by white lips that surround it. 
The opening is lined with a ciliated, columnar 
epithelium and is surrounded by muscle fibres 
that presumably act as a sphincter. The renal 
epithelium is thin and simple, composed for 
the most part of a single layer of irregular 
cells. A nephridial gland occupies most of the 
outer wall. 

The pericardium also lies immediately be- 
hind the posterior palliai wall, but on the left 
side. It contains the heart, which consists of a 
well-developed ventrical and auricle. No reno- 
pehcardial opening was observed. 

Nervous system (Fig. 43b). The nerve ring 
is embedded in a mass of spongy connective 
tissue composed partly of cells containing 
black pigment granules. The arrangement of 
the ganglia is essentially similar to that de- 
scribed for Hydrobia truncata (Vanatta) by 
Hershler and Davis (1980). The cerebral gan- 
glia (cgl) are joined by a commissure about as 
long as the width of a single cerebral gan- 
glion. Each cerebral ganglion gives off seven 
nerves anteriorly, the base of one of them, the 
tentacular nerve, being swollen. There is a 
long right pleuro-supra-oesophageal connec- 



76 



PONDER, HERSHLER & JENKINS 





int b 




dgo 



FIG. 45. Sections through stomach of Fonscochlea zeidleri brm A. Approximate positions indicated in Fig. 
44b. cr, chitin-lined ridges; cs, crystalline style; dg, digestive gland; dgo, digestive gland opening; gs, gastric 
shield; int, intestine; oes, oesophagus; ss, style sac; t1, major typhlosole; t1a, tib, folds developed from 
major typhlosole; t2, minor typhlosole. 
Scale: 0.25mm. 



tive (rp-spo) and the left pleural (Ip) and sub- 
oesophageal ganglia (sbo) are fused. 

The cerebro-pedal complex is also very 
sinnilar to that described for H. truncata except 
that the cerebropedal connectives are rela- 
tively shorter than the pleuropedal connec- 
tives. Only the cerebral, pedal and buccal 
ganglia are pigmented. 

Male genital system (Fig. 46a). The testis 
occupies the upper surface of most of the vis- 
ceral coil behind the stomach. It is complexly 
lobed, with five lobes each containing approx- 
imately 15 to 20 lobules. The visceral section 
of the vas deferens forms a seminal vesicle 
that lies coiled beneath the anterior half to 
two-thirds of the testis. When straightened the 
seminal vesicle is about one and two-thirds 
times longer than the shell. A more or less 
straight part of the seminal vesicle emerges 
from beneath the testis and runs across the 
ventral side of the stomach. This duct narrows 
before entering the prostate gland immedi- 



ately behind the posterior palliai wall. This 
large gland extends partly (0.1 to 0.45 of its 
total length) into the right side of the mantle 
cavity. The prostate has thickly glandular 
walls except in its mid-ventral portion where 
the vas deferens opens and leaves. The pal- 
liai portion of the vas deferens opens imme- 
diately in front of the posterior palliai wall and 
runs as a straight tube along the right side of 
the mantle cavity until it is close to the base of 
the penis. Here it undulates for a short dis- 
tance before entering the penis. The palliai 
vas deferens lies just beneath the surface of 
the epithelium, has a simple, ciliated epithe- 
lium and is not surrounded by muscle fibers. 
The penis (Fig. 46a), coiled twice anticlock- 
wise as seen from above, is attached to the 
midline behind the head. The distance of the 
anterior edge of the penial attachment behind 
the eyes is only slightly less than the distance 
between the tentacle bases and about two- 
thirds the length of the snout. The penial duct 



AUSTRALIAN SPRING HYDROBIIDS 



77 





FIG. 46. a. Dorsal view of penis of Fonscochlea accepta form A, Welcome Springs, preserved material, 
b. Ventral view of living penis of Fonscochlea zeidleri form A, Blanche Cup. 
e, eye; p, penis; vd, palliai vas deferens. 
Scale: 0.25mm. 



lies close to the outer edge of the penis and is 
similar to the palliai vas deferens in structure. 
It coils in the broad, proximal quarter of the 
penis and is straight in the remainder. The 
distal part of the penis is long and tapers to a 
point. Unlike the basal part it is not trans- 
versely ridged and has longitudinal stripes 
that correspond to strands of longitudinal 
muscle lying beneath the epithelium. There 
are no penial glands or cilia; the epithelium is 
covered with cuticle. 

Female genital system (Figs. 12d,g,h, F. 
accepta form A; 47, F. zeidleri). The ovary is 
a simple sac filled with about 17 eggs in a 
mature individual. It is about one-half the 
length of the digestive gland and lies behind 
the posterior end of the stomach. The thin- 
walled oviduct is lined with pale-staining, un- 
ciliated cells and passes straight across the 
ventral wall of the stomach to a position just 
behind the posterior palliai wall. At this point 
there is a sudden change to a ciliated cuboi- 
dal epithelium that is thrown into longitudinal 
folds marking the commencement of the 
coiled section of the oviduct. 

The longitudinal folds in the first part of the 



coiled oviduct persist for only a shorl dis- 
tance, the lumen becoming oval. The initial 
section of the coiled oviduct probably repre- 
sents the renal section of the oviduct. It 
passes very close to the renal organ but no 
open reno-gonadial duct was observed in 
sections or in dissection. There are, however, 
strands of tissue connecting the most proxi- 
mal portion of the duct to the kidney wall and 
some modification of the kidney tissue was 
apparent in this region. The cells increase in 
size in the section following the renal part and 
they are more or less cuboidal with a few 
blue-staining (in fVlallory's Triple Stain) gland 
cells apparent. The coiled part of the oviduct 
(со), at this point, is surrounded by a few mus- 
cle fibres. It bends sharply upwards and then 
loops down to run forward along the albumen 
gland (ag). Near the posterior end of the al- 
bumen gland it loops upwards and two spher- 
ical sperm pouches open to it. In this region 
the coiled oviduct is surrounded by an outer 
coat of circular muscle. The epithelium is 
thrown into a few low, longitudinal folds and 
sperm are attached to the ciliated epithelial 
cells. The oviduct increases in diameter and 



78 



PONDER, HERSHLER & JENKINS 

mcp> 




CO 



FIG. 47. Female genitalia of Fonscochlea zeidleri form A, from the left side. 

ag, albumen gland; be, bursa copulatrix; bed, duet of bursa eopulatrix; eg, capsule gland; eo, eoiled part of 
oviduet; go, female genital opening; mop, posterior limit of mantle cavity; sr, seminal receptacle; srd, duct of 
seminal receptacle; uo, upper oviduct; ve, ventral channel; vcp, posterior extension of ventral channel. 
Scale: 0.2mm. 



loops upwards to lie behind, and sometimes 
above, the proximal loop. It then opens ven- 
trally into the posterior end of the capsule 
gland (eg). This tubular extension (vpc) of the 
sperm groove in the ventral channel is lined 
with an epithelium similar to that of the sperm 
groove, the cuboidal cells beahng conspicu- 
ous cilia and occasional blue-staining gland 
cells. 

The two sperm pouches (be, sr) lie near the 
posterior end of the albumen gland on the 
Inner (left) side of the gland and their short 
ducts extend from their ventral walls to open 
separately into the oviduct. They are identical 
in histology and appearance and might both 



be homologous with the bursa copulatrix of 
other hydrobiids. They are lined with long, 
purple-staining cells with dense, finely- 
staining contents and basal nuclei. Unori- 
ented sperm fill the lumen in most specimens 
and additional sperm have their heads at- 
tached to the outer surface of the epithelial 
cells. Each sperm sac is surrounded by a coat 
of muscle and their ducts, which also contain 
sperm with their heads attached to the epithe- 
lial cells, are similar in structure to the oviduct 
in this region. 

The oviduct gland of F. zeidleri (Fig. 47) is 
typical of those in all the species of Fonscoch- 
lea. It consists of a blue-staining albumen 



AUSTRALIAN SPRING HYDROBIIDS 



79 



gland (ag), which lies behind the posterior 
palliai wall, and a red-purple staining capsule 
gland (eg), which lies in front of this wall. The 
two glands are, however, externally continu- 
ous. The lumen of the albumen gland is con- 
tinuous with that of the capsule gland and cil- 
iated cells line the lumina of both. The tubular 
oviduct opens to the thin-walled ventral chan- 
nel (vc) of the capsule gland, part of which, on 
the left, is separated from the main channel 
by a ciliated, nonglandular fold. This fold con- 
tinues throughout the ventral channel to the 
small, subterminal, ventral opening and sep- 
arates the sperm-conducting channel, on the 
left, from the egg-conducting channel. The 
very thin ventral wall of the egg-conducting 
channel is lined with small, cuboidal, uncili- 
ated cells. In the vicinity of the oviduct open- 
ing the gland cells in the ventral part of the 
capsule gland change from red- to pale-blue- 
staining. 

The anatomy of Fonscochlea accepta is 
typical of all the species of Fonscochlea. The 
most important character that separates this 
genus from all other genera in the family is the 
equal-sized sperm sacs that seem to have 
been developed from a subdivided bursa cop- 
ulatrix. Their arrangement differs in detail in 
the two subgenera of Fonscochlea, as de- 
scribed in the taxonomic section (compare 
Figs. 12c-h and 27a-d with Figs. 12a,b and 
47). In most other respects the anatomy of 
species of Fonscochlea is similar to that of 
other species of the family Hydrobiidae. 

Anatomical description of Trochidrobia pu- 
nicea: Head-foot (Fig. 24h). The snout is 
about two-thirds the length of the tentacles 
when at rest but when extended is about the 
same length. It has a bilobed tip that is slightly 
narrower than the rest of the snout, and is 
pigmented dark grey to black, the tip being 
unpigmented in many specimens. The ceph- 
alic tentacles are parallel-sided, held at about 
45°, sway slowly up and down through a small 
arc (species of Fonscochlea move their ten- 
tacles through a greater arc and more rapidly) 
and are pigmented light to dark grey, often 
with a narrow, white median line. A few scat- 
tered, dense-white spots lie on the inside 
proximal end of the tentacles anterior to the 
eyes and a conspicuous group of these spots 
lies on the inner side of the eyes and, some- 
times, behind them. The large, black eyes are 
in bulges at the outer bases of the tentacles 
and are, in some specimens, surrounded by 
black pigment, but in others the black pigment 



lies mainly behind the eyes. The dorsal head 
and 'neck' are grey to black and a ciliated 
rejection tract runs down both the sides of the 
head onto the foot. 

The foot is almost as long as the shell is 
wide and is about one-third as wide as long. 
Only a very short portion extends beyond the 
operculum and, normally, the foot is invisible 
when the crawling animal is viewed from 
above. There are lateral constrictions behind 
the anterior edge, and the posterior end is 
evenly rounded. A well-developed pedal 
gland opens to the anterior edge of the foot 
and the sole is supplied with subepithelial 
glands. The entire sole and the lateral edges 
of the foot are covered with posteriorly beat- 
ing cilia. The anterior edge has cilia beating 
towards the outer corners. The foot is pig- 
mented grey to black on the anterior and pos- 
terior dorsal surfaces and is paler to unpig- 
mented dorsolaterally. The sole is dark grey 
to whitish, the colour being imparted by pig- 
ment-bearing cells in the connective tissue in 
the cephalic haemocoel. 

The mantle collar is richly supplied with 
dense-white cells across the outer lip but 
these are fewer across the inner lip where 
there is more black pigment. This black pig- 
ment is predominantly in subepithelial cells, 
but, with the exception of the sole, the remain- 
der of the pigment on the head-foot is con- 
tained in epithelial cells. 

Mantle cavity (Fig. 48). The mantle cavity is 
short and broad, being slightly wider than it is 
long. The well-developed ctenidium (ct) is 
placed diagonally across the cavity and the 
apices of the filaments lie at their right edge. A 
short, oval osphradium (os) lies alongside the 
posterior end of the ctenidium. It is similar in 
structure to that of species of Fonscochlea. 
There is no hypobranchial gland. The rectum 
(r) and genital duct (eg) run down the right 
side of the cavity and the anus (a) lies close to 
the mantle edge. The ctenidium lies closer to 
the mantle edge, ending just inside the mantle 
skirt (me). 

Alimentary system. The mouth opens to a 
short, cuticle-lined oral tube with a pair of 
small jaws laterodorsally. The well-developed 
buccal mass occupies most of the snout and 
a coiled radular sac emerges posteriorly from 
it. 

The anterior part of the oesophagus (mid- 
oesophagus) has long dorsal folds, which are 
curved dorsally, occupying most of the lumen. 
The lateral and ventral walls are lined with 
a ciliated, cuboidal epithelium with purple- 



80 



PONDER, HERSHLER & JENKINS 

'^ ,mcp 




me 



FIG. 48. Dissection of palliai cavity of Trochidrobia punicea. Double-headed arrow indicates separation of 
kidney from dorsal palliai wall, a, anus; eg, capsule gland; ct, ctenidium; go, female genital opening; mcp, 
posterior limit of mantle cavity; me, mantle edge; os, osphradium; r, rectum; rg, renal organ; ro, renal 
opening. 
Scale: 0.2mm. 



Staining, granular contents. This section of 
the oesophagus terminates at the end of the 
cephalic cavity, the posterior oesophagus be- 
ing narrower and without the dorsal folds. The 
simple, tubular salivary glands lie dorsal to 
the nerve ring. 

The stomach (Fig. 44a) is similar in general 
appearance externally to that of species of 
Fonscochlea. The style sac communicates 
with the intestine along all of its length. The 
well-developed typhlosoles (t1 , t2) within the 
stomach are unpigmented and readily dis- 
cernible against the stomach wall. The major 
typhlosole (t1) extends to the posterior end of 
the stomach where it swings around the di- 
gestive gland opening after fusing with the mi- 
nor typhlosole (t2). Short left (t2a) and right 
branches of the fused minor + major typhlo- 
sole are given off that extend onto the roof of 
the posterior end of the stomach. The gastric 
shield lies close to the oesophageal (oso) and 
digestive gland (dgo) openings. These open- 
ings lie at either end of a groove that divides 



the major typhlosole (t1 ) into two arms. This 
typhlosole splits into two folds at the anterior 
end of the oesophageal opening, the right fold 
running to the anterior edge of the gastric 
shield and the left fusing with the minor 
typhlosole near the digestive gland opening at 
the posterior end of the stomach. 

The typhlosoles, style sac, and the poste- 
rior end of the stomach are ciliated, the re- 
mainder of the gastric epithelium being cutic- 
ularized. The pigmented roof of the anterior 
chamber is very indistinctly marked with 
widely separated narrow grooves (cr). 

The digestive gland and intestine are very 
similar to those of Fonscochlea. The digestive 
gland covers the inner, ventral, side of the 
stomach to half-way across the anterior 
chamber. 

Renal organ and pericardium. The renal or- 
gan (kidney) lies behind the posterior wall of 
the palliai cavity. The lumen is severely re- 
duced, particularly in females, by the invagi- 
nation of the genitalia. The renal opening (Fig. 



AUSTRALIAN SPRING HYDROBIIDS 



81 



О 




è b 



FIG. 49. Penes of Trochidrobia punicea, Blanche Cup (a.) and Trochidrobia smithi, Outside Springs (b. 
a, eye; p, penis; vd, palliai vas deferens. 
Scale: 0.1mm 



48, ro) lies in the middle of the posterior wall 
of the palliai cavity. It is a short, vertical slit 
surrounded by thickened, ciliated, white lips 
(sphincter muscle). The renal epithelium is 
simple and very thin except on the outer wall 
where it forms a thick nephridial gland. 

The pericardium lies behind the left side of 
the posterior palliai wall and the base of the 
ctenidium. Its posterior face abuts against the 
anterior end of the style sac. The ventricle 
and auricle are both well developed. 

Nervous system. The cerebral ganglia are 
joined by a commissure that is slightly shorter 
than the width of the cerebral ganglia. The 
pleural ganglia are fused to the cerebral gan- 
glia but a waist-like constriction separates 
them. The supra-oesophageal ganglion is a 
little longer than the width of the cerebral gan- 
glia, and the right pleuro-oesophageal con- 
nective is about the same length as the cere- 
bral ganglia. The suboesophageal ganglion 
abuts the left pleural ganglion. All of these 
ganglia, and the buccal and pedal ganglia, 
are pigmented except for the supra-oesoph- 
ageal ganglion. 

Male genital system. The testis consists of 
several lobes, each consisting of numerous 
lobules, about 45 in the anterior lobe. The vas 
deferens lies coiled beneath the anterior two 
lobes of the testis. It runs forward as an al- 



most straight tube, narrows across the inner 
(ventral) side of the stomach and terminates 
just behind the posterior palliai wall where it 
opens to the middle part of the prostate gland. 
The palliai section of the vas deferens leaves 
the prostate gland immediately in front of the 
posterior palliai wall and runs along the 
groove at the junction of the mantle cavity 
floor and the mantle roof. It is straight until it 
nears the base of the penis where it undulates 
across the right side of the "neck" before en- 
tering the penis. 

The large prostate gland is reniform, 
narrowly oval in section, thickly glandular, 
with a thin ventral wall only in the vicinity of 
the point of entry and departure of the vas 
deferens. It lies partly in the palliai roof and 
partly behind the posterior palliai wall. Its 
extent of penetration of the palliai roof varies 
from one-third to one-half of its total length. 
The penis (Fig. 49a) lies just to the right side 
of the midline of the head at a distance 
behind the eyes about equal to the length of 
the snout. It is coiled twice anticlockwise in 
preserved material. The base of the penis is 
swollen and unpigmented, at least in the 
proximal part, and has clearly defined 
creases running across its surface. The 
remainder of the organ is pale to dark grey 
along much of its length, the proximal part 



82 



PONDER, HERSHLER & JENKINS 



often being unpigmented. It is smooth and 
tapers to a point. The inner side of the penis, 
i.e. the edge on the inner side of the coil, is 
flattened to almost channelled in some 
individuals and rounded in others. The penial 
duct is, like the palliai vas deferens, ciliated, 
thin-walled and very narrow. The penis is 
surrounded by an unciliated, non-cuticular- 
ized cuboidal epithelium. It contains some 
pale-blue staining subepithelial gland cells 
amongst the muscle and connective tissue. 
Distinct penial glands are absent. 

Female genital system (Fig. 36b). The 
ovary is short relative to the digestive gland. 
The upper oviduct (uo) is a straight, thin- 
walled tube that passes across the ventral 
surface of the stomach before reaching a 
point close to the pericardium and the renal 
organ. Here its walls thicken and the ciliated 
epithelium is raised into longitudinal hdges. 
There is no gonopericardial or renogonadial 
duct although a tissue connection (st) with the 
pericardium can be seen in dissection. The 
renal section of the oviduct is extremely short 
and is invaginated within the renal wall. 

The coiled oviduct (со), the first, very short 
part of which is the renal oviduct, is coiled 
behind the posterior palliai wall (mcp) on the 
left side of the albumen gland. It is consider- 
ably swollen in this species, a character not 
seen in other species of the genus. It invagi- 
nates into the renal organ, considerably re- 
ducing the volume of the renal lumen. The 
outer wall of the coiled oviduct is surrounded 
by an outer layer of circular muscle fibres and 
a thicker inner layer of longitudinal fibres and 
is lined with a ciliated cuboidal epithelium. 
Spermatozoa are stored in the lumen of the 
coiled oviduct, and are aligned more or less 
longitudinally, apparently by ciliary action. 
The large bursa copulatrix lies behind the 
coiled oviduct and its right (outer) wall is em- 
bedded in the albumen gland. There is a 
short, free bursal duct (about one-fifth the 
length of the bursa), the remainder of the duct 
merging with the coiled oviduct and running 
back along it. The bursal duct eventually 
opens to the coiled oviduct but the exact point 
of opening was not established because the 
two tubes are enveloped in a common sheath 
of connective tissue. The bursa copulatrix 
(be) is lined with an unciliated, purple-staining 
columnar epithelium with granular cytoplasm 
and basal nuclei. In all specimens sectioned, 
the bursa did not contain sperm. 

The oviduct anterior to the bursal duct con- 
tinues as a short, broad tube, for a distance 



TABLE 10. Shell heights for the snail taxa used in 
the physiology experiments. 

Range of 

shell heights 

(means, sexes 

pooled) among 

Species populations (mm) 



F. accepta form A 
F. accepta form В 
F. aquatica 
F. variabilis 
F. cónica 
F. zeidleri 
T. punicea 



3.16-3.57 
2.83-3.41 
3.93-4.50 
1.41-2.52 
1.70-2.18 
2.97-4.37 
1.60-1.91 

(Shell width) 



approximately equal to the length of the 
bursa, to the posterior wall of the palliai cavity 
where it opens to the capsule gland (eg) as 
the ventral channel. This oviducal tube is 
lined with ciliated cells, amongst which are 
scattered larger, blue-staining gland cells. 

The oviduct gland is clearly divided into a 
blue-staining albumen gland (ag) lying behind 
the palliai cavity and, continuous with it, a red- 
staining capsule gland (eg) in front of the pos- 
terior palliai wall. The albumen gland opens to 
the capsule gland which, immediately in front 
of the junction of the two glands, receives the 
oviduct. This tube opens to the ventral chan- 
nel (vc) of the capsule gland into a ciliated 
gutter, similar to that in species of Fonscoch- 
lea, which runs to a slit-like opening (go) sit- 
uated about one-third of the length of the cap- 
sule gland from its anterior end. The capsule 
gland is, however, relatively shorter and 
broader than that of species of Fonscochlea. 
In the vicinity of the genital opening the glan- 
dular epithelium in the ventral part of the cap- 
sule gland forms a pale-blue zone. 

The main feature of interest in the anatomy 
of this genus is the lack of a seminal recep- 
tacle and the development of accessory 
sperm storage in the coiled oviduct. In Troch- 
idrobia smithi sperm storage takes place in 
the ventral cliannel. In other respects the 
anatomy is typical of the family Hydrobiidae. 

Physiology 

The taxa examined fall into four main 
groups, distinguished by differences in shell 
size (Table 10) and habits: 1) F. zeidleri iorm 
A, the amphibious species; 2) large aquatic 
species (F. aquatica form A and cf. form A, 
and F. accepta); 3) small, aquatic Fonscoch- 



AUSTRALIAN SPRING HYDROBIIDS 



83 



TABLE 11. Survivorship of snails in dry dishes. Ten snails were used in each experiment. 
T1 = Trochidrobia punicea, Fl = Fonscochlea accepta, F2 = Fonscochlea aquatica, F3 = 
variabilis, F4 = Fonscochlea cónica. F5 = Fonscochlea zeidleri. ВС = Blanche Cup, 
Springs Railway Bore, FS = Finniss Springs. 



Fonscochlea 
CS = Coward 













Species (population) 










Number 
of hours 


T1 
(run 1) 


T1 
(run 2) 


F4 


F3 
(BO) 


F3 

(CS) 


Fl 


Fib 


F2 


F5 
(FS) 


F5 
(ВС) 


F5 

(CS) 


1 


8 


5 


2 


7 


8 


10 


10 


9 


10 


10 


10 


2 


5 


2 





4 


5 


7 


10 


9 


10 


8 


10 


4 


2 








1 


4 


6 


10 


9 


9 


9 


10 


6 














3 


1 


8 


6 


10 


9 


10 


12 


0^ 

















5 


5 


9 


10 


7^ 


24 


0' 














2 


4 


3 


93 


9 


9 


48 


02 























83 


9 


6 


Date& 

time 

commenced 


27-8 
1 1 :30AM 


2-9 
8:15AM 


3-9 
10:15AM 


21-8 
9:00AM 


1-9 
8:00AM 


3-9 
9:25AM 


29-8 
8:00AM 


31-8 
9:30AM 


28-8 
8:00AM 


30-8 
8:40 AM 


1-9 
7:45AM 



'commenced at 6:30PM ^commenced at 3:34PM ^commenced at 8:40AM ''disti checked after 10 minutes, but not after one 
hour. 



lea species (F. cónica, F. variabilis form A); 
and 4) Trochidrobia punicea, small and 
aquatic. 

Desiccation. During the 48 hours that these 
experiments were run, there was no mortality 
in any of the wet, control dishes of any of the 
species. The results for the moist dishes were 
the same, except that at 48 hours the two 
populations of F. variabilis tested had 90% 
(Coward Springs Railway Bore) and 100% 
(Blanche Cup) mortality (results significantly 
different from those for the other species, 
Fisher's Exact Test, P < 0.005). The results 
are summarized in Table 1 1 . 

Only F. zeidleri survived well (60-90% in 
the three populations tested) after 48 hours in 
the dry dishes (Figs. 50-52). Fonscochlea 
aquatica cf. form A (Kewson Hill) had 10% 
survival after 48 hours (significantly less than 
that of any F zeidleri population, P < 0.05) 
and 50% mortality after only one hour. The 
other species had higher mortalities. F cón- 
ica had 80% mortality after one hour and 
100% mortality after two hours, T. punicea 
50-80% mortality after two hours and 100% 
mortality after 6 hours, F variabilis 50% mor- 
tality after two hours and 100% mortality after 
12 hours and F aquatica and F accepta 
100% mortality after 48 hours (see below for 
details). After 24 and 48 hours, all three F 
zeidleri populations had significantly higher 
survival than that of the next best "survivor", 
F accepta form В (for all pairwise compari- 



sons, P < 0.05). At 12 hours, only the 
Blanche Cup F zeidleri population had a sig- 
nificantly higher survival than that of F ac- 
cepta form В (P = 0.025). There was no sig- 
nificant difference in survival amongst the 
three F zeidleri populations at any time. 

Survival of two of the three large aquatic 
Fonscochlea taxa, F. accepta form A, F ac- 
cepta form В and F. aquatica, was good 
through 12 hours (50%) and higher than that 
of the small aquatic Fonscochlea species (F 
cónica and F variabilis): for F accepta form 
В, this difference was significant at two, four, 
six, 12, and 24 hours (for all pairwise compar- 
isons, P < 0.05), for F aquatica, the difference 
was significant at four and 12 hours (P < 
0.05). Both F accepte form В and F. aquatica 
had higher survival than the other large 
aquatic species, F accepta form A, after six 
and 12 hours (all pairwise comparisons, P < 
0.05), but not at 24 hours. At no point during 
the experiments did the former two taxa differ 
significantly from each other in survival. 

There were no significant differences in 
survival seen in any of the painA/ise compari- 
sons for any time between the two popula- 
tions of F variabilis and the two runs of T. 
punicea (Finniss Springs), except for that be- 
tween T. punicea and F variabilis from Cow- 
ard Springs Railway Bore after four hours 
(P = 0.05). Both of these species showed a 
fairly rapid onset of mortality. Yet after two 
hours both of the above species had a signif- 



84 



PONDER, HERSHLER & JENKINS 




> 

Ш 

30 û; 



0700 
2-9 



0700 
3-9 



0700 
4-9 



0815 
5-9 



0700 
6-9 



0830 
7-9 



TIME OF DAY 



FIG. 50. Running record of air temperature and humidity in tent for duration of experiments. Readings taken 
hourly, generally from 0600 to 2200; dashed lines indicate intervals at night during which readings were not 
taken. 



icantly higher survival than F. cónica (all com- 
parisons, P < 0.05), which already had 100% 
mortality at that time. 

Fonscochlea aquatica from Kewson Hill 
showed the peculiar pattern of fairly rapid on- 
set of mortality (50% after one and two 
hours), followed by survival of 10% of the 
snails after 12, 24, and 48 hours. 

Salinity: Nearly 100% of the snails, for all 
species, remained active in the control jars for 
the duration of the experiment. At 24 hours, 
98% of the snails (for all species pooled) were 
active. The results for salinities of 6, 9, and 
12%o are given in Table 12. 

In 6%o salt water, nearly 100% of the spec- 
imens of F. zeidleri, F. aquatica, and F. ac- 
cepta form В remained active throughout the 
experiment: after 24 hours, for these species 
pooled, 91% of the snails were active and 
there were no significant differences in activ- 
ity among these species. However, activity 
did decline markedly in F variabilis after two 
hours and in T. punicea after 24 hours. At 12 
and 24 hours, the number of active snails of 
F variabilis (either population) was signifi- 
cantly less (Fisher's Exact Test, P < 0.05) 
than that of the species listed above for all 



pairwise comparisons but one (F accepta 
form B-F variabilis, Coward Springs Railway 
Bore). For T. punicea, at 12 hours, the num- 
ber of active snails was significantly less than 
that of only F aquatica from Kewson Hill and 
F zeidleri and Coward Springs Railway Bore 
(P < 0.005). At 24 hours, the number was 
significantly lower than that seen in any of the 
above group of species (P < 0.005). A signif- 
icantly larger number of T. punicea were ac- 
tive than F variabilis (both populations) at 
two, three, six, and 12 hours (for all compar- 
isons, P < 0.025). 

In 12%o salt water, activity of F zeidleri and 
the large aquatic species remained high. After 
24 hours, for all species pooled, 80% of the 
snails were active and there were no signifi- 
cant differences among species. There were, 
however, several significant differences seen 
in the early hours of the experiment when ac- 
climitization was apparently occurring. There 
was no activity for both F variabilis and T. 
punicea for the duration of this experiment. At 
six, 12, and 24 hours, the number of active 
snails for these species (0) was significantly 
lower than that of the above group of species 
(for all comparisons, P < 0.025). 



AUSTRALIAN SPRING HYDROBIIDS 



85 




NO. OF HOURS 

FIG. 51 . Survivorship of large-sized species of Fons- 
cochlea in dry dishes. Fonscochlea accepta form A, 
solid circles; F. accepta form B, open triangles; 
Fonscochlea aquatica form A, open circles; F. zeid- 
leri form A represented by line with error bars, rep- 
resenting range of results among populations of 
that species. 



At 24%o salt water, with two exceptions (In 
which case a few snails were active at only 
one point in the experiment), activity was nil 
for all species throughout the experiment. 

Deoxygenated water. The results are given 
in Table 13. In the control tubes, initially sup- 
plied with oxygenated water, activity generally 
decreased markedly by six hours, and only 
26% of the snails (for all species pooled) were 
active by 20 hours. In four expehments, the 
snails in the control tubes were tested for sur- 
vival, in the same manner as were those 
snails in the tubes with deoxygenated water, 
after 20 hours; 90% of these snails (all spe- 
cies pooled, N = 80) were alive, although 
some were sluggish. The decrease in activity 
and occasional mortality could have been due 
to deoxygenation of the water in the small 
15 cc test tubes during the course of the ex- 
periment. 

In the test tubes initially supplied with deox- 
ygenated water, again activity decreased 
markedly during the course of the experi- 
ments, with only 26% of the snails (all species 
pooled) active at six hours, and 13% at 20 
hours. Despite this decrease in activity, sur- 



I 4 




12 4 6 12 

NO. OF HOURS 

FIG. 52. Survivorship of small-sized species of Fons- 
cochlea and Trochidrobia punicea in dry dishes. 
Trochidrobia punicea, solid squares (two runs 
pooled); Fonscochlea cónica, open circles; Fons- 
cochlea variabilis form A, solid circles (runs for dif- 
ferent populations of this species pooled). Error 
bars represent ranges of results. 



vival for all species, except T. punicea, was 
near 100% at all times. In general, the snails 
became active during the first ten minutes af- 
ter being placed in oxygenated water. Sur- 
vival of T. punicea was significantly less than 
that of all other species tested at four hours 
(all pairwise comparisons, Chi-Square Test of 
Independence, one-tailed, P < 0.005), six 
hours (P < 0.05) and 20 hours (P < 0.005). 
There were no significant differences in sur- 
vival between any two of the other species. 

Temperature. The results (Table 14) indi- 
cate that, in general, almost all individuals of 
all species tested remained active at 10- 
32°C., and a large percentage of individuals 
were active at 5° (76% for all species pooled), 
35° (77%) and 37° (41%). At the lower end of 
the temperature range, the snails generally 
became more and more sluggish, whereas at 
the upper range of the temperature range, ac- 
tivity greatly increased and, at a slightly 
higher temperature, was followed by slug- 
gishness and cessation of activity. 

Considerable variation was seen in the in- 
stances in which several populations of a spe- 



86 



PONDER, HERSHLER & JENKINS 



TABLE 12. Activity of snails over time in water of salinities of 6%o, 97oo, and 12%o. Ten snails were used 
in each experiment unless otherwise indicated. The approximate natural salinity of the water used in the 
experiments is given for each sample (calculated from the conductivity). 
FS = Finniss Springs, CSRB = Coward Springs Railway Bore, ВС = Blanche Cup. 



Species 
(population) 


Salinity 




6%o 

Number of hours 






97oo 
Number of hours 




1 


12%o 
Mumber of hours 




Date, time 
commenced 


1 


2 


3 


6 


12 


24 


1 


2 


3 


6 


12 


24 


1 


2 


3 


6 


12 


24 


F, zeidleri (FS) 


1.8 


9 


6 


10 


9 


6 


10 


4 


5 


8 


5 


9 


9 


2 


4 


9 


6 


9 


9 


29.8 


10:55AM 


F, zeidleri (CSRB) 


2 


10 


10 


10 


10 


10 


9 


10 


10 


10 


10 


10 


10 


2 


3 


7 


8 


9 


8 


1.9 


9:30AM 


F zeidleri (ВС) 


3.6 


10 


10 


10 


10 


9 


10 


10 


10 


10 


7 


8 


9 


10 


9 


8 


7 


6 


9 


30.8 


10:20AM 


F. aquatica 


3.6 


10 


10 


8 


10 


9 


10 


8 


8 


7 


7 


8 


9 


9 


6 


6 


5 


3 


7 


30.8 


12:52PM 


F. accepta form В 


1.8 


10 


10 


9 


7 


7 


9 


9 


9 


9 


7 


4 


10 


8 


2 


5 


5 


5 


9 


29.8 


9:10AM 


F. variabilis 

(всу 


3.6 


6 


9 


9 


9 


6^ 


9 





4 


1 


2 


0' 























31.8 


10:50AM 


F. variabilis 
(CRSB) 


2 


10 


9 


9 


7 


10 


7 


10 


2 








1 


2 




















2.9 


9:30AM 


T. punicea 


1.8 


10 


10 


9 


9 


10 


10 


8 


10 


8 


9 


6 


2 




















28.8 


8:20AM 



' 1 1 specimens used 



TABLE 13. Survivorship and activity of snails in deoxygenated water. Activity of snails in control tubes 
(initially supplied with oxygenated water) also shown. Twenty snails were used in each experiment 
unless otherwise indicated. 
FS = Finniss Springs, ВС = Blanche Cup, and CS = Coward Springs Railway Bore. 



% of snails surviving 



% of snails active 
in tube 



% of snails active 
In control tube 



Number of hours 



Species 
(population) 


1 


2 


4 


6 


20 


1 


2 


4 


6 


20 


1 


2 


4 


6 


20 


Date, Time 
Comm. 


F zeidleri (CS) 


100 


100 


100 


100 


100 


45 


40 


45 


75 





100 


100 


80 


95 


5 


1.9. 


1 1 :05AM 


F zeidleri (ВС) 


100 


100 


100^8 


100'^ 


100 


87 


15 


30 








100 


90 


90 


85 


80 


30.8. 


12:00PM 


F zeidleri (FS) 


100 


100 


100 


90 


100 


15 


10 


5 








100 


100 


80 


25 


10 


28.8. 


10:30 AM 


F. accepta 
form В (FS) 


100 


100 


ЮО'э 


100 


100 


95 


95 


30 


60 


10 


100 


100 


80 


75 


20 


29.8. 


10:55AM 


F. variabilis 
(10 snails/tube) 


90 


100 


100 


100 


100 


30 


30 


40 


10 





100 


100 


100 


45 





31.8. 


1 1 :25AM 


F. cónica 


100 


100 


100 


90 


100 


60 


90 


90 


30 





100 


100 


100 


90 


50 


3.9. 


12:30PM 


T. punicea 


95 


100 


50 


60 


35 


65 


60 


10 


10 





90 


60 


30 


25 


5 


2.9. 


3:00PM 



Мб specimens used '^18 specimens used ^^19 specimens used 



cíes were tested. For F. zeidleri, at 2° the 
Blanche Cup population had a significantly 
larger nunnber of individuals active than did 
the other two populations (P < 0.0005 Chi- 
Square); at 35° the Coward Springs Railway 
Bore population had significantly larger num- 
ber of active snails than did the other two (P < 
0.01 , Chi-Square); at 37°, the Coward Springs 
Railway Bore population had significantly 



higher activity than did the Blanche Cup pop- 
ulation (P = 0.025, Fisher's Exact Test), which 
in turn had higher activity than did that of Fin- 
niss Springs (P < 0.01, Chi-Square); at 40 
and 42° the Coward Springs Railway Bore 
population had a significantly higher activity 
than had the other two populations (P < 
0.0005, Chi-Square). While F. accepta form В 
had significantly higher activity than did F. ac- 



AUSTRALIAN SPRING HYDROBIIDS 87 

TABLE 14. Numbers of snails active at various water temperatures. Twenty snails were used for each 

experiment unless otherwise indicated. 

FS = Finniss Springs, ВС = Blanche Cup, CS = Coward Springs Railway Bore. 





















Temperature (°C) 


















Species 
(population) 





.12 


.25 


.50 


1 


2 


5 


10 


15 


20 


25 


30 


32 


35 


37 


40 


42 


45 


47 


50 


Date 
Comm. 


F. zeidleri (FS) 


- 





2 


3 


5 


7 


20 


20 


20 


20 


20 


20 


20 


7 


6 





— 


— 


— 


— 


2.9 


F. zeidleri (ВС) 





— 


3 


2 


3 


20 


20 


20 


20 


20 


20 


20 


18 


11 


15 


8 





— 


— 


— 


30.8 


F. zeidleri (CS) 


- 


— 


- 


— 





9 


15 


20 


20' 


20 


20 


20 


20 


19 


20 


20 


1 





— 


— 


1.9 


F. aquatica 





— 


- 


2 


4 


4 


17 


19 


20 


20 


20 


20 


18 


16 


18 





— 


— 


— 


— 


30.8 


F. accepta form В 


- 


- 


- 





9 


16 


20 


20 


20 


20 


20 


20 


20 


15 


11 


2 


4 


02 


— 


— 


1.9 


F. accepta form A 


— 


— 





2 


3 


5 


19 


20 


20 


20 


20 


20 


20 


20 


20 


10 


4 





— 


— 


3.9 


F. variabilis (ВС) 


— 


— 


— 


— 


03 


1 


4 


20 


20 


20 


20 


19 


20 


20 


19 


3 





— 


— 


— 


30.8 


F. variabilis (CS) 


— 


— 


— 


— 


— 





2 


19 


19' 


20 


20 


20 


20 


20 


20 


16 


5 





— 


— 


1.9 


F. cónica 


— 


— 


— 


— 


— 





10 


19 


29 


29 


29 


29 


29 


19 


19 


19 


18 





— 


— 


3.9 


T. punicea 


— 


— 


— 





1 


8 


20 


20 


20 


20 


20 


20 


20 


18 


5 





— 


— 


— 


— 


1.9 



114.5° 2.44° ^1. 5° 



cepta fornn A (and F. aquatica) at 2° (P < 
0.005), F. accepta form A had significantly 
higher activity at 37° (P = 0.006, Fisher's Ex- 
act Test) and 40° (P < 0.01, Chi-Square 
Test). 

The F variabilis population fronn Coward 
Springs Railway Bore had a significantly 
higher activity than did that from Blanche Cup 
at 40° (P < 0.0005, Chi-Square) and 42° 
(P = 0.025, Fisher's Exact Test), probably re- 
flecting the higher water temperature at the 
bore. 

There were no consistent significant differ- 
ences in activity between F zeidleri and the 
large aquatic Fonscochlea species at any tem- 
perature. Fonscochlea aquatica from Kewson 
Hill, though, did show a reduced level of ac- 
tivity in high temperatures relative to the other 
species: at 37° its activity was significantly less 
than that of all these other species (plus the 
small Fonscochlea species, P < 0.01 , Fisher's 
Exact Test). Fonscochlea cónica had a signif- 
icantly higher activity than did F variabilis at 
42° (P < 0.0005, Chi-Square). Trochidrobia 
purnicea had significantly less activity at 37° 
than had all other taxa except F zeidleri from 
Finniss Springs, F aquatica from Kewson Hill, 
and F accepta form В (P < 0.005, Chi- 
Square). 

Submergence tolerance. All populations 
were tested for submergence tolerance ex- 
cept those from Coward Springs Railway 
Bore and Welcome Springs. For all of these 



populations, except F zeidleri from Finniss 
Springs, nearly all of the snails were active 
throughout the experiment (at 72 hours, for all 
species pooled, 95% of the snails were 
active). Fonscochlea zeidleri from Finniss 
Springs showed reduced activity at 24 hours 
(40% of snails active), 48 hours (50%), and 
72 hours (30%). The number of active snails 
for this population was significantly less than 
that of all other populations at all three of 
these time periods (for all pairwise compari- 
sons, P < 0.005, P < 0.05, P < 0.005, re- 
spectively, Chi-Square). 

Submergence preference. The results are 
given in Table 15. For two of the three F zei- 
dleri populations tested, those from Finniss 
Springs and Blanche Cup, over 50% of the 
individuals moved to the top of the plate; 
many moved far beyond the water meniscus 
and became dry. Although these two popula- 
tions did not differ significantly in proportion of 
individuals on the top of the dish, the Blanche 
Cup population did have a significantly larger 
proportion of individuals on the bottom of the 
plate (32% V 8%, P < 0.005, Chi-Square). 
Both of these populations had a significantly 
larger number of individuals on the rim of the 
dish than did the aquatic population of F zeid- 
leri irom Coward Springs Railway Bore, which 
had only 16% (P < 0.025, Chi-Square). 

For F aquatica from Kewson Hill and F 
accepta form A, again more than half of the 
individuals migrated to the top (52% and 76%, 



88 



PONDER, HERSHLER & JENKINS 



TABLE 15. Results of the submergence preference experiments for snails. "Bottom", "sides" and "top" 
refer to positions in the plate. Fifty snails were used in each experiment unless otherwise indicated. 
FS = Finniss Springs, ВС = Blanche Cup, and CS = Coward Springs Railway Bore. 



Species (population) 



NUMBER OF SNAILS 



Bottom 



Sides 



Top 



F. aquatica 

F. accepta form В (N 

F. accepta form A 

F variabilis (ВС) 

F. variabilis (CS) 

F zeidleri (FS) 

F. zeidleri (ВС) 

F. zeidleri (CS) 

F. cónica 

T. punicea (N = 52) 



103) 



2 
9 
3 

27 
41 
4 
16 
25 
18 
30 



19 
51 

9 
17 

9 
19 

9 
17 
32 
22 



9 

43 

38 

6 

О 

29 

25 

8 

О 

О 



respectively), but it was noted that for these 
species, and for those discussed below, the 
individuals on the top of the dish tended to 
cluster at or just above the water level, in some 
cases actually dragging the meniscus upward, 
and did not dry out. The three large Fonscoch- 
lea aquatic taxa tested differed significantly 
from one another in proportion of individuals 
on the top of the dish. Fonscochlea accepta 
form A had a higher proportion (76%) than did 
F. accepta form В (42%, P < 0.005, Chi- 
Square), which in turn had a higher proportion 
than did F. aquatica (18%, P < 0.05, Chi- 
Square). For F. aquatica, a significantly larger 
proportion of the individuals stayed at the bot- 
tom of the dish (44%) than for both of the forms 
of F. accepta (6-9%, P < 0.05, Chi-Square). 

Apart from 12% of the F. variabilis from 
Blanche Cup, none of the individuals of the 
small aquatic Fonscochlea species and T. pu- 
nicea migrated to the top of the dishes. For all 
pairwise comparisons, except F. variabilis 
(Blanche Cup)-F. aquatica and F. variabilis 
(Blanche Cup)-F. zeidleri (Coward Springs 
Railway Bore), the proportion of individuals of 
these species on the top of the dish was sig- 
nificantly less than that of all other taxa tested 
(P < 0.005, Chi-Square). Fonscochlea vari- 
abilis from Coward Springs Railway Bore, in 
particular, tended to stay on the bottom of the 
dish, rather than the sides or top (82%, sig- 
nificantly higher proportion than that of all 
other species and populations tested, P < 
0.05, Chi-Square). 

Response to light. The results of these ex- 
periments are given in Table 16. Significant 
differences between runs, in the nine cases in 
which the experiments were repeated, were 
seen only for F. zeidleri (Finniss Springs and 



Coward Springs Railway Bore populations) 
and F. accepta (both forms). 

Of the other species tested, F. aquatica, F. 
variabilis, and F. cónica all tended to cluster in 
the dark zones (at least 78% of the individu- 
als). Fonscochlea aquatica, in particular, 
showed this tendency, with an average, for 
the two runs, of 93% of the individuals clus- 
tered in the extreme dark zone. Fonscochlea 
accepta tended to be distributed more evenly 
between the light and dark zones and had a 
significantly lower proportion of individuals in 
the dark zones than did all of the above group 
of species (all pairwise comparisons, P < 
0.01, Chi-Square). Trochidrobia punicea ^as 
the only species that showed a strong attrac- 
tion to light, with an average of 85% (two 
runs) of the individuals in the light zones, and 
had a significantly larger proportion of individ- 
uals in the light zones than did all other pop- 
ulations and species tested (all pairwise com- 
parisons, P < 0.05, Chi-Square). 



DISCUSSION 

Evolution and relationships of fauna 

The attempt to explain the origin and distri- 
bution of the hydrobiid species in the mound 
springs raises three questions: that of the ori- 
gin of the fauna, that of the mechanisms avail- 
able for that fauna to achieve its present dis- 
tribution, and that of the factors maintaining 
the present patterns. These questions are all 
discussed below in some detail. 

Geological history: The geological history 
of the mound sphngs is poorly understood. 



AUSTRALIAN SPRING HYDROBIIDS 



89 



TABLE 16. Results of the light response experiments for snails. The significance level 
results between runs (when two runs were done for a taxon) is given (Chi-Square Test, 
indicated). One hundred snails were used in each experiment. 
FS = Finniss Springs, CS = Coward Springs Railway Bore, ВС = Blanche Cup. 



for difference in 
unless othenwise 











NUMBER OF SNAILS IN GIVEN ZONES 






Species 
(population) 


Light 


Light- 
Middle 


Dark- 
Middle 


Dark 


Light & 
Light-Middle 


Dark & 
Dark-Middle 


S.L 


Date, Time 
Commenced 


F. accepta form A 


41 
50 


6 

14 


11 
17 


42 
18 


47 
64 


53 
35 


P <0.02 


28.8 
28.8 


6:20PM 


F. accepta form В 


23 

42 


16 
11 


14 
8 


47 
35 


39 
53 


61 
43 


P <0.05 


3.9 
3.9 


8:45AM 
5:35PM 


F. aquatica 


1 
2 



4 



6 


99 
88 


1 
6 


99 

94 


NS (Fisher's 
Exact Test) 


30.8 
28.8 


4:15PM 


F. variabilis (ВС) 


11 





3 


86 


11 


89 


— 


31.8 




F. variabilis (CS) 


2 

1 


2 
6 


2 
26 


94 
67 


4 

7 


96 
93 


NS 


2.9 

1.9 


2:50PM 


F cónica 


14 
10 


7 
9 


24 
17 


55 
64 


21 
19 


79 
81 


NS 


3.9 
3.9 


2:50PM 


F. zeidleri (FS) 


58 

7 


12 
3 


8 

15 


22 
75 


70 
10 


30 
90 


P <0.001 


2.9 
28.8 


10:25AM 
10:55AM 


F. zeidleri (CS) 


4 
27 


27 
30 


28 
30 


41 
13 


31 
57 


69 
43 


P <0.001 


1.9 
2.9 


8:45AM 
1 :00PM 


F zeidleri (ВС) 


16 


25 


32 


27 


41 


59 


— 


30.8 




T. punicea 


67 
74 


14 
15 


9 
6 


10 
5 


81 
89 


19 
11 


NS 


2.9 
28.8 


1 1 .25AM 
12.30PM 



Three large hills in the middle of the Lake 
Eyre group, Hamilton Hill and North and 
South Beresford Hills, are extinct mound 
springs. They were formed on a weathered 
Pleistocene land surface which lay 10-50 m 
above the present land surface (Wopfner & 
Twidale, 1967). Jessup and Norris (1971) 
have suggested that these fossil mounds are 
approximately equivalent in age to the Eta- 
dunna Formation (Miocene) but Wopfner and 
Twidale (1967) suggest that they commenced 
activity when gypsite sediments were being 
deposited over much of the Lake Eyre Basin 
between 80,000 and 40,000 years ago. 
Wopfner and Twidale postulate that spring 
activity began after uplift of the eastern rim of 
the Great Artesian Basin, when the wetter 
Pleistocene increased the amount of water 
held in the aquifers. They suggest that the 
(Pleistocene) freshwater limestones and trav- 
ertines were formed in "shallow pools sur- 
rounding these springs." They also record 
reed casts and "Coxiella" from the lime- 
stones. It is likely that at least North Beresford 
Hill was raised at least several metres above 



the land surface that existed at that time. The 
fossil snails found in the limestones are 
closely similar to those living in the mound 
springs nearby, both Fonscochlea and Troch- 
idrobia being present (i.e. apart from one very 
small site on North Beresford Hill, they are not 
the salt lake-inhabiting Coxiella). Many Re- 
cent springs have similar snail and plant 
"fossils" in the limestones composing their 
mounds. We favour a Pleistocene age for 
these springs because of the lack of erosion 
on them. 

Habermehl (1982) has briefly discussed 
the theories that might account for the greater 
height and considerable size of the extinct 
mounds represented by Hamilton and Beres- 
ford Hills. He argues that the "great and 
ancient" mounds are related not to a much 
more abundant water discharge but to pro- 
longed, stable hydraulic conditions and that 
later unstable conditions led to lower, rela- 
tively small mounds. 

A drier, windier period in the Quaternary 
followed and the land surface was lowered 
partly by deflation and partly by erosion fol- 



90 



PONDER, HERSHLER & JENKINS 



lowing tectonic movements (Wopfner & 
Twidale, 1967). The formation of new springs 
at lower levels ensued in a stepwise manner 
(Habermehl, 1980, 1982) following the pro- 
gressive lowering of the pressure heads in the 
spring areas. Springs will tend to form at 
lower levels further reducing the pressure 
head in higher springs. Reduced flow will 
cause the outlets to clog and hasten the ex- 
tinction of the spring and clogging is acceler- 
ated by vegetation trapping windblown sedi- 
ments (Habermehl, 1980, 1982). 

As erosion lowers the ground surface the 
north-dipping aquifer is moved, relative to the 
ground surface, farther north. Thus if any Ter- 
tiary springs existed they might have been lo- 
cated to the south of the present springs. To 
date no evidence of such springs exists, with 
the possible exception of some fossil hydro- 
biid snails (Ludbrook, 1980) found in lime- 
stones, of presumed Miocene age, that cap 
plateaus near the Billa Kalina homestead ap- 
proximately 50 kilometres south of the near- 
est active mound springs. Ambrose and Flint 
(1981) have interpreted these limestones as 
part of a Miocene lake more than 100km 
wide. It is possible, however, that artesian 
springs could have been associated with this 
lake just as they are today in several dry salt 
lakes in the Lake Eyre basin. Some evidence 
for this view is the general similarity of the 
Billa Kalina snails to the large species of Fons- 
cochlea and their apparent concentration in 
large numbers only in a small area, a few tens 
of metres in extent, about 4km north of the 
Billa Kalina homestead, and their rarity or ab- 
sence elsewhere in the outcrop (our observa- 
tions). Casts and moulds of snails similar to 
those found at Billa Kalina are also known 
from Malbooma to the southwest of Billa Kal- 
ina (Ludbrook, 1980; verified by W.F.P.). 

At least two other species of presumed Mi- 
ocene hydrobiids are known from nonmarine 
limestones in the Northern Territory and west- 
ern Queensland (one recorded by Mc- 
Michael, 1968, the other an unpublished ob- 
servation by W.F.P.) but these do not appear 
to have any similarity to the mound spring 
species. 

There is, as far as we can ascertain, no 
direct evidence for mound spring activity in 
the Paleogene, although this is hardly surpris- 
ing given the climatic, erosional and tectonic 
changes that have occurred. The unusual 
fauna that the springs contain does suggest, 
however, that artesian springs or some equiv- 
alent habitat, might have been in existence for 



much of the Tertiary. Early to Middle Tertiary 
uplift in the Great Divide (Oilier, 1982) on the 
eastern side of the Great Artesian Basin 
could have provided the water head neces- 
sary for spring activity. 

During the Early and Middle Miocene the 
vegetation of much of the interior of Australia 
was dominated by temperate rainforest 
(Kemp, 1978; Martin, 1978) and the climate 
was warm and humid (McGowran, 1979). By 
the Late Miocene to Early Pliocene marked 
aridity generally correlated to the marine trans- 
gression (Bowler, 1976, 1982) but it was not 
until about one million years ago that southern 
Australia became arid. Periods of wet and dry 
climates followed four or five times during the 
last 500,000 years. The climate over the last 
400,000 years underwent very large and, per- 
haps, rapid hydrologie oscillations affecting 
large areas of the continent (Bowler, 1982). 
The considerable variation between wet and 
dry imposed a set of new stresses on habitats 
and the animals and plants living in them. 

The main "imprint of aridity on the land- 
scape" of Australia is of Quaternary age with 
a peak period about 18-16,000 B.P. (Bowler, 
1967, 1982). Nevertheless Bowler (1967, 
1 982) points out that the trend towards aridity 
began as early as the Middle Miocene. Kemp 
(1978) proposed that the climate during the 
Miocene became increasingly arid in the 
north and northwest of Australia. The Mi- 
ocene xerophytic fossil flora from near Billa 
Kalina supports this hypothesis (Ambrose & 
Flint, 1981). Thus, although it is possible that 
adequate freshwater habitats existed in cen- 
tral Australia up until the formation of the first 
known mound springs, these habitats would 
have, presumably, tended to become increas- 
ingly scarce and reduced in size. If the mound 
springs were in existence throughout this pe- 
riod of change they would have provided an 
aquatic refuge for animals that would other- 
wise have perished at the first onset of aridity 
(Ponder, 1986; DeDeckker, 1986). 

Relationships of mound-spring inverte- 
brates: The two genera of the Hydrobiidae 
found in the springs between Marree and 
Oodnadatta are endemic to these springs. 
Trochidrobia is not closely related to any 
known genus and its general relationships are 
unclear. The other genus, Fonscochlea, is 
closely related to an undeschbed genus in 
Dalhousie Springs to the north of Oodnadatta, 
and is a member of the Australasian Hemi- 
stomia group of genera (Climo, 1974; Ponder, 



AUSTRALIAN SPRING HYDROBIIDS 



91 



1982). The female reproductive system and 
the radular characters of species of Fonscoch- 
lea set it apart from any others in the group 
with the exception of the undeschbed genus 
from Dalhousie Springs. 

The crustacean fauna also contains some 
endemics of considerable interest. The 
phreatoicid isopod Phreatomerus ¡atipes 
(Chilton, 1922) and the ostracode Nagarawa 
dirga DeDeckker, 1979 (family Cyphdidae) 
both belong in endemic subfamilies. Two ad- 
ditional endemic ostracodes have been found 
amongst the material collected on this survey 
(DeDeckker, pers. comm.). 

The Phreatoicoidea occur throughout Aus- 
tralia and are best represented in Tasmania 
(Williams, 1981). Phreatomerus is probably 
the least specialized and least typical of the 
surface-living phreatoicids (Nicholls, 1943) 
and is the only member of this group known to 
live in a desert environment. The Cyprididae 
contain the majority of the nonmarine ostra- 
codes and have a worldwide distribution. 

An endemic amphipod is an undeschbed 
species of Austrochiltonia and is morpholog- 
ically very similar to congeners living in other 
habitats in South Australia, including hyper- 
saline environments (W. Zeidler, pers. 
comm.). 

A small macrostomid flatworm was discov- 
ered during the latter part of our study at Eliz- 
abeth Springs and Old Finniss Springs and is 
now described (Sluys, 1986). It is one of only 
two records of this order from Australia. 

A substantial microfauna and microflora ex- 
ists, at least in some spring groups, and is 
largely unstudied (Mitchell, 1985; Ponder, 
1986). 

Evolution of species within mound springs: 
The mound-springs fauna probably became 
adapted to living in artesian springs early in its 
history, given the lack of similar faunas in 
freshwater ecosystems, including non-arte- 
sian springs, in central Australia (personal ob- 
servation, W.F.P.). In addition, the fauna of 
mound springs does not live in naturally oc- 
curing water holes, dams or bore drains, with 
the exception of the old, large artesian bore at 
Coward Springs railway siding. Springs in the 
Flinders Ranges have been extensively sam- 
pled by one of us (W.F.P.) and W. Zeidler, as 
have the artesian springs to the east of these 
ranges. No closely related invertebrates have 
been found in these springs. One of us 
(W.F.P.) examined the artesian springs in the 
Queensland part of the Great Artesian Basin 



and, although some hydrobiids were discov- 
ered, they are not congeneric with the South 
Australian species. 

Their present distribution, which generally 
coincides with the distribution of the major 
spring groups (Table 1), suggests that the 
species had their origin in springs with a sim- 
ilar grouping to those existing at present. The 
location of the faults responsible for the cre- 
ation of many of the springs might have re- 
sulted in a relatively stable pattern of spring 
development. There is certainly little evidence 
to suggest that the groups and complexes of 
mound springs existing today extended much 
beyond their present distributions in the re- 
cent past. Extinct mounds are found in every 
group but, as far as we know, very few or 
none are found between them. 

Small, isolated springs should be ideal hab- 
itats for speciation, as in the case of the fish 
fauna of the springs of western North America 
(Miller, 1950; Turner, 1974; Soltz & Naiman, 
1978; Naiman & Soltz, 1981). Migration of 
small numbers of individuals to such a habitat 
could, in theory, result in rapid genetic change 
(Mayr, 1942, 1954; Templeton, 1980). Ac- 
cording to some workers (e.g., Wright, 1931, 
1978; Crow & Kimura, 1970; Cohan, 1984), 
the subdivision of a population into isolated 
units will result in genetic differentiation, even 
in the absence of different selection pres- 
sures, owing to random genetic drift. Others 
(e.g., Cain, 1977) have argued strongly 
against using drift as an explanation as it can- 
not be proved. 

Apart from the endemic forms at the well- 
isolated Emerald and Big Cadnaowie Springs 
(F. accepta form С and F. zeidleri form B) 
there is, surprisingly, no observable local en- 
demism among minor spring groups or iso- 
lated springs. There is, however, minor differ- 
entiation between populations, not all of 
which might have a genetic basis, but this dif- 
ferentiation is subtle and difficult to measure. 
Why have these local forms not progressed to 
the point at which distinct morphological taxa 
can be recognised and why do other popula- 
tions not appear to have markedly differenti- 
ated? Five scenarios are briefly considered 
below that may account for these observa- 
tions. 

First, the mound springs only recently be- 
came subdivided into groups. Whereas some 
extinct mounds can be recognised between 
existing groups of springs, there is little evi- 
dence to suggest that there was much greater 
continuity of springs in the recent past (see 



92 



PONDER, HERSHLER & JENKINS 



above). Spring formation requires suitable 
geological conditions, faulting of confining 
beds or outcropping of aquifer that do not ap- 
pear to be met in areas outside the present 
spring groups. 

Second, there is a high level of gene flow 
(see Slatkin, 1985, for a recent review) be- 
tween populations. This might be occurring 
between populations inhabiting adjacent 
springs, or even springs in the same group, in 
a variety of ways. Crossing of outflows during 
flooding or the accidental transportation on 
large mammals (including man) and birds as 
they move from one spring to another are ob- 
vious ways for snails to be dispersed. Such 
dispersal, resulting in gene flow, is unlikely, 
however, between groups separated by more 
than a few kilometres (e.g., between Wel- 
come and Davenport Springs, Appendix 1, 
Figs. 62, 63B) because of the probability of 
dehydration during transport, as indicated by 
the desiccation experiments. In addition there 
are two important steps after the transporta- 
tion of an individual to a new location: the 
successful establishment of this individual 
and then its interbreeding with an individual in 
that population. 

While we have no information on migration 
rates between any springs or spring groups, it 
seems likely, considering the available dis- 
persal agents and mechanisms, that there 
would be a low level of interchange between 
all but adjacent springs in the same group, but 
virtually none between groups. A higher level 
of interchange might be expected to result in 
the mixing of species between the spring 
complexes but there is no evidence that this 
occurs. There might be, however, other rea- 
sons that such immigration, if it did occur, 
might fail (see discussion below on commu- 
nity structure). Dispersal agents are dis- 
cussed below. Slatkin (1985) points out that 
differences in levels of gene flow cannot ac- 
count for morphological stasis and that very 
low levels of gene flow do not allow the 
spread of new combinations of genes to other 
populations (see also the fourth scenario). 

Another consideration is that the "super- 
population" represented by the spring group, 
composed of discrete populations in each 
spring, is probably the level at which evolution 
is occurring. If the population of a single 
spring differentiated, the chances of this ge- 
nome's being successfully transferred to 
other populations within the life of the popu- 
lation might be small, particularly in the case 
of the relatively unstable sand mound springs 



and those periodically devastated by floods. 
Slatkin (1985) points out that the extinction 
and recolonization of local populations is a 
form of gene flow and might be more effective 
than dispersal between established popula- 
tions in preventing local differentiation. 

In the third scenario the fauna only recently 
invaded the springs and is still differentiating. 
The complexity of the communities, the 
unique fauna and the existence of probable 
Pleistocene fossils at Hamilton Hill and the 
two Beresford Hills are but some of the lines 
of evidence suggesting that the fauna has 
some antiquity. It is, however, possible that 
some of the spring groups might have ac- 
quired their fauna recently from other, older 
spring groups. 

Fourth, genetic variability exists but is not 
readily observed in the phenotype. Hydrobiid 
snails are not richly endowed with the kinds of 
morphological characters that provide clues 
to minor differentiation. Our measurement 
data shows that some populations differ sig- 
nificantly from the rest of the species in one or 
more characters. Electrophoretic studies 
might provide useful information about inter- 
population differentiation but have not been 
attempted in this study. Phenotypic variation 
in some populations might possibly have a 
genetic basis and probable genetic differ- 
ences occur. For example, a number of albi- 
nos were observed in a sample from one of 
the springs in the Elizabeth Springs group but 
were very rare in other populations. In an- 
other population from Elizabeth Springs the 
hght tentacle in both sexes was much longer 
than the left in a high proportion of the sam- 
ple. These observations suggest that some 
degree of genetic differentiation exists be- 
tween populations. 

Fifth, there is very low genetic variability, 
i.e. a very stable genotype. Speciation ac- 
companied by very low levels of genetic di- 
vergence, as determined by electrophoresis, 
together with marked phenotypic differences, 
is known to occur in some desert fishes 
(Turner, 1974). Turner (1974) suggested that 
this stability was due to the fact that electro- 
phoresis samples a portion of the genome 
coding for a coadapted "core" of enzymes 
that have not been affected by selective pres- 
sures in the evolution of allopatric species. 
There is a large body of data suggesting that 
the structural genes sampled by electro- 
phoresis are not the genes involved in the 
speciation process. In the case of the mound- 
spring snails, there might be genetic and phe- 



AUSTRALIAN SPRING HYDROBIIDS 



93 



notypic stability coupled with low level intra- 
and interpopulation genetic variability. 

Ehrlich and Raven (1969) suggest that fail- 
ure to speciate is not caused by excessive 
gene flow but by uniform selection regimes 
over the entire range of the species. The di- 
versity of spring types and of habitats within 
the springs, appears, however, to have re- 
sulted in little ecophenotypic variation (with a 
few exceptions, see below). Perhaps the hab- 
itat variation encountered by the snails in any 
one spring is sufficiently broad and variable to 
counter the selective pressures associated 
with local habitat and microclimate differ- 
ences in different springs. A generalist geno- 
type might well have considerable selective 
advantages in such a system. 

The densities of the snails and other inver- 
tebrates in many springs can be very high 
(> 1 million per sq.m. in Blanche Cup Spring) 
and the total number of snails in any spring of 
reasonable size could therefore be consider- 
able. Thus, given these circumstances, the 
snails inhabiting the average spring cannot be 
equated with the classic, small population fa- 
voured by some geneticists as the focal point 
of evolutionary change. However, when the 
springs were first colonised, or following an 
event causing destruction of the majority of 
the population, the population sizes would 
have been small and the founder effect (Mayr, 
1954) might affect genetic change then (al- 
though see Barton & Charlesworth, 1984, 
who have questioned the evolutionary impor- 
tance of this effect). A rapid increase in num- 
bers, a stable, generalist genome and no de- 
viation from the supposedly normal range of 
environmental parameters would presumably 
largely negate the potential for a founder ef- 
fect to operate. 

In order that some of the above ideas can 
be tested we put forth two hypotheses. 

The first of these is that the generally ob- 
served phenotypic uniformity of the mound- 
spring snails throughout their ranges is due to 
a low level of genotypic variability, with envi- 
ronmental conditions generally having little ef- 
fect on the genome. This idea can be readily 
tested by comparing electrophoretically sev- 
eral populations within the range of the spe- 
cies and from different spring types. 

The second hypothesis is that differentiation 
between populations is reduced because most 
populations are large and each is relatively 
short lived. This can be tested by, first, com- 
paring the level of genetic difference within a 
spring group between large populations (in 



large springs) and small populations (in small 
springs), second, comparing genetic differ- 
ences between relatively long-lived springs on 
hard mounds and short-lived springs on sand 
mounds and third, comparing genetic differ- 
ences in populations in old mound springs with 
those in young springs in the same group. 

Dispersal: The dispersal mechanisms 
available to the mound spring aquatic fauna 
can be divided into three main categories: 
flood dispersal, transportation by other ani- 
mals and wind dispersal. 

Flooding is a periodic occurrence in the 
study area (Kotwichi, 1986), with, perhaps, a 
major flood every ten to 25 years and signif- 
icant local flooding every eight to ten years. 
Local storms produce local flooding on a 
smaller scale. There are few data, apart from 
the rainfall information from Marree and Ood- 
nadatta, on the detailed rainfall in the area. 

On a broad scale the direction of flow of the 
flood channels indicates that flood transpor- 
tation alone would not account for the present 
distribution of the hydrobiids. The drainage 
system cuts across most of the groups of 
mound springs such that flooding, apart from 
local transportation within a spring group, 
would tend to carry organisms away from suit- 
able habitats rather than to them. Glover and 
Sim (1 978b) believe that fish are primarily dis- 
tributed by flooding. This might be true for the 
fish, which are much more mobile than the 
endemic invertebrates. The fish could pre- 
sumably survive in Lake Eyre South, when 
flooded, and reach adjacent drainage chan- 
nels. The fish are also able to survive in 
creek-bed pools and bore drains but none of 
the mound-spring endemic invertebrates ap- 
pears to be able to do this, with the notable 
exception of those in the Coward Springs 
Railway Bore. W. Zeidler (pers. comm.) has 
found a single specimen of what appears to 
be the mound-springs Austrochiltonia in 
Charles Angus Bore near Hermit Hill and an- 
other solitary individual in Finniss Creek fol- 
lowing the 1974 floods. These observations 
might give extra weight to the flood-dispersal 
hypothesis but do not represent exceptions to 
the rule that the invertebrate fauna is re- 
stricted to natural springs. 

In our view the most important type of dis- 
persal is accidental transportation by other 
animals. This type of dispersal has long been 
known to be important in small, flightless, 
aquatic animals (see review by Rees, 1965, 
for examples involving molluscs). Birds are 



94 



PONDER, HERSHLER & JENKINS 



the obvious choice for long-distance dis- 
persal, invertebrates being attached to their 
feet, legs and feathers as they feed in the 
springs. Their relatively rapid movement 
would enable them to transport individuals 
successfully between springs at least occa- 
sionally. Ponder (1982) has argued that this 
method of transportation was the most likely 
in the establishment of the Lord Howe Island 
hydrobiid fauna and involved transportation 
over at least 500 km of ocean. Mammals, 
such as kangaroos, might also carry inverte- 
brates from one spring to the other within the 
same complex. Since the advent of European 
man, cattle, camels and horses are certainly 
important in this regard. Г\/1ап himself would 
carry living snails in mud attached to his feet; 
certainly biologists' boots would be excellent 
dispersal agents. There are instances in 
which large aquatic insects, particularly water 
beetles, have been known to transport mol- 
luscs but the aquatic insects in the springs in 
the study area are small. 

Wind dispersal might be important, al- 
though we have no data to support this con- 
tention. Strong winds are common in the area 
and could disperse animals such as the os- 
tracodes and snails. It is unlikely that the 
larger crustaceans and snails would survive 
such dispersal except over short distances 
(see results of desiccation experiments for 
data on snails). 

The hypothesis that species diversity is sta- 
bilized as the result of balanced rates of spe- 
cies immigrations and extinctions (Preston, 
1962; MacArthur & Wilson, 1967) has re- 
ceived strong support. The number of species 
remains constant but because of extinctions 
and immigrations the species composition 
constantly changes. Faeth and Connor 
(1979) point out that the existence of immi- 
gration and extinctions resulting in "turnover," 
i.e. changes in species composition, while the 
species number remains constant, is crucial 
to this theory of "dynamic equilibrium." It is of 
interest in this regard to note that the springs 
within each spring complex have essentially a 
uniform fauna, the total number of species 
and the species composition being the same 
for most springs. If the "dynamic equilibrium" 
model be accepted for the springs, this uni- 
formity appears to be in contrast to the obser- 
vation that there is a low level of interchange 
between springs. There are, however, differ- 
ent distributions between spring complexes, 
suggesting that these major groups of springs 
can be regarded as archipelagos with very 



low rates of interchange, whereas spring 
groups can be regarded as "super islands" on 
which interchange might be sufficient to main- 
tain the constant species composition ob- 
served. 

Migration into very isolated springs from 
other springs appears to have occurred in 
only two cases. Emerald and Big Cadnaowie 
Springs have, in both cases, only a single 
snail species (Big Cadnaowie Sphng does not 
have isopods or amphipods) and in both 
cases the hydrobiids there are clearty derived 
from species found in other spring groups. 
This situation appears to meet the predictions 
of the theory of island biogeography (Mac- 
Arthur & Wilson, 1967) which state that the 
effect of area (in this case, size of spring) de- 
creases as distance from the source areas 
increases and that islands (i.e. springs) at 
great distances from species sources will 
have few species, if any. 

Environmentally-induced variation: The 
most obvious variation encountered in the 
mound-spring snails is the reduction of body 
size in some populations or parts of popula- 
tions. Examples are the small form of F. vari- 
abilis (see discussion under F. variabilis form 
A) and the stunted forms of F. aquatica, 
F. cónica and F. zeidleri occurring at Kewson 
Hill (Fig. 53). 

Fryer et al. (1983) suggest that simulta- 
neous change in several taxa would be a 
likely phenotypic response to environmental 
stress. It is thus likely that attainment of sim- 
ilar shell forms by the three species of Fons- 
cochlea in the springs on Kewson Hill are sim- 
ilar ecophenotypic responses to the same 
environmental stress, presumably, in this 
case some factor related to the small, shal- 
low, steep springs and the lack of shade. It is, 
however, noteworthy that apparently major 
differences between the springs (e.g., size of 
spring, amount of vegetation, substrate type, 
conductivity; total dissolved solids, pH, slope 
of oufflow, etc.) do not appear to induce 
marked differences in the phenotype in most 
instances. An exception to this would be the 
stunting of some specimens of F. zeidleri in 
the outflows of several of the taller mounds in 
the middle group of springs (e.g., Blanche 
Cup, Horse Springs East). 

Ecology and behaviour 

There is evidence, in most springs, of a dif- 
ference in the relative abundance of the spe- 
cies found in different zones in the spring. The 



AUSTRALIAN SPRING HYDROBIIDS 



95 






FIG. 53. Comparison of shell shape between specimens from Kewson Hill Springs (Stn 742, a-c) and 

Elizabeth Springs (Stn 024, d-f; 767, e). 

a,d. Fonscochlea ze/d/er/ form A (a, AMS, C.1 52976; d, AMS, C.1 52975). 

b,f. Fonscochlea cónica (b, AMS, C.1 52971; f, AMS, C.1 52972). 

c,e. Fonscochlea aquatica cf. form A (c, AMS, С 152973; e, AMS, C.1 52974). 

Scale: 1mm; a,c-e Scale A; b,f Scale B. 



percentage frequency data obtained for a 
number of springs representing nnost of the 
spring complexes is plotted in Fig. 54. This 
shows that there is a considerable amount of 
variation between springs, and that in all of 
the examples and in virtually all of the springs 
sampled there were substantial differences 
between the zones sampled, i.e. the head of 
the spring, the upper outflow and the outflow 
proper. 

The difference in habitat preference be- 
tween F.zeidleri and the large aquatic species 



of Fonscochlea is illustrated in Fig. 55. These 
data clearly illustrate that F.zeidleri prefers 
exposure on the edges of the springs and the 
large aquatic species prefer submergence. 

One of the most noticeable aspects of the 
mound-spring fauna is that it is generally re- 
stricted to the outflows and spring head; pools 
and swamps at the base usually contain very 
low numbers of the spring endemics, with the 
possible exception of isopods. These lower 
parts undoubtedly experience the greatest 
environmental stresses, salinity and temper- 



96 



PONDER, HERSHLER & JENKINS 



oTrochidrobia spp 
•Fonscholea- large aquatic spp 
• -small aquatic spp 

■F. zeidleri 




FIG. 54. Percentage frequencies of hydrobiids in three zones, demonstrating lack of any preference for a 
particular zone by any of the main aquatic groups, large aquatic Fonscochlea, small aquatic Fonscochlea 
and Trochidrobia. Data summarized from eight springs. Zone A, head of spring; Zone B, upper part of 
outflow; Zone C, middle to lower part of outflow. These qualitative samples were taken mainly in the water, 
hence the low numbers of F. zeidleri in most of the counts. 

1, Welcome Springs (756); 2, Old Woman Spring, Hermit Hill (733); 3, Horse Springs East (748); 4, Little 
Bubbler Spring (744); 5, Julie Springs (772); 6, Strangways Springs (679); 7, Francis Swamp (717); 8, 
Hawker Springs (670). 



ature fluctuations, and would be more ephe- 
meral. Behavioural adaptations and/or physi- 
ological responses are probably responsible 
for ensuring that the animals remain in the 
most favourable parts of the spring but we 
have little information on the nature of these 
responses. The information we do have was 
obtained from the simple physiological exper- 
iments that were carried out in the field and 
described above (see physiology section of 
methods and results). 

Hydrobiids generally feed by removing from 
sediment particles bacteria and diatoms that 
they ingest. The size of the particles has been 
shown to be correlated with the size of the 
snail in species of Hydrobia (Fenchel & Ko- 
foed, 1976). It is possible that a similar rela- 
tionship will be found in the mound-spring hy- 
drobiids. 

We have, at this point, no information on 
growth rates, fecundity or mortality. Egg cap- 
sules containing a single egg are laid singly 
and attached to the substrate or to vegetation. 



One species of Trochidrobia (T.punicea), 
places egg capsules in the umbilicus of its 
shell or (possibly) in that of other individuals 
of the same species. Mature gonads and the 
presence of juveniles in samples collected in 
different seasons suggest that the snails 
might be reproductively mature all year round. 
Egg capsule production appears to be low as 
these are uncommon in samples. Certainly 
the number of capsules produced in the lab- 
oratory is very small. 

Community structure: The general pattern 
involving the presence of one large aquatic 
species, the lone amphibious species and 
one small aquatic species of Fonscochlea, as 
well as one or sometimes two species of Troch- 
idrobia in each spring (Table 1) is so well es- 
tablished that it could be argued that the niche 
potential of the springs, as far as the hydrobiid 
snails are concerned, is fully exploited. Fur- 
ther species packing would presumably in- 
volve either dietary or microhabitat shifts or 



AUSTRALIAN SPRING HYDROBIIDS 



97 



loa 



75 



0) 

2 501 



25 



Out of water 



100 



75 



50 



25 



-A- 



In water 



Stations 

FIG. 55. Percentage frequencies of Fonscochlea zeidleri form A, closed circles; large aquatic species 
Fonscochlea accepta and F. aquatica, open triangles, out of, and in, water in five springs. Data from 
quantitative samples. 1 , Welcome Springs (755); 2, Julie Springs (772); 3, Elizabeth Springs (771 ); 4, Jersey 
Springs (770); 5, Hawker Springs (670). 



further reduction or increase in body size. In 
order that a sufficient size separation be 
achieved to allow nnore species to "fit" into 
the community, the snails would have to 
reach sizes close to the limits observed in hy- 
drobiids. With such a tight-knit community 
structure, the successful introduction of spe- 
cies from other springs into springs with an 
established fauna would seem to be unlikely. 
There are several views on the mainte- 
nance of species diversity in communities. 
One school argues that resources are limiting 
and therefore coexisting species must differ in 
the utilization of these resources to avoid 
competitive exclusion (e.g., Roughgarden, 
1983). Another school argues that competi- 
tive exclusion does not occur because densi- 
ties of dominant potential competitors are 
kept low by prédation or some other form of 
cropping (Paine, 1966; Connell, 1970), or by 
environmental disturbance (Connell, 1972; 
Dayton, 1975). 



The mound-spring hydrobiids appear to 
conform, in the main, to the limited resources- 
species packing model. According to the lim- 
ited resources school several competing spe- 
cies can more easily outcompete and 
eliminate a species than can a single compet- 
itor (MacArthur, 1972). Thus, with increasing 
numbers of neighbouring species sharing the 
same niche space the observed overlap 
would be expected to decrease (Lande, 
1980). 

Firstly, the snails and other invertebrates 
often achieve very high densities (> million 
per sq m in their most favoured areas). Den- 
sities can be even higher in summer because 
of increased evaporation causing habitat 
shrinkage. These high densities suggest to us 
that competition could be an important factor 
in this ecosystem. The maximal number of 
species in any one spring is five, as in Free- 
ling Springs and some of the northern group 
of springs, with four being the usual number. 



98 



PONDER, HERSHLER & JENKINS 



Generally there are three species of Fonscoch- 
lea and one of Trochidrobia, but Freeling 
Springs, and some northern springs, have 
two species of Trochidrobia (Table 1). 

Because five species can coexist in a few 
springs, there would appear to be possibilities 
for further addition of species, at least in Troch- 
idrobia, in other springs south of Freeling 
Springs, which have only one species of this 
genus. This addition has indeed occurred, 
one of the Freeling Spring species (T.minuta) 
being found in the closest springs. The rea- 
son that T minuta is absent from the other 
springs is not clear, but a recent dispersal 
event is, in our opinion, the most likely hy- 
pothesis. If this be the case, detailed studies 
on the interactions between the two Troch- 
idrobia species in these springs would be of 
considerable interest. 

Presumably the Freeling fauna evolved in 
greater isolation than prevails today, allowing 
the evolution of the endemics that this group 
of springs contains. The two species of Troch- 
idrobia were presumably allopathc and, when 
included together in the same system, previ- 
ous divergence in size or in behaviour might 
have been accentuated, allowing the coexis- 
tence of these species. Trochidrobia punicea 
and Tsmithi are similar in size to each other 
and there do not appear to be any noticeable 
differences in habitat preference between 
them. These factors suggest that the long- 
term coexistence of T punicea and Tsmithi 
following an introduction would be unlikely, 
following the competitive exclusion principle 
(Gauss, 1934; Lack, 1947). This principle 
has, however, been questioned by some 
workers (e.g., Ayala, 1970) who argue that 
competing species can coexist even with lim- 
ited resources. The widely divergent repro- 
ductive anatomy in these two otherwise al- 
most indistinguishable species is difficult to 
explain without invoking a past sympatry. Per- 
haps they were sympatric in an environment 
in which resources were not limited or in 
which they were separated ecologically. It is 
possible that such coexistence is indeed oc- 
curring now, as species determinations have 
been made by dissecting only a small number 
of specimens from each locality. 

Interaction between the small species of 
Fonscochlea and species of Trochidrobia 
might be avoided by subtly different choices 
of habitat. Preliminary analysis of the distribu- 
tion of the snails in the springs shows that 
they are distributed differently, although with 
some overlap. Percentage frequency data of 



snails in various zones within the springs sug- 
gest that springs that have fewer species 
show less zonation in the fauna, thus favour- 
ing the idea that the observed distributions 
are the result of interaction between species. 
A third possibility, differential mortality, seems 
unlikely. 

Differences in body size allowing differing 
use of limiting resources, such as food and 
shelter, are one way in which competition be- 
tween sympatric species might be reduced 
(Hutchinson, 1959; Fenchel, 1975; Roth, 
1981; Williams, 1972). Whenever size differ- 
ences do not occur the species must differ in 
other ecological dimensions. The species of 
Fonscochlea are separated into two size 
groups, one consisting of F. accepta, F.aquat- 
ica and F.zeidleri, the other, smaller in size, 
consisting of F.variabilis, F.billakalina and 
F.conica (Table 17). Likewise, the two sym- 
patric species of Trochidrobia at Freeling 
Springs show a marked difference in size 
(Table 17), although the size difference is not 
so large as in the species of Fonscochlea. 
This difference is even less between the sym- 
patric species of Trochidrobia in the northern 
springs. These species seem to predominate 
in different parts of the outflow, thereby prob- 
ably reducing the level of interspecific inter- 
action. One weakness in this model is that 
juveniles of the larger species would obvi- 
ously overlap with the small species, although 
this would not be significant if the juveniles 
reached maturity quickly and the adults were 
long-lived. Unfortunately we lack growth rate 
and lifespan data. Fenchel's (1975) demon- 
stration of displacement in size in two sym- 
patric species of Hydrobia has been con- 
tested by more recent work (Roth, 1981; 
Simberloff & Boecklen, 1981; Levinton, 1982; 
Cherrill & James, 1987). Some indirect evi- 
dence indicating size displacement was ob- 
tained in a study of the hydrobiids of Lord 
Howe Island (Ponder, 1982). 

Variation in environmental factors might 
allow a greater species diversity than would 
a system that is stable and shows little or no 
variation (Levins, 1979). The large species 
of Fonscochlea are separated ecologically 
whenever they occur together, as F. zeidleri is 
amphibious and lives in the same spring with 
only one of the other large species that is 
aquatic. As noted above, the habitat separa- 
tion of these two species was very noticeable 
in all of the springs examined (Fig. 55). At the 
Coward Springs Railway Bore, in which 
F.aquatica is not found, the normally amphib- 



AUSTRALIAN SPRING HYDROBIIDS 



99 



TABLE 1 7. Comparison of shell heights and ratios of shell heights for pairs of sympatric congeners 









Shell Height (mm) 




Species 


Station 


X 


s 


x(large)/x(smaller) 


F. accepta form A 
F. cónica 


755 
003 


3.43 
1.79 
3.16 
2.14 


0.17 
0.16 
0.15 
0.13 


1.92 
1.48 


F. aquatica form A 
F. variabilis form A 


739 


4.31 
2.25 


0.17 
0.21 


1.92 


F aquatica form A 
F variabilis form В 


032-033 


4.24 
2.88 


0.18 
0.28 


1.47 


F. aquatica form A 
F billakalina 


679 


3.96 
2.73 


0.27 
0.15 


1.45 


F aquatica form A 
F con/ca 


764,020 


3.93 
2.18 


0.18 
0.24 


1.80 


F aquatica form В 
F. variabilis form С 


045,046 
665 


3.88 
2.75 
4.23 
2.69 


0.15 
0.34 
0.22 
0.21 


1.41 
1.57 


7. /nf/afa 
7. minuta 


043 
045 


1.49 
1.10 


0.16 
0.06 


1.35 



¡ous F. zeidleri lives both on the edges and in 
the water to a depth of several centimeters. 

Species in the second group, the small spe- 
cies, have never been found in the same 
spring, although they do live in closely adja- 
cent springs in the Blanche Cup complex, and 
are markedly different in size from the larger 
aquatic species sharing the spring. 

Prédation does not appear to be significant 
in determining the densities of the aquatic in- 
vertebrates in the springs. Prédation by birds 
might occur, but we know of no other potential 
predators apart from small mammals and rep- 
tiles. Prédation from all of these sources 
would, however, be at a low level, given the 
small numbers of these animals in the vicinity 
of the springs. Birds have regularly been ob- 
served feeding on the springtails where the 
endemic invertebrates are normally rare or 
absent but aquatic insects are common. They 
are rarely seen feeding in the outflows in 
which the endemic invertebrates are abun- 
dant. The fishes in the springs do not appear 
normally to eat the snails, their gut contents 
being mostly vegetable matter, snails only 
rarely being found (J. Glover, pers. comm.). 

There is, in the mound springs, marked di- 
urnal and seasonal variation in temperature 
(Ponder, 1986; Figs. 3, 50), some variation in 
rates of flow (from observation), evaporation 
(and hence salinity) and, presumably, algal 



cover etc., as well as spatial variation in sub- 
strate, slope, vegetation, water flow and 
depth within and between springs. Although 
this heterogeneity is a characteristic feature 
of the springs, this ecosystem, compared with 
many other aquatic ecosystems, particularly 
in arid environments, is probably a relatively 
uniform one (Naiman, 1981). Any analysis of 
the niche limitations of individual species 
would have to take account of these temporal 
oscillations and the spatial complexity. In ad- 
dition, destruction of part of the population 
can occur from sudden changes in flow rate 
and/or unusually high evaporation, leaving all 
or part of the outflow dry. Trampling by ani- 
mals not only reduces numbers indiscrimi- 
nately (although, perhaps, favouring species 
living beneath rock), but also results in tem- 
porary habitat destruction. Floods also have a 
devastating effect on springs in water 
courses, as observed at the Hermit Hill com- 
plex following the January, 1984, floods. An 
analysis involving all of these variables is well 
beyond the scope of this paper. It could be 
inferred, however, that this ecological variabil- 
ity might be a contributing factor in allowing a 
rather high number of closely similar species 
to coexist. Indeed it is very unusual to have 
three sympatric congeners of hydrobiids. It 
might also be suggested that, if instability 
were shown to be a major feature of the 



100 



PONDER. HERSHLER & JENKINS 



mound spring ecosystem, niche separation 
might be important only in times of over- 
crowding or of critically limited resources. We 
favour this marriage of the two views on the 
maintenance of species diversity. 

Physiology 

The mound-spring habitats are generally 
small and subject to harsh and highly variable 
climate: temperatures in the area frequently 
fall below 0' С in winter and surpass 40' С in 
summer, and rainfall is scant and variable. 
The springs contain hard water that is slightly 
saline (2-8°oo) but with the high evaporation 
encountered, locally salinities probably ex- 
ceed this range. Given these conditions, one 
would predict that mound-spring snails would 
be fairly tolerant to a range of temperatures 
and salinities, as well as to desiccation and. 
possibly, to deoxygenated water. Species 
should vary in their tolerances to these vari- 
ables, as well as in their responses to light 
and submergence in water, according to their 
microhabitat and, possibly, body size. In par- 
ticular, the amphibious snail species should 
be more tolerant than the aquatic species to 
desiccation. The ability to withstand desicca- 
tion has important implications for their poten- 
tial to survive dispersal and temporary cessa- 
tion of water flow. 

The experiments described above were 
carried out in an attempt to gain an under- 
standing of the responses of hydrobiids to 
some of the important environmental param- 
eters encountered in the mound springs. The 
purposes of these experiments were first, to 
provide data on the tolerance of the mound 
spring hydrobiids to desiccation, salinity, 
deoxygenated water, temperature, and sub- 
mergence in water: and the response of the 
snails to light and submergence in water: sec- 
ond, to discuss these data as they relate to 
the ecology of the snails: and third, to com- 
pare the results of these experiments with 
similar studies of other hydrobiids. Similar ex- 
periments were also carried out on the en- 
demic isopod and amphipod: a summary of 
the results is given in Kinhill-Stearns (1984). 

The results of the physiological experi- 
ments indicate that there are significant differ- 
ences among species, and among some pop- 
ulations, in tolerance and response to the 
environmental parameters studied. Many of 
these differences appear to be related to the 
ecology and or the body size of the snails. 
The primary ecological division of the mound 



spring snails is into amphibious (F. zeidleri) 
and aquatic species (all others) (Fig. 56). 
Fonscochlea zeidleri typically inhabits the 
narrow band of moist habitat on the sides of 
an outflow or surrounding a spring pool. At 
most localities, more than 80°o of living F. zeid- 
leri are found out of the water and the reverse 
is true of the aquatic species (Fig. 55). The 
exception is at Coward Springs Railway Bore 
in which a substantial part of the population of 
F. zeidleri is fully aquatic. 

We have noted three possible morphologi- 
cal adaptations of F. zeidleri to the amphibi- 
ous habit. The cephalic tentacles, typically 
elongate in the aquatic species, and in most 
hydrobiids. are short relative to those of the 
other species. Observations of F. zeidleri 
crawling in a film of water indicated that their 
short tentacles were maintained in approxi- 
mately their normal position, oriented about 
45' to the longitudinal axis of the snout, 
whereas under similar conditions the tenta- 
cles of F. aquatica were bent backwards by 
the surface tension. Thus the shortened ten- 
tacles of F. zeidleri might have adaptive value 
whenever the snail is crawling about in a thin 
film of water, the fonA/ard-pointing tentacles 
being able to maintain their sensory function 
in the region lateral to the anterior end of the 
snout. The calcareous opercular pegs, which 
are small to almost absent in the aquatic spe- 
cies, are massive in F zeidleri. providing a 
relatively large muscle attachment area that 
presumably enables the operculum to be held 
tightly against the aperture whenever the 
snail is retracted into the shell, and thus help 
resist desiccation. The gill filaments are 
fewer, shorter, and thicker relative to body 
size than those of the aquatic species. When- 
ever a snail is out of the water it is likely that 
the palliai cavity will contain air bubbles as 
well as water. Such air bubbles could abut 
against long gill filaments, and cause them to 
fold over, which folding would inhibit the lat- 
eral ciliary activity and hence the flow of water 
through the mantle cavity, and, consequently, 
interfere with respiratory activity. It is less 
likely that the air bubbles would so affect the 
shortened, stubby filaments of F zeidleri. 
Also note that an air bubble held in a damp 
mantle cavity could also assist in maintaining 
a lower body temperature compared to snails 
with a water-filled cavity. This has been found 
to be the case in experiments with land snails 
(Schmidt-Nielsen et al.. 1972). As predicted. 
F zeidleri in all three populations tested had a 
significantly higher tolerance to desiccation 



AUSTRALIAN SPRING HYDROBIIDS 



101 




I mm 





amphibious 



F.zeidleri 



aquatic 




П 



F.z. form В 



F.aquatica 



F billakalina 
F.accepta F.variabilis 



F. cónica 




FIG. 56. Diagrammatic representation of probable relationships of species of Fonscochlea, as well as sizes 
and habitats. This figure is not a cladogram and the distances between branches are not intended to indicate 
degree of taxonomic separation. 



than had the aquatic species tested. Apart 
from F. zeidleri, only F. aquatica from the 
small, harsh Kewson Hill springs survived for 
48 hours in the dry dishes. 

Considering their amphibious habit, it was 
not surprising that, for two of three popula- 
tions, large percentages of F. zeidleri crawled 
out of the water in the submergence prefer- 
ence experiments. While large numbers of 



snails of some of the aquatic species also 
crawled to the tops of the dishes, they did not 
venture beyond the meniscus and remained 
at least partly submerged. 

The differences in results between popula- 
tions of F. zeidleri in the submergence pref- 
erence and light response experiments can 
be explained partly by differences in micro- 
habitat of these populations. Blanche Cup is a 



102 



PONDER, HERSHLER & JENKINS 



large calcrete mound, with a spring pool on 
top and outflow to one side. Fonscochlea zeid- 
leri lives there on moistened rock, and most of 
the individuals are fully exposed to the sun. At 
Finniss Springs, the mound is soft, being 
composed of a sandy substrate, allowing the 
snails to burrow to shallow depths. The pop- 
ulation of F. zeidleri at the Coward Springs 
Railway Bore has been introduced, presum- 
ably recently, from a nearby spring, but F. 
aquatica has not been introduced in the 80- 
90 years that the bore has been flowing. Fons- 
cochlea zeidleri occupies both the amphibi- 
ous and aquatic microhabitats at this locality, 
possibly because F. aquatica, which is similar 
in size to F zeidleri, is absent. The specimens 
on which the experiments were conducted 
were all submerged when collected. These 
microhabitat differences correlated well with 
the results of the submergence preference 
experiments. Over 50% of F zeidleri from 
Blanche Cup and Finniss Springs crawled out 
of the water in these experiments, but only 
16% of the snails from Coward Springs Rail- 
way Bore did so. 

Despite its reduced ctenidium, F zeidleri 
did not show significantly higher mortality or 
reduction in activity than did the aquatic spe- 
cies during the experiments on tolerance to 
deoxygenated water and submergence. In 
the controls of the deoxygenation experiment, 
too, the activity of F zeidleri did not decrease 
faster than that of the aquatic species. This 
fact might suggest that the differences ob- 
served in the ecology of these snails might 
not be due soley to simple physiological lim- 
itations, at least in the ability of F zeidleri to 
tolerate a submerged existence. Certainly the 
existence of an aquatic population at Coward 
Springs Railway Bore would support this ob- 
servation. 

Given the variation seen amongst runs of F 
zeidleri from Coward Springs Railway Bore 
and Finniss Springs, it is difficult to generalize 
as to the response of these snails to light. 
One possible explanation for the variable re- 
sults is that the snails from these populations 
are adapted to avoid light in their natural hab- 
itats, as their microhabitat distribution would 
suggest, but while held in sunlight-exposed 
containers, the snails used in one of the runs 
might have become light adapted and hence 
did not avoid light during the experiment. It is 
also possible that the snails used for the sep- 
arate runs were collected from slightly differ- 
ent habitat types. The Blanche Cup popula- 
tion of F zeidleri lives exposed to the sunlight. 



but only 41% of the snails tested for this pop- 
ulation were in the light zones. 

The two similar-sized forms of F accepta 
differ in the height of the gill filaments; F ac- 
cepta form В has shortened gill filaments, 
similar to those of F zeidleri, whereas F ac- 
cepta form A has tall filaments like those of F 
aquatica. Their habitats are generally similar 
as both species are abundant in shallow wa- 
ters in outflows, but F accepta form A is com- 
monly found in deeper pools as well, whereas 
F accepta form В does not seem to prefer 
this habitat. As might be predicted from their 
morphology, F accepta form В survived bet- 
ter than did F accepta form A during the des- 
iccation experiments. 

Trochidrobia punicea is often found on ex- 
posed surfaces in the water whereas most of 
the other aquatic species seem to prefer 
shaded microhabitats. This difference cor- 
responds well with the fact that T. punicea 
was the only species tested that had a 
strongly positive response to light. The 
aquatic Fonscochlea species, however, are 
also frequently encountered in the open, often 
in large numbers, but were negatively photo- 
tropic in the experiments. Their natural oc- 
curence might be due, in part, to the lack of 
suitable shelter. 

The tolerances of the various species to 
desiccation and salinity might be determined, 
in part, by body size. Desiccation rate is partly 
a function of exposed surface area of tissue. 
When retracted in the shell, a snail can lose 
water either through the shell or through, or 
around the edges of, the operculum. A small 
snail has larger ratios of shell surface area to 
shell volume and shell apertural area to shell 
volume than has a large snail of similar shell 
geometry. Small snails therefore should des- 
iccate more rapidly than large snails. This 
would be accentuated by the fact that, for the 
mound-spring snails, small snails have thinner 
shells than do large snails. The desiccation 
experiments clearly showed that the large- 
sized species, apart from F accepta form A 
(see above), had higher survival in the dry 
dishes than did the small-sized species (f. 
punicea, F variabilis, F. cónica). As noted 
above, these differences obviously are at least 
partly due to divergent adaptation as well. 

Salinity tolerance was also correlated with 
body size among the species tested. The 
large species were fully active in 12%o salt 
water whereas the small species had reduced 
activity in 9%o and no activity in 12%o. It is not 
clear the extent to which body size itself is 



AUSTRALIAN SPRING HYDROBIIDS 



103 



responsible for these differences. Although 
osmotic ргоЫеглз of water loss and salt up- 
take encountered in high-salinity water are 
again dependent on surface area, and hence 
related to body size, physiological adapta- 
tions might be more important. The maximal 
salinity known for the spring groups from 
which the snails were collected for these ex- 
periments is about 4.5%o and about 5.2%o for 
springs known to contain hydrobiids (Kinhill- 
Stearns, 1984). It is noteworthy that the snails 
can tolerate salinities that are twice this value. 
The mound-spring snails are members of a 
large group of freshwater animals that can tol- 
erate salinities of approximately 3-1 0%o 
(Bayly, 1972). As discussed below, their sa- 
linity tolerances do not approach those of the 
inhabitants of athalassic nonmarine waters 
(salinity of 10-300%o, sensu Bayly, 1972). 

It would be of great interest to compare the 
tolerances of mound-spring snails to temper- 
ature, salinity, and water oxygenation with 
fluctuations of these parameters within the 
springs from which the snails came. Unfortu- 
nately such habitat data are not generally 
available, although we do have some data 
concerning temperature. For an 11 -day pe- 
riod during winter, beginning 26/8/83, the 
temperature in one of the largest of the Fin- 
niss Springs varied from 1 1 .0-27.8° С just 
below the springhead, and from 13.0- 
31.0° С in a downstream pool. The air tem- 
perature varied from 3.0-36.0° С during the 
same period. Maximal diurnal fluctuations 
were 16.1° near the springhead and 1 5° in the 
downstream pool, values approaching the 
maximal such fluctuations recorded in desert 
aquatic habitats (Deacon & Minckley, 1974; 
Hershler, 1984). 

An aspect of snail morphology that might 
bear on thermal tolerance is body pigmenta- 
tion. In most of the populations of mound- 
spring snails the degree of pigmentation of 
the head/foot is highly variable but some con- 
spicuous trends have been observed. In gen- 
eral, there is an increase in black pigment 
(melanin?) in populations inhabiting the most 
exposed habitats (e.g., Kewson Hill) where 
shelter (e.g., vegetation) is virtually absent. 
Individuals exposed on hard rock outflows 
tend to be darker than those that can gain 
shelter by burrowing in the sand. This color- 
ation does not appear to be in any way cryptic 
because in many outflows the dark snails are 
very conspicuous against the pale sediment 
or rock. 

Hydrobiids living in caves and other 



phreatic habitats are always unpigmented 
(Vandel, 1965), whereas species living in 
surficial waters are often pigmented, usually 
black. This pattern, together with our obser- 
vations on the pigmentation of mound-spring 
snails, suggests that the degree of pigmenta- 
tion is correlated with exposure to sunlight. As 
the pigment in the mound spring snails is 
largely restricted to the upper visceral mass 
(including the gonad), head/foot and snout, 
areas that are exposed to the sunlight, and 
hence ultraviolet rays, it is likely that such pig- 
ment has a screening function in this group. 

While there are no data available on toler- 
ance to environmental parameters in other 
spring-dwelling hydrobiids, some data are 
available for species in the related family Po- 
matiopsidae, which inhabit ephemeral water 
bodies in arid lands {Coxiella in Australia, 
Tomichia in Africa) (Bayly & Williams, 1966; 
DeDeckker & Geddes, 1980; Davis, 1981), 
and moist amphibious habitats in non-arid re- 
gions {Oncomelania in Asia, Pomatiopsis in 
North America) (van der Schalle & Getz, 
1963). Some information is also available for 
hydrobiids of brackish waters (Hydrobia, Pot- 
amopyrgus) (Newell, 1964; Avens, 1965; 
Winterbourn, 1970; Bayly, 1972; Fenchel, 
1975; Wells, 1978). These various data sets 
can be compared only in a general fashion 
because of differences in experimental de- 
sign and methods. 

Tomichia and Coxiella typically tolerate at 
least several months of desiccation, and a 10 
to 20-fold change in water salinity. These tol- 
erances are considerably broader than those 
of the mound-spring hydrobiids, although the 
desiccation tolerance of Fonscochlea zeidleri 
can approach that of the permanent stream- 
dwelling Tomichia differens (Davis, 1981). 
Such broad tolerances are expected, consid- 
ering the typical habitats of Tomichia and 
Coxiella, ephemeral water bodies subject to 
extreme salinity fluctuations. The mound- 
spring habitat, while often quite shallow, is 
permanent and not subject to great salinity 
fluctuations. Fonscochlea zeidleri does not 
occupy dry habitats, nor do any of the mound- 
spring snails inhabit downstream pools, pos- 
sibly because they might be subject to high 
temperature and salinity fluctuations and 
might even dry up in summer. Pomatiopsis 
and Oncomelania appear to have tempera- 
ture tolerances slightly broader than those of 
the mound spring snails. While Fonscochlea 
zeidleri had no mortality after submersion for 
72 hours, there was significant mortality after 



104 



PONDER, HERSHLER & JENKINS 



this lapse of time in some of the species of 
Oncomelania and Pomatiopsis, perhaps re- 
flecting more specialization for a terrestrial ex- 
istence in the latter group. Most of the species 
of Oncomelania and Pomatiopsis tested ap- 
pear to survive desiccation better than do F. 
zeidleri. again implying more specialization 
for near-terrestrial life. After 48 hours in dry 
dishes, there was mortality in F. zeidleri 
whereas there was 100% survival in all spe- 
cies of Pomatiopsis and Oncomelania. While 
it is unlikely that F zeidleri would survive 30 
or 42 days in dry dishes, it might well survive 
a week and therefore be as tolerant to desic- 
cation as Pomatiopsis cincinnatiensis. 

Hydrobia totteni and the mound-spring hy- 
drobiids were active throughout a similar 
range of temperatures. The Hydrobia and 
Potamopyrgus species tested had high per- 
centages of snails active in a range of salini- 
ties exceeding M%c and as much as 33%o 
(Winterbourn, 1970), whereas the mound 
spring snails were active throughout a salinity 
range of only 12 o/oo. This difference is prob- 
ably a reflection of the estuarine habitat of 
Hydrobia and Potamopyrgus. Fonscochlea 
zeidleri, but not the other mound-spring spe- 
cies, appears to have a higher tolerance to 
desiccation than has Potamopyrgus (an aver- 
age of 73% survival versus 0% survival in dry 
dishes after 48 hours) and possibly Hydrobia 
totteni, but probably not H. ulvae. Obviously 
the estuarine Hydrobia would not be exposed 
to the semi-dry conditions that F zeidleri ex- 
periences for more than the length of a tidal 
cycle. Fish and Fish (1977) have shown that 
the embryonic development of Hydrobia ul- 
vae has an optimal temperature/salinity com- 
bination. At temperature/salinity combinations 
differing from the optimum, hatching was pro- 
longed and mortality increased. It is probable 
that temperature and salinity changes in the 
mound springs have similar effects on the de- 
velopment of the hydrobiid eggs. 

Hydrobiid fauna 

The discussion thus far has concentrated 
on the general problems and theoretical con- 
siderations concerning the fauna as a whole. 
A scenario is suggested within the framework 
proposed above to provide an explanation for 
the differentiation of the taxa. 

The mound springs provide a gradation of 
degrees of isolation from completely isolated, 
through single springs, to local spring groups 
with scattered to interconnected springs. Any 



hypothesis that attempts to explain the evo- 
lution of a taxon only in terms of the details of 
present-day spring disthbution would be inad- 
equate but, as suggested above, the general 
pattern of spring distribution is likely to be 
fairly stable. Obviously any links between, or 
greater isolation of, present groups would 
have been of significance. Other past events 
that might have been important in the devel- 
opment of the present-day taxa are changes 
in climate, drainage patterns and, possibly, 
different ecological and physiological require- 
ments of the hydrobiid fauna, perhaps en- 
abling some of the species to live in other 
water bodies. This last possibility we consider 
unlikely and, consequently, do not develop it 
further. A possible exception is the amphipod, 
Austrochiltonia, which might have invaded the 
springs recently from other water bodies, 
closely similar species being found farther 
south. 

The sympatric species of snails occurring in 
the majority of the springs represent four ra- 
diations. One radiation is that of Trochidrobia 
with two very distinct sympatric species at 
Freeling Springs, one of which is endemic 
and the other, as noted above, also found in 
some of the northern springs to the south of 
Freeling Springs, and two morphologically 
similar, allopatric species in the other springs. 
Fonscochlea (Fig. 56) has radiated in two 
main directions, one toward an amphibious 
species, F zeidleri from which the aquatic 
form, F zeidleri form B, is secondarily de- 
rived, and the other, probably less derived, 
including all the other taxa. These groupings 
are reflected in the subgeneric classification. 
The larger, aquatic group split into two groups 
that radiated in parallel with each other but 
differ markedly in size. The species in these 
two "aquatic" radiations are very similar mor- 
phologically and differ from F zeidleri in a 
number of important characters. It is thus 
likely that the two subgenera in Fonscochlea 
represent an ancient speciation. The species 
distributions within the radiations follow the 
existing pattern of springs closely enough to 
indicate that the speciation events are similar 
in antiquity to the present major spring 
groups. 

There are several patterns of distribution 
demonstrated by the mound-spring hydrobi- 
ids (Figs. 13, 26, 31, 39; Appendix 1, Figs. 
57-63; Table 1). These fall into three main 
groups. The first pattern is restriction to a sin- 
gle spring. This applies to only two infraspe- 
cific forms (F zeidleri form В not included in 



AUSTRALIAN SPRING HYDROBIIDS 



105 



distribution maps but occurring at Big Cad- 
naowie Spring, Fig. 63A; and F. accepta form 
C, Fig. 13). The evolution of both of these 
forms is presumably quite recent as they are 
not greatly differentiated from related taxa. 
They presumably differentiated in isolation af- 
ter dispersal, or might be relictual popula- 
tions. 

The second pattern is restriction to a single 
spring group or complex. Three of the taxa 
occurring at Freeling Springs (Fig. 58), T. in- 
flate, F. aquatica form В and F. variabilis form 
С fall into this category, as do F. accepta form 
В (Fig. 13) and F. variabilis form A (Fig. 26). 
The "taxa" of Fonscochlea in this category 
are considered to be of infraspecific status 
only, i.e. "forms" that might be subspecies, 
and their relatively recent divergence is prob- 
able. Whether these forms represent differen- 
tiation following dispersal or the partial frag- 
mentation of a wider-ranging taxon following 
greater isolation of spring groups, is unclear. 
The two species of Trochidrobia found at 
Freeling Springs are, on the other hand, very 
different from their congeners and no close 
relatives occur elsewhere, facts suggesting a 
considerable period of isolation and continuity 
with the ancient spring habitat of a group dif- 
ferent from the rest of the mound springs. If 
this were indeed the case, the endemic forms 
of Fonscochlea found at Freeling Springs 
would probably be of relatively recent origin 
and derived from the springs to the south. The 
occurrence of T.minuta in some of the north- 
ern springs might be due to recent dispersal 
events. 

The third pattern is occurrence in several 
spring complexes. The majority of taxa, in- 
cluding geographic forms, fall into this cate- 
gory. Fonscochlea accepta form A (Fig. 13) is 
found in Welcome and Davenport Springs 
(Figs. 62, 63B), whereas the species (F. ac- 
cepta) ranges from Welcome to Emerald 
Springs, a range of about 82 km (Figs. 13,61, 
63B). Fonscochlea aquatica form A ranges 
through the Blanche Cup group to the north- 
ern springs south of Freeling Springs (165 km 
range) (Figs. 13, 61, 63B), with a closely re- 
lated form (subspecies?) in Freeling Springs 
(Figs. 13, 58). The amphibious F.ze/d/er/ form 
A has the largest range of any species (270 
km) and is found from Freeling Springs to 
Welcome Springs (Figs. 31, 58, 63B). One of 
the smaller species of Fonscochlea, F. vari- 
abilis, has differentiated into what we are re- 
garding as forms but which might well be 
equivalent to subspecific taxa. One form is 



found in the scattered northern springs, an- 
other even farther north in Freeling Springs 
(Fig. 58), and another in the Blanche Cup 
spring group (Figs. 26, 61). Fonscochlea cón- 
ica, on the other hand, while showing some 
morphological variation, ranges from Beres- 
ford Spring to Welcome Springs (124 km) 
(Figs. 26, 61, 63B). Fonscochlea billakalina 
ranges through the Billa Kalina-Francis 
Swamp-Strangways spring complexes (Figs. 
26, 60, 61). The two species of Trochidrobia 
that occur in the springs south of Freeling 
Springs are distributed differently from the 
FonscocWea species (Table 1 ; Fig. 39). Troch- 
idrobia smithi extends from the northern 
springs to the Billa Kalina complex and the 
Beresford group (Figs. 60, 61). Trochidrobia 
punicea, like F.conica, is found in the middle 
springs and extends to Welcome Springs 
(Fig. 63B) but, unlike that species, is found in 
most of the springs in the area. 

The different distributions of the larger 
aquatic species of Fonscochlea, F accepta 
and F. aquatica, compared with T punicea 
and F.conica suggest that there might have 
been an extinction of the fauna in the middle 
springs followed by the differential invasion of 
F. aquatica form A and F variabilis from the 
northern springs and T.punicea and F.conica 
from the south. It is possible that the original 
population of F aquatica in the area is still 
represented by at least some of the popula- 
tions in the Jersey-Elizabeth-Kewson Hill 
Springs area (Fig. 61), as these appear to 
have differentiated (see discussion in taxo- 
nomic section under F aquaticaiorm A). Fons- 
cochlea variabilis has been successful in es- 
tablishing itself only in the larger springs in the 
Blanche Cup spring group (Fig. 61) whereas 
the very similar F.conica occurs throughout 
the rest of the middle group. This hypothesis 
would also help to explain the lack of notice- 
able differentiation in the species found in the 
middle spring group, with the exception of 
F. variabilis form A. Fossil specimens from the 
middle of the area (from the top of Hamilton 
Hill, Fig. 61) include only F.zeidleri and a spe- 
cies of Trochidrobia that could be either 
Tsmithi or T.punicea, whereas small-sized 
Fonscochlea are abundant on North Beres- 
ford Hill (Fig. 60), a similar fossil mound on 
the northwest edge of the middle springs. 

Absence of fauna 

Several springs and groups of springs in 
the study area did not contain hydrobiids (Ap- 
pendix 1) and many of these same springs 



106 



PONDER, HERSHLER & JENKINS 



also lacked the endemic crustaceans. Individ- 
ual springs in some spring groups also lacked 
the snails and crustaceans whereas neigh- 
bouring springs did not. Water chemistry does 
not appear to explain the absence of fauna in 
many cases (see Kinhill-Stearns, 1984, for 
details of water chemistry of most of the rel- 
evant springs), although poor water quality 
and the lack of running, oxygenated water is 
certainly relevant in some cases. At least two 
springs. Pigeon Hill Spring and Dead Boy 
Spring in the Hermit Hill Spring Complex (Fig. 
62), are closely associated with fauna- 
bearing springs but have sulphate-rich water 
that renders them unsuitable for the mound- 
spring invertebrates. 

Several springs along the southern edge of 
Lake Eyre South (Jacobs, Fred, Smiths, 
Gosse and McLachlan, Fig. 62) are similar in 
water chemistry to the Hermit Hill springs and, 
at least in most cases, have potentially good 
habitat available. The invertebrates do not ap- 
pear to have become established in these 
springs in the recent past as there were no 
traces of snail shells in the spring sediments. 
Flooding in this area results in the submer- 
gence of many of these springs (our observa- 
tions and С Woolard, pers. comm., based on 
the Jan. 1984, floods) and it seems likely that 
even if one of the invertebrates were occa- 
sionally introduced naturally and if a popula- 
tion were established, it would not be suc- 
cessful in the long term. Some of the smaller, 
more isolated springs might never have 
achieved a successful introduction or, per- 
haps, because of their small size, are much 
more susceptible to devastating stock dam- 
age or occasional natural fluctuations in flow 
which might obliterate the habitat. 

Conservation 

The importance of the mound springs as 
unique natural ecosystems that contain a va- 
riety of endemic biota has been addressed 
elsewhere (Casperton, 1979; Mitchell, 1980, 
1985; Harris, 1981; Kinhill-Stearns, 1984; 
Ferguson, 1985; Ponder, 1985, 1986). The 
fragility of these ecosystems, their suscepti- 
bility to damage by livestock and, particularly, 
the real probability of their extinction as a re- 
sult of the extraction of larger amounts of ar- 
tesian water from the aquifers of the Great 
Artesian Basin, would suggest that special 
provisions for their maintenance are required. 
To date none of the springs of the Lake Eyre 
Supergroup that contain endemic fauna is of- 



fered special protection apart from a few 
springs that recently have been fenced to pre- 
vent stock damage. 



ACKNOWLEDGMENTS 

This project would never had started with- 
out the enthusiasm and support of Wolfgang 
Zeidler of the South Australian Museum. We 
thank him for this and for his help and com- 
pany in the field. We also thank D. Winn, D. 
Bushell, J. Ponder, W. Ponder Jnr, С Wool- 
ard and E. Hershler for their assistance in the 
field. 

Mr. D. Beechey assisted greatly with the 
computer analyses and Miss S. Yom-Tov as- 
sisted with the development of the programs 
used in the data processing on the microcom- 
puter. D. Winn, J. Gillespie, M. Fletcher, S. 
Carter, G. Clark, J. Howarth and the late J. 
Kerslake assisted in the preparation of this 
paper. A grant from the South Australian Gov- 
ernment and financial assistance from Roxby 
Management Services (RMS) assisted great- 
ly with this project and provided a salary for 
one of us (R.H.). Considerable additional sup- 
port has been provided by ARGS grants 
(grant numbers D1 781 51 82, D1 851 61 86) and 
support from the Australian Museum Trust. 
RMS also provided logistical support, espe- 
cially the provision of helicopter time and the 
campsite in which the physiological experi- 
ments were performed. The tent used as a 
field laboratory was kindly loaned by the 
South Australian Museum. We should like to 
acknowledge the cooperation and support re- 
ceived from Colin Woolard of RMS, and from 
С Harris and R. Habermehl. We gratefully 
acknowledge the assistance of the staff of the 
drafting section of RMS who prepared the 
maps used in Appendix 1. Ms. S. Dunlop is 
responsible for many of the drawings. 

We gratefully acknowledge the efforts of 
three anonymous reviewers whose construc- 
tive comments markedly improved this paper. 



REFERENCES 

AMBROSE, G. J. & FLINT, R. B. 1981, A regres- 
sive Miocene lake system and silicified strand- 
lines in Northern South Australia: implications for 
regional stratigraphy and silcrete genesis. Jour- 
nal of the Geological Society of Australia, 28: 81 - 
94. 

AVENS, A. C, 1965, Osmotic balance in gastropod 



AUSTRALIAN SPRING HYDROBIIDS 



107 



molluscs-ll. The brackish water gastropod, Hy- 
drobia ulvae Pennant. Comparative Biochemistry 
and Physiology 16: 143-153. 

AYALA, F. J., 1970, Competition, coexistence and 
evolution. Pp. 121-158 in HECHT, M. К. & 
STEERE, W. С, eds.. Essays in Evolution and 
Genetics in honor of Theodosius Dobzhansky. A 
supplement to Evolutionary Biology. North-Hol- 
land, Arnsterdam. 

BARTON, N. H. & CHARLESWORTH, В., 1984, 
Genetic revolutions, founder effects, and specia- 
tion. Annual Review of Ecology and Systematics, 
15: 133-164. 

BAYLY, I. A. E., 1972, Salinity tolerance and os- 
motic behavior of animals in athalassic saline 
and marine hypersaline waters. Annual Review 
of Ecology and Systematics, 3: 233-268. 

BAYLY, I. A. E. & WILLIAMS, W. D., 1966, Chem- 
ical and biological studies on some saline lakes 
in south-east South Australia. Australian Journal 
of Marine and Freshwater Research, 17: 177- 
228. 

BAYLY, I. A. E. & WILLIAMS, W. D., 1973, Inland 
waters and their ecology. Longman Australia Pty 
Limited, Hawthorn, 316 pp. 

BOWLER, J. M., 1967, Recent developments in re- 
constructing late Quaternary environments in 
Australia. Pp. 55-77 in KIRK, R. L. & THORNE, 
A. G., eds.. The Origin of the Australians. Aus- 
tralian Institute for Aboriginal Studies, Canberra/ 
Humanities Press, New Jersey. 

BOWLER, J. M., 1976, Aridity in Australia: age, or- 
igins and expression in aeolian landforms and 
sediments. Earth-Science Reviews, 12: 279- 
310. 

BOWLER, J. M., 1982, Aridity in the late Tertiary 
and Quaternary of Australia. Pp. 35-45 in 
BARKER, W. R. & GREENSLADE, J. M., eds. 
Evolution of the Flora and Fauna of Arid Austra- 
lia. Peacock Publications, Frewville, South Aus- 
tralia. 

BROOKSTEIN, F. L., CHERNOFF, В., ELDER, R. 
L, HUMPHRIES, J. M., SMITH, G. R. & 
STRAUSS, R. E., 1985, Morphometries in evolu- 
tionary biology: the geometry of size and shape 
change, with examples from fishes. Academy of 
Natural Sciences of Philadelphia, Special Publi- 
cation, 15, 277 pp. 

CAIN, A. J., 1977, The efficacy of natural selection 
in wild populations. Pp. 111-133 in GOULDEN, 
C, ed.. The Changing Scene in Natural Sci- 
ences, 1776-1976. Academy of Natural Sci- 
ences of Philadelphia, Special Publication 12. 

CASPERTON, K. C, 1979, Mound springs of South 
Australia. Part 1 — Physical features, history, 
biota and conservation requirements. Depart- 
ment of the Environment and Planning, South 
Australia, unpublished report. 

CHERRILL, A. J. & JAMES, R., 1987, Character 
displacement in Hydrobia. Oecologia (Berlin), 71 : 
618-623. 

CHILTON, C, 1922, A new isopod from central 
Australia belonging to the Phreatoicidae. Trans- 



actions of the Royal Society of South Australia, 
46: 23-33. 

CLIMO, F. M., 1974, Description and affinities of 
the subterranean molluscan fauna of New 
Zealand. New Zealand Journal of Zoology, 1: 
247-284. 

COBB, M. A., 1975, Sampling and measurement of 
mound springs, Great Artesian Basin, South Aus- 
tralia. Progress Report No. 2. Marree, Curdi- 
murka and Billa Kalina sheets. Open File Report, 
Department of Mines and Energy, South Austra- 
lia, Geological Survey, Engineering Division, 
70 + 4 + 2 pp. 

COHAN, F. M., 1984, Can uniform selection retard 
random genetic divergence between isolated 
conspecific populations? Evolution, 38 (3): 495- 
504. 

COLE, G. A., 1968, Desert limnology. Pp. 423-486 
in BROWN, G. W., JR., ed.. Desert Biology Vol- 
ume I. Academic Press, New York. 

CONNELL, J. H., 1970, A predator-prey system in 
the marine intertidal region. I. Balanus glándula 
and several predatory species of Thais. Ecolog- 
ical Monographs, 40: 49-78. 

CONNELL, J. H., 1972, Community interactions on 
marine rocky intertidal shores. Annual Review of 
Ecology and Systematics, 6: 169-192. 

CROW, J. F. & KIMURA, M., 1970, An introduction 
to population genetics theory. Harper & Row, 
New York. 

DAVIS, G. M., 1979, The origin and evolution of the 
gastropod family Pomatiopsidae, with emphasis 
on the Mekong River Triculinae. Academy of Nat- 
ural Sciences of Philadelphia, Monograph 20: 
120 pp. 

DAVIS, G. M., 1981, Different modes of evolution 
and adaptive radiation in the Pomatiopsidae 
(Prosobranchia: Mesogastropoda). Malacologia, 
21 : 209-262. 

DAVIS, G. M., MAZURKIEWICZ, M. & MANDRAC- 
CHIA, M., 1982, Spurwinkia: morphology, sys- 
tematics, and ecology of a new genus of North 
American marshland Hydrobiidae (Mollusca: 
Gastropoda). Proceedings of the Academy of 
Natural Sciences of Philadelphia, 134: 143-177. 

DAYTON, P. K., 1975, Experimental evaluation of 
ecological dominance in a rocky intertidal algal 
community. Ecological Monographs, 45: 137- 
159. 

DEACON, J. E. & MINCKLEY, W. L, 1974, Desert 
fishes. Pp. 385-488 in BROWN, G. W., JR., ed.. 
Desert Biology. Academic Press, New York. 

DeDECKKER, p., 1979, Ostracods from the mound 
springs area between Strangways and Curdi- 
murka. South Australia. Transactions of the 
Royal Society of South Australia, 103: 155- 
168. 

DeDECKKER, P., 1986, What happened to the 
Australian biota 18,000 years ago? Pp. 487-496 
in DeDECKKER, P. & WILLIAMS, W. D., eds.. 
Limnology in Australia. C.S.I.R.O., Melbourne. 

DeDECKKER, P. & GEDDES, M. C, 1980, Sea- 
sonal fauna of ephemeral saline lakes near the 



108 



PONDER, HERSHLER & JENKINS 



Coorong Lagoon, South Australia. Australian 
Journal of Marine and Freshwater Research, 31 : 
677-699. 

EHRLICH, P. R. & RAVEN, P. H., 1969, Differentia- 
tion of populations. Science, 165: 1228-1232. 

FAETH, S. H. & CONNOR, E. F., 1979, Supersat- 
urated and relaxing island faunas: a critique of 
the species-age relationship. Journal of Biogeog- 
raphy, 6: 311-316. 

FENCHEL, T., 1975, Factors determining the dis- 
tribution patterns of rnud snails (Hydrobiidae). 
Oecologia. 20: 1-18. 

FENCHEL, T. & KOFOED, L. H., 1976, Evidence 
for exploitative competition in mud snails (Hydro- 
biidae). Oikos, 27: 367-376. 

FERGUSON, D., 1 985, The mound springs: lens on 
a looming tragedy for Australia's desert lands. 
Habitat, 13 (2): 32-33. 

FISH, J. D. & FISH, S., 1977, The effects of tem- 
perature and salinity on the embryonic develop- 
ment of Hydrobia ulvae (Pennant). Journal of the 
/Marine Biological Association of the United King- 
dom, 57: 213-218. 

FORBES, B. G., 1961, Boring on mound springs 
near Coward Springs, /[fining Review (1959), 
South Australia, 3: 145-149. 

FRYER, G., GREENWOOD, P. H., & PEAKE, J. F., 
1983, Punctuated equilibria, morphological stasis 
and the palaeontological documentation of spe- 
ciation: a biological appraisal of a case history in 
an African lake. Biological Journal of the Linnean 
Society, 20: 195-205. 

GAUSS, G. F., 1934, The struggle for existence. 
Williams & Wilkins, Baltimore. 

GLOVER, С J. M., 1982, Adaptations of fishes in 
arid Australia. Pp. 241-246 in BARKER, W. R. & 
GREENSLADE, P. J. M., eds.. Evolution of the 
flora and fauna of arid Australia. Peacock Publi- 
cations, Frewville, South Australia. 

GLOVER, С J. M. & SIM, T. C, 1978a, Studies on 
Central Australian fishes: a progress report. 
South Australian Naturalist, 52: 35-44. 

GLOVER, С J. M. & SIM, T. C, 1978b, A survey of 
Central Australian ichthyology. Australian Zoolo- 
gist, 19: 245-256. 

HABERMEHL, M. A., 1980, The Great Artesian Ba- 
sin, Australia. Bureau of Mineral Resources, Jour- 
nal of Australian Geology & Geophysics, 5: 9-38. 

HABERMEHL, M. A., 1982, Springs in the Great 
Artesian Basin, Australia — their origin and na- 
ture. Bureau of Mineral Resources, Australia, Re- 
port 235, 50 pp. 

HABERMEHL, M. A. & SEIDEL, G. E., 1979, 
Groundwater resources of the Great Artesian 
Basin. Pp. 71-93 in HALLSWORTH, E. G. & 
WOODCOCK, J. T., eds.. Proceedings of the 
Symposium Land and Water Resources of Aus- 
tralia, Australian Academy of Technological Sci- 
ences, Sydney 30 October-1 November 1978. 
Australian Academy of Technological Sciences, 
Melbourne. 

HARRIS, C, 1981 , Oases in the desert: the mound 
springs of northern South Australia. Royal Geo- 



graphical Society of Australasia — South Australia 
Division, Proceedings, 81 : 26-39. 

HERSHLER, R., 1984, The hydrobiid snails of the 
Cuatro Ciénegas Basin: systematic relationships 
and ecology of a unique fauna. Journal of the 
Arizona-Nevada Academy of Sciences, 19:61- 
76. 

HERSHLER, R., 1985, The systematics and evolu- 
tion of the hydrobiid snails (Gastropoda: Rissoa- 
cea) of the Cuatro Ciénegas Basin, Coahuila, 
Mexico. Malacologia, 26 (1-2): 31-123. 

HERSHLER, R. & DAVIS, G. M., 1980, The mor- 
phology of Hydrobia trúncala (Gastropoda: Hy- 
drobiidae): relevance to systematics of Hydrobia. 
Biological Bulletin, 158: 195-219. 

HERSHLER, R. & LANDYE, J. J., 1988, Arizona 
Hydrobiidae (Prosobranchia: Rissoacea). Smith- 
sonian Contributions to Zoology, 459: 1-63. 

HUBBS, С L & MILLER, R. R., 1948, The zoolog- 
ical evidence: correlation between fish distribu- 
tion and hydrographie history in the desert basins 
of western United States. Pp. 17-166 in The 
Great Basin, with emphasis on glacial and post- 
glacial times. Bulletin of the University of Utah, 
38, Biology Series 10. 

HUBBS, С L., MILLER, R. R. & HUBBS, L. C, 
1974, Hydrographie history and relict fishes of 
the north-central Great Basin. Memoirs of the 
California Academy of Sciences, 7: 1-259. 

HUTCHINSON, G. E., 1959, Homage to Santa 
Rosalia or Why are there so many kinds of ani- 
mals? American Naturalist, 93: 145-159. 

JESSUP, R. W. & NORRIS, R. M., 1971, Cainozoic 
stratigraphy of the Lake Eyre Basin and part of 
the arid region lying to the south. Journal of the 
Geological Society of Australia, 18: 303-331 . 

KEMP, E. M., 1978, Tertiary climatic evolution and 
vegetation history in the southeast Indian Ocean 
region. Palaeogeography, Palaeoclimatology, 
and Palaeoecology, 14: 169-208. 

KINHILL-STEARNS, 1983, Olympic Dam project: 
supplement to the draft environmental impact 
statement. Prepared by Kinhill-Stearns Roger 
Joint Venture for Roxby Management Services 
Pty. Ltd., 103 pp. 

KINHILL-STEARNS, 1984. Olympic Dam Project. 
Supplementary environmental studies, mound 
springs. Prepared by Kinhill Stearns for Roxby 
Management Services Pty. Ltd. 1-4, 2-7, 3-14, 
4-23, 5-16, 6-7, 7-15, 8-6, 9-10, A-6. 

KINHILL-STEARNS ROGER, 1982, Olympic Dam 
project: draft environmental impact statement. 
Prepared by Kinhill-Stearns Roger Joint Venture 
for Roxby Management Services Pty. Ltd., 488 
pp. 

KOTWICKI, v., 1986, Floods of Lake Eyre. Govern- 
ment Printer, South Australia, Adelaide. 

LACK, D., 1947, Darwin's finches. Cambridge Uni- 
versity Press, Cambridge. 

LANDE, R., 1980, Genetic variation and phenotypic 
evolution during allopatric speciation. American 
Naturalist, 116: 463-479. 

LEVINS, R., 1979, Coexistence in a variable 



AUSTRALIAN SPRING HYDROBIIDS 



109 



environment. American Naturalist, 114: 765- 
783. 

LEVINTON, J.S., 1982, The body size-prey size hy- 
pothesis: the adequacy of body size as a vehicle 
for character displacement. Ecology, 63: 869- 
872. 

LUDBROOK, N. H., 1980, Non-marine molluscs 
from dolomitic limestones in the north of South 
Australia. Transactions of the Royal Society of 
South Australia, 104: 83-92. 

MACARTHUR, R. H., 1972, Geographical Ecology. 
Patterns in the distribution of species. Harper & 
Row, New York. 

MACARTHUR, R. H. & WILSON, E. O., 1967, The 
theory of island biogeography. Princeton Univer- 
sity Press, Princeton. 

MARTIN, H. A., 1978, Evolution of the Australian 
flora and vegetation through the Tertiary: evi- 
dence from pollen. Alcheringa, 2: 181-202. 

MAYR, E., 1 942, Systematics and the origin of spe- 
cies. Columbia University Press, New York. 

MAYR, E., 1954, Change of genetic environment 
and evolution. Pp. 157-180 in HUXLEY, J., ed.. 
Evolution as a process. Allen and Unwin, London. 

MCGOWRAN, В., 1979, The Tertiary of Australia: 
foraminiferal overview. t\Aahne t^icropaleontol- 
ogy, 4: 235-264. 

MCMICHAEL, D. F., 1968, Non-marine Mollusca 
from Tertiary rocks in northern Australia. Palae- 
ontological Papers, Bureau of f^ineral Resources 
of Australia Bulletin. 80: 135-156. 

MILLER, R. R., 1950, Speciation in fishes of the 
genera Cyprinodon and Empetrichthys, inhab- 
iting the Death Valley region. Evolution, 4: 155- 
163. 

MINCKLEY. W. L., HENRICKSON, D. A. & BOND, 
С е., 1986, Geography of western North Amer- 
ican freshwater fishes: Description and rela- 
tionships to intracontinental tectonism. Pp. 519- 
613 in HOCUTT, С H. & WILEY, E. O., eds.. 
Zoogeography of North American freshwater 
fishes. John Wiley & Sons, New York. 

MITCHELL, B. D., 1980, Limnology of mound 
springs and temporary pools south and west of 
Lake Eyre. Nature Conservation Society of South 
Australia, Adelaide. Mound Spring Survey Report 
(unpublished), 16 pp. 

MITCHELL, B. D., 1985. Limnology of mound 
springs and temporary pools south and west of 
Lake Eyre. Pp. 51-63 in GREENSLADE, J., 
JOSEPH, L. & REEVES, A., eds.. South Austra- 
lia's mound springs. Nature Conservation Soci- 
ety of South Australia, Adelaide. 

NAIf\^AN, R. J., 1981, An ecosystem overview: 
desert fishes and their habitats. Pp. 493-531 in 
NAIMAN, R. J. & SOLTZ, D. L., eds.. Fishes in 
North American Deserts. John Wiley & Sons, 
New York. 

NAIMAN, R. J. & SOLTZ, D. L., eds., 1981, Fishes 
in North American deserts. John Wiley & Sons, 
New York, 352 pp. 

NEWELL, R., 1964, Some factors controlling the 
upstream distribution of Hydrobia ulvae (Pen- 



nant) (Gastropoda, Prosobranchia). Proceedings 
of the Zoological Society of London, 142: 85- 
106. 
NICHOLLS, G. E., 1943, The Phreatoicoidea. Part 

I. The Amphisopidae. Papers and Proceedings 
of the Royal Society of Tasmania (for 1942): 1- 
145. 

OLLIER, С. D., 1982, The Great Escarpment of 
eastern Australia: tectonic and geomorphic sig- 
nificance. Journal of the Geological Society of 
Australia, 29: 13-24. 

PAINE, R. T., 1966, Food web complexity and spe- 
cies diversity. American Naturalist, 100: 65-75. 

PIMENTEL, R.A. & SMITH, J. D., 1986, BIOSTAT 

II. A multivariate statistical toolbox. Tutorial man- 
ual. Sigma Soft, Placentia, California, pp vii + 21 2. 

PITTMAN, E. F. & DAVID, T W. E., 1903, Irrigation 
geologically considered with special reference to 
the artesian area of New South Wales. Journal of 
the Royal Society of New South Wales, 37: 1 03- 
153. 

PONDER, W. F., 1981, Posticobia norfolkensis 
(Sykes), an apparently-extinct, fresh-water snail 
from Norfolk Island (Gastropoda: Hydrobiidae). 
Proceedings of the Linnean Society of New 
South Wales, 105: 17-21. 

PONDER, W. F., 1982, Hydrobiidae of Lord Howe 
Island (Mollusca: Gastropoda: Prosobranchia). 
Australian Journal of Marine and Freshwater Re- 
search, 33: 89-159. 

PONDER, W. F., 1985, South Australian mound 
springs. Relict faunas in the desert. Australian 
Natural History, 21(8): 352-355. 

PONDER, W. F., 1 986, Mound springs of the Great 
Artesian Basin. Pp. 403-420 in DeDECKKER, P. 
& WILLIAMS, W. D., eds.. Limnology in Australia 
C.SJ.R.O., Melbourne. 

PONDER, W.F., 1988, The truncatelloidean 
( = nssoacean) radiation — a preliminary phylog- 
eny. Symposium on Prosobranch Phytogeny, 
Ninth International Malacological Congress. Ma- 
lacological Review Supplement, 4: 129-166. 

PONDER. W. F. & HERSHLER, R., 1984, Hydro- 
biid studies. Pp. 5.1-5.16 in Olympic Dam 
Project. Supplementary environmental studies — 
mound sphngs. Prepared for Roxby Manage- 
ment Services by Kinhill-Stearns Pty. Ltd. 

PRESTON, F. W., 1962, The canonical distribution 
of commonness and ranty. II. Ecology, 43: 410- 
432. 

RADOMAN, P., 1966, Die Gattungen Pseudamni- 
cola und Horatia. Archiv für f\^olluskenkunde, 95: 
243-253. 

RADOMAN. P., 1983, Hydrobioidea superfamily of 
Prosobranchia (Gastropoda) 1 Systematics. Ser- 
bian Academy of Sciences and Arts l\/1ono- 
graphs. DXLVII. Department of Sciences, 
Beograd, No. 57: 256 pp. 

REES, W. J., 1965, The aerial dispersal of Mol- 
lusca. Presidential address. Proceedings of the 
Malacological Society of London, 36: 269-282. 

ROHLF, F. J. & SOKAL, R. R., 1969, Statistical 
Tables. W. H. Freeman & Co., San Francisco. 



110 



PONDER, HERSHLER & JENKINS 



ROTH, V. L., 1981, Constancy in the size ratios of 
sympatric species. American Naturalist, 118: 
394-404. 

ROUGHGARDEN, J., 1983, Competition and tine- 
ory in community ecology. American Naturalist, 
122: 583-601. 

SCHMIDT-NIELSEN, K., TAYLOR, С R. & 
SHKOLNIK, A., 1972, Desert snails: problems of 
survival. Pp. 1-14 in MALOIY, G. M. O., ed., 
Comparative ptiysiology of desert animals. Sym- 
posia of the Zoological Society of London, Num- 
ber 31, Academic Press, London. 

SIEGEL, S., 1956, Nonparametric statistics for the 
befiavioral sciences. McGraw-Hill Inc., New York, 
312 pp. 

SIMBERLOFF, D. S. & BOECKLEN, W., 1981, 
Santa Rosalia reconsidered: size ratios and com- 
petition. Evolution, 35: 1206-1228. 

SLATKIN, M., 1985, Gene flow in natural popula- 
tions. Annual Review of Ecology and Systemat- 
ics, 16: 393-430. 

SLUYS, R., 1986, First representative of the Order 
Macrostomida in Australia (Platyhelminthes, Mac- 
rostomidae). Records of the South Australia Mu- 
seum, 19: 400-404. 

SMITH, G. R., 1978, Biogeography of intermoun- 
tain fishes. Pp. 17-42 in HARPER, K. T. & RE- 
VEAL, J. L., eds., Intermountain Biogeography: A 
Symposium. Great Basin Naturalist Memoirs 2. 

SOLTZ, D. L. & HIRSHFIELD, M. F., 1981, Genetic 
differentiation of pupfishes (Genus Cyphnodon) 
in the American Southwest. Pp. 291-333 in 
NAIMAN, R. J. & SOLTZ, D. L., eds., Fishes in 
North American Deserts. J. Wiley & Sons, New 
York. 

SOLTZ, D. & NAIMAN, R. J., 1978, The nat- 
ural history of native fishes in the Death Valley 
system. Natural History Museum of Los Ange- 
les County, Science Series, 30: 76 pp. 

STARMÜLNER, F., 1970, Die Mollusken der 
Neukaledonischen Binnengewässer. Cahiers, 
ORSTOM, Serie Hydrobiologie, 4: 3-127. 

SYMON, D. E., 1985, Botanical notes on mound 
springs and bores, and plant species list. Pp. 27- 
48 in GREENSLADE, J., JOSEPH, L. & 
REEVES, A., eds., South Australia's mound 
sphngs. Nature Conservation Society of South 
Australia, Adelaide. 

TAYLOR, D. W., 1966a, A remarkable snail fauna 
from Coahuila, Mexico. The Veliger, 9: 152-228. 

TAYLOR, D. W., 1966b, Summary of North Ameri- 
can Blancan nonmarine mollusks. Malacologia, 
4: 1-171. 

TEMPLETON, A. R., 1 980, The theory of speciation 
via the founder principle. Genetics, 94: 1011- 
1038. 

THOMPSON, F. G., 1968, The aquatic snails of the 
family Hydrobiidae of peninsular Florida. Univer- 
sity of Florida Press, Gainesville, 268 pp. 

THOMSON, R. & BARNETT, S., 1985, Geology, 
geomorphology and hydrogeology. Pp. 3-26 in 
GREENSLADE, J., JOSEPH, L. & REEVES, A., 
eds.. South Australia's mound springs. Nature 



Conservation Society of South Australia, Ade- 
laide. 

THORPE, R. S., 1976, Biometrie analysis of geo- 
graphic variation and racial affinities. Biological 
Reviews, 51: 407-452. 

TURNER, B. J., 1974, Genetic divergence of Death 
Valley pupfish populations: biochemical versus 
morphological evidence. Evolution, 28: 281-294. 

VANDEL, A., 1965, Biospeleology. The Biology of 
Cavernicolous Animals. Pergamon Press, Lon- 
don and New York. 

VAN DER SCHALIE, H. & GETZ, L L, 1963, Com- 
parison of temperature and moisture responses 
of the snail genera Pomatiopsis and Oncomela- 
nia. Ecology, 44: 73-83. 

WATTS, S. H., 1975, Australian landform example 
No. 26. Mound Springs. Australian Geographer, 
13: 52-53. 

WELLS, F. E., 1978, The relationship between en- 
vironmental variables and the density of the mud 
snail Hydrobia totteni in a Nova Scotia salt 
marsh. Journal of Molluscan Studies, 44: 120- 
129. 

WILLIAMS, A. F., 1974, Sampling and measure- 
ment of mound springs. Great Artesian Basin, 
South Australia. Progress Report 1. Dalhousie 
Springs. Open File Report, Department of Mines 
and Energy, South Australia, Geological Survey, 
Engineering Division. 

WILLIAMS, A. F., 1979, Sampling and measure- 
ment of mound springs. Great Artesian Basin, 
South Australia. Progress Report No. 3, Depart- 
ment of Mines and Energy, South Austrailia, Re- 
port Book 79/66. 

WILLIAMS, A. F. & HOLMES, J. W., 1978, A novel 
method of estimating the discharge of water from 
mound springs of the Great Artesian Basin, Cen- 
tral Australia. Journal of Hydrology, 38: 263-272. 

WILLIAMS. E. E., 1972, The origin of faunas. Evo- 
lution of lizard congeners in a complex island 
fauna: a trial analysis. Evolutionary Biology, 6: 
47-89. 

WILLIAMS, W. D., 1981, The Crustacea of Austra- 
lian inland waters. Pp. 1 103-1 138 in KEAST, A., 
ed., Ecological biogeography of Australia. W. 
Junk, The Hague. 

WINTERBOURN, M., 1970, The New Zealand spe- 
cies of Potamopyrgus (Gastropoda: Hydrobi- 
idae). Malacologia, 10: 283-322. 

WOPFNER, H. & TWIDALE, С R., 1967, Geo- 
morphological history of the Lake Eyre Basin. Pp. 
1 1 9-1 43 in JENNINGS, J. N. & MABBUT, J. A., 
eds., Landform studies from Australia and New 
Guinea. Cambridge University Press, London. 

WRIGHT, S., 1931, Evolution in Mendelian popu- 
lations. Genetics, 16: 97-159. 

WRIGHT, S., 1978, Evolution and the genetics of 
populations. Vol. 4. Variation within and among 
natural populations. University of Chicago Press, 
Chicago. 



Revised Ms. accepted 1 June 1988 



AUSTRALIAN SPRING HYDROBIIDS 



111 



APPENDIX 1 



List of stations 



The stations are listed in order of our sta- 
tion numbers and are referred to in the text 
and tables by these numbers. The spring 
name is followed by the latitude and longi- 
tude, the name of the appropriate 1 :250 000 
map sheet and the grid reference for that 
sheet. A reference for Cobb (1975) or 
Williams (1979) is given if appropriate al- 
though the citing of these references does not 
imply that exactly the same spring was sam- 
pled. Additional chemical and flow data are 
given by Kinhill-Stearns (1984) for many of 
the springs listed. The collectors and the date 
of collection are given as are brief details 
about the substations. The numbers in brack- 
ets following the substation data for some of 
the Southern Springs are the numbers allo- 
cated to these springs by Roxby Management 
Services during their survey. Full data about 
each station are not given. Generally informa- 
tion on the dimensions of the spring, the sub- 
strate, habitat, vegetation cover, condition, 
and details of the substations were noted for 
each station. In many temperature and, in 
some, pH, were recorded. 

Abbreviations used; BJ — B. Jenkins, CW — С 
Woolard, DB— D. Bushell, DW— D. Winn, EH— E. 
Hershler, helic — helicopter, RH — R. Hershler, 
WP— W. Ponder, WPj— W. Ponder Jnr, WZ— W. 
Zeidler. 

002 ( = 41) Welcome Springs-northern one. 
29°40.09'S, 137°44'E. Curdimurka 594324 (Cobb, 
1975:1). Coll. WZ, 10 Sept.81. General. 

003 ( = 42) Welcome Springs-southwest. 
29°40.77'S, 137°49.75'E. Curdimurka 594324. 
(Cobb, 1975:1). Coll. WZ, 11 Sept.81. General. 

004 ( = 49A) Davenport Springs. 29°40.09'S, 
137°35.31'E. Curdimurka 567325. (Cobb, 1975: 
11). Coll. WP, WZand BJ. 13 May 81. General. 

005 (49B) Davenport Springs. 29°40.09'S, 
137°35.56'E. Curdimurka 567325. (Cobb, 1975: 
11). Coll. WP, WZ, WPj and BJ, 13 May 81. Gen- 
eral. 

006 ( = 71) Mount Hamilton Homestead ruins. 
29°29.71'S, 136°53.95'E. Curdimurka 496346. 
Coll. WP, WZ, BJ and WPj, 1 6 May 81 . Pool at top. 

007 ( = 69) Strangways Spring, E. of Blanche Cup. 
29°29.06'S, 136°53.64'E. Curdimurka 495357. 
Coll. WP, WZ, BJ and WPj, 16 May 81. Upper part 
of outflow. 

008-012 ( = 65) Blanche Cup Spring. 29"27.35'S, 
136°51.57'E. Curdimurka 491351 (Cobb, 1975:51). 
Coll. WP, WZ, BJ and WPj, 1 5 May 81 . 008: In pool. 



009: Upper outflow. 010: Middle outflow. 01 1 : Out- 
flow at base of mound. 012: Near end of outflow. 

013-017 ( = 66) The Bubbler Spring. 29°26.8'S, 
136°51.4'E. Curdimurka 492352 (Cobb, 1975:49). 
Coll. WP, WZ, BJ and WPj, 1 5 May 81.013: Upper 
outflow, just below pool. 014: Lower outflow. 015: 
Swampy pool at base. 016: In seep on edge of pool 
at top. 017: On sedges and algae in pool at top. 

018 ( = 63) Coward Springs Railway Bore. 
29°24.21'S, 136°48.89'E. Curdimurka 357486. 
Coll. WP, WZ, BJ and WPj, 15 May 81. General. 

019-022 ( = 64) Coward Springs. 29°24.78'S, 
136°47.28'E. Curdimurka 484357 (Cobb, 1975:56). 
Coll. WP, WZ, BJ and WPj, 1 5 May 81 . 01 9: Pool at 
top. 020: Upper outflow. 021 : Outflow near base of 
mound. 022: Lower outflow. 

023 ( = 64E) Coward Springs. 29°24.78'S, 
136°47.28'E. Curdimurka 484357 (Cobb, 1975:56). 
Coll. WP, WZ, BJ and WPj, 15 May 81. Separate 
seepage at top of mound. 

024 ( = 52) Elizabeth Springs. 29°21.36'S, 
1 36°46.30'E. Curdimurka 482363. (Cobb, 1 975:59). 
Coll. WP, WZ, BJ and WPj, 14 May 81. General. 

025 ( = 60) Jersey Springs. 29°20.81'S, 
136°45.37'E. Curdimurka 481364. Coll. WP, WZ, 
BJ and WPj, 15 May 81. General. 

026-027 ( = 53) Old Billa Kalina Spring. 
29°27.66'S, 136°29.75'E. Billa Kalina 453350. Coll. 
WP, WZ, BJ and WPj, 14 May 81. 026: Top of 
outflow. 027: Lower outflow. 

028 ( = 75) Beresford Spring (N. side of Beresford 
Hill). 29°16.0'S, 136°39.7'E. Curdimurka 471374 
(Cobb, 1 975:65). Coll. WZ, 1 Sept. 81 . Near top of 
outflow. 

029-030 ( = 76) Strangways Springs (near Irrapa- 
tana). 29°09.9'S, 136°33.1'E. Curdimurka 458387 
(Cobb, 1975:68, Williams, 1979:64). Coll. WZ, 5 
Sept. 81. 029: Near top of outflow. 030: Outflow. 

031-033 ( = 77) The Fountain Spring. 28°21.1'E, 
136°17.0'E. Warrina 431485 (Williams, 1979:14). 
Coll. WZ, 9 Sept. 81 . 031 : Top of outflow. 032: Near 
outflow of top pond. 033: Near bottom of outflow. 

034 ( = 78) Big Perry Springs (West). 28°20.4'S, 
136°20.6'E. Warrina 438487 (Williams, 1979:16). 
Coll. WZ, 9 Sept. 81 . Top and middle of outflow. 

035-037 ( = 79) Twelve Mile Spring. 28°18.5'S, 
136°15.4'E. Warrina 427490 (Williams, 1979:13). 
Coll. WZ, 6 Sept. 81 . 035: Top of spring. 036: Base 
of mound. 037: Near top of outflow. 

038-040 ( = 80) Outside Springs (middle one). 
28''16'S, 136°12.5'E. Warrina 422496 (Williams, 
1979:8). Coll. WZ, 6 Sept. 81. 038: Top of outflow. 
039: Middle of outflow. 040: Near bottom of out- 
flow. 

041 ( = 81) Outside Springs (southern one). 
28°17'S, 136°12.7'E. Warrina 422495 (Williams, 
1979:8). Coll. WZ, 6 Sept. 81. Middle of outflow. 



112 



PONDER, HERSHLER & JENKINS 



042-044 ( = 82) Freeling Springs (southernmost). 
28°4.3'S, 135°54.4'E. Warrina 390518 (Williams, 
1979:27). Coll. WZ, 7 Sept. 81 . 042: Top of outflow. 
043: Middle of outflow. 044: Near bottom of outflow. 

045-046 ( = 83) Freeling Springs (one crossing 
track). 28°4.3'S, 135°54.4'E. Warrina 390518 
(Williams, 1979:27). Coll. WZ, 7 Sept. 81. 045: 
Near top of outflow. 046: Near bottom of outflow. 

047 Lodden ( = Louden) Spring. 28°35.2'S, 
136°24.0'E. Warrina 443456 (Williams, 1979:48). 
Coll. WZ, Sept. 81. Spring dry. 

048 Melon Spring. 28°15.3'S, 136°4.9'E. Warrina 
408496. Coll. WZ, Sept. 81. General. 

049 Levi Spring. 28°22.9'S, 136°09'E. Warrina 
416482. Coll. WZ, Sept. 81. General. 

050 Spring in creek bed NE of Nilpinna Springs. 
28°14'S, 135°43'E. Warrina 367503. Coll. WZ, 
Sept. 81 . General. 

051 The Vaughan Spring. 28°17.4'S, 136°10.1'E. 
Warrina 426493 (Williams, 1979:12). Coll. WZ, 
Sept. 81 . General. 

659 Unnamed spring. 27°47.1'S, 135°39.9'E. 
Oodnadatta 364553 (Williams, 1979:49). Coll. WP 
and WZ, 1 June 83. General. 

660 Okenden Spring and Bore. 27°50.8'S, 
135°44.0'E. Oodnadatta 372547 (Williams, 1979: 
54). Coll. WP and WZ. 1 June 83. General. 

661 Big Cadnaowie Spring. 27°51.5'S, 135°40.1'E. 
Oodnadatta 364545 (Williams, 1979:53). Coll. WP 
and WZ. 1 June 83. Outflow and top pool. 

662 Little Cadnaowie Spring. 27°47.4'S, 
135°56.5'E. Oodnadatta 367554 (Williams, 1979: 
51). Coll. WP and WZ, June 83. General. 

663 Freeling Springs, main spring (southernmost). 
28°4.3'S, 135°54.4'E. Warrina 390518 (Williams, 
1979:27). Coll. WP and WZ, 2 June 83. Ouantita- 
tive samples taken. 

664 Freeling Springs, "Well Spring". 28°4.1'S, 
135°54.3'E. Warrina 389518 (Williams, 1979:27). 
Coll. WP and WZ, 2 June 83. A1: Pool, mud and 
weed on bottom. A2: Pool on calcrete near water 
surface. B: Beginning of outflow. C: ca.50m down 
outflow. 

665 Freeling Springs, near "Well Spring". 28°4.2'S, 
135°54.5'E. Warrina 390518 (Williams, 1979:27). 
Coll. WP and WZ, 2 June 83. A:Head of spring. 
B:21m down outflow. C:50m down outflow. 

666 Unnamed spring, ca.2.5km N. of Freeling 
Springs. 28°2.0'S, 135°44.1'E. Warrina 389521 
(Williams, 1979:29). Coll. WP and WZ, 3 June 83. 
General. In Peake Creek bed. 

667 Tidiamurkuna waterhole-spring. 28°2.3'S, 
135°48.9'E. Warrina 380523. Coll. WP and WZ, 3 
June 83. General. 

668 Melon and Milne springs. 28°15.3'S, 
136°4.9'E. Warrina 408496 (Williams, 1979:7). 
Coll. WP and WZ, 4 June 83. General. 



670 Hawker Springs, 4.1km from N. turnoff on N. 
side of track. 28°24.4'S, 136°11.0'E. Warrina 
419478 (Williams, 1979:20). Coll. WP and WZ, 4 
June 83. A:Head of spring. B: Beginning of outflow. 
C:Outflow. 

671 Hawker Springs, 6.3km from N. turnoff, N.E. of 
track. 28°25.3'S, 136°11.3'E. Warrina 421484 
(Williams, 1979:20). Coll. WP and WZ, 4 June 83. 
General. 

672 Hawker Springs, 7.3km from N. turnoff, W. of 
track. 28°26.0'S, 136°11.6'E. Warrina 421475 
(Williams, 1979:20). Coll. WP and WZ, 4 June 83. 
A:Head of spring. B:12m down outflow. C:40m 
down outflow. D:Outflow of subsidiary spring. 

673 Hawker Springs. 8.3km S.E. from N. turnoff to 
springs. 28°26.8'S, 136°11.6'E. Warrina 421474 
(Williams, 1979:20). Coll. WP and WZ, 4 June 83. 
General. 

674 Spring Hill Springs, S. side of Spring Hill. 
28°25.3'S, 136°9'E. Warrina 416476 (Williams, 
1979:23). Coll. WP and WZ, 5 June 83. General. 

675 Edith Spring. 28°28.0'S, 136°5.4'E. Warrina 
409472 (Williams, 1979:24). Coll. WP and WZ, 5 
June 83. General. 

676 Taitón Springs. 28°31.6'S, 136°5.7'E. Warrina 
410463 (Williams, 1979:46). Coll. WP and WZ, 5 
June 83, General. 

677 Brinkley Springs. 28°30.4'S, 136°16.9'E. War- 
rina 432466 (Williams, 1979:44). Coll. WP and WZ, 
5 June 83. General. 

678 Strangways Springs (near Irrapatana), 
ca.lOOm S.W. of ruins. 29°9.88'S, 136°33.09'E. 
Warrina 458386 (Cobb, 1975:68, Williams, 1979: 
64). Coll. WP and WZ, 6 June 83. A:Upper outflow. 
B:Pool on top of mound. 

679 Strangways Springs (near Irrapatana), 
ca.200m S.W. of ruins. 29°9.79'S, 136°33.09'E. 
Warrina 458386 (Cobb, 1975:68, Williams, 1979: 
64). Coll. WP and WZ, 6 June 83. AI :Pool at top of 
mound on edges out of water. A2:Pool and upper 
outflow. A3: Lower outflow. 

680 Strangways Springs (near Irrapatana), 
ca.130m N.W. of ruins. 29°9.98'S, 136°32.87'E. 
Warrina 458386 (Cobb, 1975:68, Williams, 1979: 
64). Coll. WP and WZ, 6 June 83. General. 

681 Warburton Spring. 29°16.68'S, 136°40.31'E. 
Curdimurka 471373 (Cobb, 1975:65). Coll. WP and 
WZ, 7 June 83. A: Pool at top, AI on edge, A2 in 
pool. B:Upper outflow, 81 from edges, B2 from wa- 
ter. C: Lower outflow. 

682 Unnamed spring near Warburton Spring. 
29°16.57'S, 136°40.19'E. Curdimurka 472373. 
Coll. WP and WZ, 7 June 83. General. 

683 Jersey Springs. 29°20.81 'S, 136°45.37'E. Cur- 
dimurka 481364, Coll. WP and WZ, 7 June 83. A: 
Beginning of seepage. B:Outflow. 



AUSTRALIAN SPRING HYDROBIIDS 



113 



684 Coward Springs Railway Bore. 29°24.21'S, 
136°48.89'E. Curdimurka 357486. Coll. WP and 
WZ, 7 June 83. Exit from pool and upper outflow. 

685 Blanche Cup Spring. 29°27.35'S, 1 36°51 .57'S. 
Curdimurka 491 351 (Cobb, 1 975:51 ). Coll. WP and 
WZ, 7 June 83. Quantitative samples. Also quanti- 
tatively sampled 29 Jan. 84. 

686 Priscilla Spring. 29°34.30'S, 137°13.52'E. Cur- 
dimurka 528336 (Cobb, 1975:41). Coll. WP and 
WZ, 8 June 83. General. 

687 Venable Spring/bore. 29°40.78'S, 
137°22.03'E. Curdimurka 544323 (Cobb, 1975:28). 
Coll. WP and WZ. 9 June 83. General. Low mound 
with bore. 

688 Beatrice Spring/bore. 29°37.46'S, 
137°21.95'E. Curdimurka 544330 (Cobb, 1975:25). 
Coll. WP and WZ, 9 June 83. Bore and large mound 
with seepages. 

689 Dead Boy Spring. 29°36.08'S, 137°24.44'E. 
Curdimurka 547333. Coll. WP and WZ, 9 June 83. 
General. Very small spring in large abiotic spring 
(HDB005). 

690A Finniss Swamp West. 29°35.68'S, 
1 37°24.66'E. Curdimurka 549333 (Cobb, 1 975:1 9). 
Coll. WP and WZ, by helic, 9 June 83. Small 
spring— general (HWF039). 

690B Finniss Swamp West. 29°35.68'S, 
137°24.66'E. Curdimurka 549333 (Cobb, 1975:19). 
Coll. WP and WZ, by helic, 9 June 83. Small 
spring — general (HWF042). 

690C Finniss Swamp West. 29°35.68'S, 
137°24.66'E. Curdimurka 549333 (Cobb, 1975:19). 
Coll. WP and WZ, by helic, 9 June 83. Small 
spring — general (HWF041). 

691 A Finniss Swamp West. 29°35.68'S, 
137°24.66'E. Curdimurka 549333 (Cobb, 1975:19). 
Coll. WP and WZ, by helic, 9 June 83. A:Head of 
spring in swampy, shallow pool. B:Upper outflow. 
C:Upper part of middle outflow. D:Lower outflow 
(HWF031). 

692A Bopeechee (or Zeke) Springs. 29°36.49'S, 
1 37°23.1 5'E. Curdimurka 547332 (Cobb, 1 975:21 ). 
Coll. WP and WZ, 9 June 83. Very small mound and 
seepage, ca.40m S.S.W. of 692B (HBO003). 

692B Bopeechee (or Zeke) Springs. 29°36.49'S, 
1 37°23.1 5'E. Curdimurka 547332 (Cobb, 1 975:21 ). 
Coll. WP and WZ, 9 June 83. General (HBO002). 

693 Old Finniss Springs. 29°34.97'S, 137°26.79'E. 
Curdimurka 553336. Coll. WP and WZ, by helic, 12 
June 83. Quantitative samples. Also sampled in 
Aug. 1983 (non-quantitative) and Jan. 1984 (quanti- 
tative) (HHOF092). 

694 Old Finniss Springs. 29°34.97'S, 137°26.79'E. 
Curdimurka 553336. Coll. WP and WZ, by helic, 
10 June 83. General. A: Spring 13 x 24m 
(HOF089). B:Spring 15x37m (HOF088). C:Spring 
8xl7m (HOF087). Three small springs grouped 
together. 



695 Smith Springs. 29°30.37'S, 137°21.42'E. Cur- 
dimurka 544344 (Cobb, 1975:31). Coll. WP and 
WZ, by helic, 11 June 83. General examination of 
all springs. 

696 Gosse Springs. 29°28.0'S, 137°20.6'E. Curdi- 
murka 542349 (Cobb, 1975:34). Coll. WP and WZ, 
by helic, 11 June 83. General (3 separate springs 
examined). Main spring also examined 29 Jan. 84. 

697 McLachian Spnngs. 29°27.8'S, 137°19.0'E. 
Curdimurka 539349 (Cobb, 1975:37). Coll. WP and 
WZ, by helic, 11 June 83. General (a large sand 
mound). 

698-9 Unnamed springs near McLachian Springs. 
29°28'S, 137°19.1'E. Curdimurka 540348. Coil. 
WP and WZ, by helic, 11 June 83. General. 

700 Unnamed spring 1.5km S.E. of McLachian 
Springs. 29°28'S, 137°19.1 'E. Curdimurka 540348. 
Coll. WP and WZ, by helic, 1 1 June 83. General- 
several small seeps. 

701 Unnamed spring in W. Lake Eyre South. 
29°19.9'S, 137°10.9'E. Curdimurka 526366. Coll. 
WP and WZ, by helic, 11 June 83. General. 

702 Unnamed spring in S. end of Lake Eyre South. 
29°21.60'S, 137°16.54'E. Curdimurka 535363. 
Coll. WP and WZ., by helic, 11 June 83. General. 

703 Emerald Spring. 29°23.14'S, 137°3.70'E. Cur- 
dimurka 513359 (Cobb, 1975:45, Williams, 1979: 
61). Coll. WP and WZ, by helic, 11 June 83. A: 
Upper outflow. B:Middle outflow. 

704 Fred Springs West. 29°31.08'S, 137°16.85'E. 
Curdimurka 536344 (Cobb, 1975:38). Coll. WP and 
WZ, by helic, 1 1 June 83. General. Very little sur- 
face water. Fred Springs East was also visited but 
no station number was allocated. 

710 Old Finniss Springs (nearest ruin). 29°35.08'S, 
137°27.0'E. Curdimurka 553336. Coll. WP and WZ, 
by helic, 12 June 83. General (one of several sim- 
ilar mounds examined) (HOF081). 

71 1 A Hermit Hill Springs. 29°34.32'S, 137°25.56'E. 
Curdimurka 551336 (Cobb, 1975:16). Coll. WP and 
WZ, by helic, 1 2 June 83. General (HHS1 72). Sev- 
eral similar mounds examined (B-V). 

71 1W Hermit Hill Springs. 29°34.24'S, 
137°25.86'E. Curdimurka 552336 (Cobb, 1975:16). 
Coll. WP and WZ, by helic, 12 June 83. General 
(HHS149). Firmer sediment in outflow than 711 A. 

712 Hermit Hill Springs (E.group). 29°34.24'S, 
137°25.86'E. Curdimurka 552336 (Cobb, 1975:16). 
Coll. WP and WZ, by helic, 12 June 83. General 
(HHS064-077). Group of 3 small spnngs with com- 
mon outflow. 

714 Cardajalburrana Spring. 28°11.1'S, 
135°33.1'E. Warrina 352505 (Williams, 1979:31). 
Coll. WP and WZ, by helic, 13 June 83. General. 

715 Weedina Springs. 28°23.6'S, 135°38.6'E. War- 
rina 362480 (Williams, 1979:37). Coll. WP and WZ, 
13 June, 83. General. 



114 



PONDER, HERSHLER & JENKINS 



716 Eurilyana Spring, on S. side of Lake Cadibar- 
rawirra. 28°55.5'S, 135°26.9'E. Warrina 341416 
(Williams, 1979:43). Coll. WP and WZ, 13 June, 83. 
General. 

717 Loyd Bore, Francis Swamp. 29°7.3'S, 
136°17.7'E. Warrina 432393 (Cobb, 1975:1, 
Williams, 1979:58). Coll. WP and WZ, 13 June, 83. 
A:At point of outlet. B:General swamp around main 
outlet. C:ln outflow draining out of main part of 
spring. 

718 Anna Springs East (?bore). 29°31.90'S, 
136°59.32'E. Curdimurka 506345. Coll. WP and 
WZ, by helic, 13 June 83. General. 

719 North West Springs. 29°33.51'S, 137°24.11'E. 
Curdimurka 548337. Coll. WP and WZ, by helic, 13 
June 83. General. A-C:3 small springs in S.E. of 
group (HNW005,007,010). 

71 9D North West Springs. 29°33.51'S, 
137°24.11'E. Curdimurka 548337. Coll. WP and 
WZ, by helic, 13 June 83. General. Small spring in 
N. of group (HNW003). 

720 Francis Swamp, one of springs in middle part 
of swamp. 29°8.6'S, 136°17.3'E. Billa Kalina 433 
388 (Cobb, 1975:1). Coll. WP and WZ, by helic, 14 
June 83. A:ln middle of spring outlet area. B:ln 
swamp surrounding outlet. C:ln outflow. 

721 Francis Swamp, springs near south end. 
29°10'S, 136°19.2'E. Billa Kalina 434386 (Cobb, 
1975:1). Coll. WP and WZ, by helic, 14 June 83. 
Three springs samples (A-C). 

722 Margaret Spring. 29°13.2'S, 136°20.8'E. Billa 
Kalina 436739. Coll. WP and WZ, 1 4 June 83. Gen- 
eral. 

723 Fenced Spring (Billa Kalina). 29°29.1'S, 
136°26.9'E. Billa Kalina 447347. Coll. WP and WZ, 
by helic, 14 June 83. A:Pool at top. Mostly open 
water. B:Upper outflow. C:Middle outflow. D:Edge 
of outflow. 

730 Finniss Swamp West, near main road. 
29°35.92'S, 137°24.57'E. Curdimurka 548333. 
Coll. RH and EH, 27 Aug. 83. General collection 
(HWF048). 

731 Old Woman Springs. 29°35.41 'S, 137°27.35'E. 
Curdimurka 554334. Coll. WP and BJ, 30 Aug. 83. 
General (HOW024). Small spring reactivated after 
seismic work in area. 

732A Old Woman Springs. 29°35.46'S, 
137°27.35'E. Curdimurka 554334. Coll. WP and 
BJ, 30 Aug. 83. General-small mound near 732B 
(HOW015). 

732B Old Woman Springs. 29°35.46'S, 
137°27.35'E. Curdimurka 554334. Coll. WP and 
BJ, 30 Aug. 83. General (HOW013). 

733 Old Woman Springs, main spring. 29°35.57'S, 
137°27.28'E. Curdimurka 554334. Coll. WP and 
BJ, 30 Aug. 83. A:Top pool. B:Beginning of outflow. 
C:Upper part of outflow. D:Lower outflow. E:Seep- 
age at head of pool (HOW009). 



734 Old Finniss Mound Spring. 29°35.00'S, 
137°28.18'E. Curdimurka 556335. Coll. WP and 
BJ, 30 Aug. 83. General (HOF094). 

735 Sulphuric Springs. 29°36.51'S, 137°24.20'E. 
Curdimurka 548333. Coll. WP and BJ, 30 Aug. 83. 
General (HSS016). 

736 Sulphuric Springs. 29°36.68'S, 137°24.20'E. 
Curdimurka 558332. Coll. WP and BJ, 30 Aug. 83. 
General (HSS014). 

737 Sulphuric Springs. 29°36.61'S, 137°24.01'E. 
Curdimurka 547332. Coll. WP and BJ, 30 Aug. 83. 
General (HSS006). 

738 Jacobs Spring. 29°29.38'S, 137°8.95'E. Curdi- 
murka 523347 (Cobb, 1975:44). Coll. WP, RH and 
DB, 31 Aug. 83. General. 

739 Blanche Cup Spring. 29°27.35'S, 136°51.57'E. 
Curdimurka 491 351 (Cobb, 1 975:51 ). Coll. WP, RH 
and DB, 31 Aug. 83. Transect of pool. 

740 Kewson Hill Springs. 29°22.31'S, 136°47.13'E. 
Curdimurka 484362. Coll. WP, RH and DB, 31 Aug. 
83. General. On side of very large mound. 

741 Kewson Hill Springs. 29°22.28'S, 136°47.16'E. 
Curdimurka 484362. Coll. WP, RH and DB, 31 Aug. 
83. Upper 10m of outflow. 

742 Kewson Hill Springs. 29°22.23'S, 136°47.16'E. 
Curdimurka 484362. Coll. WP, RH and DB, 31 Aug. 
83. A:Upper outflow. B:Lower outflow. 

742B Kewson Hill Springs. 29°22.23'S, 
136°47.16'E. Curdimurka 484362. Coll. WP, RH 
and DB, 31 Aug. 83. Lower outflow. 

743 Coward Springs Railway Bore. 29°24.21'S, 
136°48.89'E. Curdimurka 357486. Coll. WP, RH 
and DB, 31 Aug. 83. Beginning of outflow. A:Pool at 
bore on edge. B:On surface of damp mud near 
large clump of bullrushes. C:ln water near large 
clump of bullrushes. 

744 Little Bubbler Spring. 29°27.35'S, 136°51.91'E. 
Curdimurka 492351 (Cobb, 1975:51). Coll. WP, BJ 
and CW, 1 Sept. 83. A:Beginning of outflow. B:34m 
down outflow. C:Lower outflow. 

745 Strangways Spring E. of Bubbler group. 
29°29.06'S, 136°53.64'E. Curdimurka 495357. 
Coll. WP, BJ and CW, 1 Sept. 83. A:Upper outflow. 
B:Middle outflow. 

746 Horse Springs West. 29°29.50'S, 136°54.80'E. 
Curdimurka 497347 (Cobb, 1975:48). Coll. WP, BJ 
and CW, 1 Sept. 83. A:General — mostly upper out- 
flow. B:ln solution hole on side of mound. 

747 Horse Springs East. 29°29.50'S, 136°55.25'E. 
Curdimurka 498347 (Cobb, 1975:48). Coll. WP, BJ 
and CW, 1 Sept. 83. A:Top pool, mostly under 
stones. B:Outflow. 

748 Horse Springs East. 29°29.58'S, 136°55.25'E. 
Curdimurka 498347 (Cobb, 1975:48). Coll. WP, BJ 
and CW, 1 Sept. 83. A:Crater-like pool at top. B: 
Outflow. C:Outiflow at base of mound. 



AUSTRALIAN SPRING HYDROBIIDS 



115 



749 Spring at Mt. Hamilton ruins. 29°29.71'S, 
136°53.95'E. Curdimurka 496346. Coll. WP, BJ 
and CW, 1 Sept. 83. Pool at top. 

750 Anna Springs West. 29°32.04'S, 136°59.26'E. 
Curdimurka 506345. Coll. WP, BJ and CW, 1 Sept. 
83. Pool. 

751 Anna Spring/bore East. 29°31.90'S, 
136°59.32'E. Curdimurka 506345 (Cobb, 1975:47). 
Coll. WP, BJ and CW, 1 Sept. 83. General. 

752 Main bore/spring, Davenport Springs. 
29°40.09'S, 137°35.31'E. Curdimurka 567325 
(Cobb, 1975:11-1). Coll. WP, RH and DW, 2 Sept. 
83. A:15m down outflow. B:25m down outflow. C: 
60m down outflow. 

753 Davenport Springs. 29°40.09'S, 137°35.56'E. 
Curdimurka 567325 (Cobb, 1975:11-1). Coll. WP, 
RH and DW, 2 Sept. 83. A:Head and uppermost 
outflow. B:Lower outflow. 

754 Welcome Springs. 29°40.09'S, 137°49.44'E. 
Curdimurka 594324 (Cobb, 1975:1-3). Coll. WP 
RH and DW, 2 Sept. 83. A:Uppermost outflow. В 
20m down outflow. C:Pool 25m down outflow. D 
80m down outflow. 

755 Welcome Springs. 29°40.42'S, 137°49.75'E. 
Curdimurka 594323 (Cobb, 1975:1-3). Coll. WP, 
RH and DW, 2 Sept. 83. A:Head of spring. B:20m 
down outflow. C:50m down outflow. D:12m down 
outflow. 

756 Welcome Springs. 29°40.77'S, 137°49.75'E. 
Curdimurka 594323 (Cobb, 1975:1-3). Coll. WP, 
RH and DW, 2 Sept. 83. A:Pool 4m from beginning. 
B:Upper outflow. C:Lower outflow. 

757 Wangianna Spring/well/bore. 29°40.55'S, 
137°42.65'E. Curdimurka 581323 (Cobb, 1975:8). 
Coll. WP, RH and DW, 2 Sept. 83. General. 

758 Welcome Bore/spring. 29°21.02'S, 
136°37.38'E. Curdimurka 465364. Coll. WP, RH 
and DB, 3 Sept. 83. General. 

759 Spring at Old Billa Kalina ruin. 29°27.66'S, 
136°29.75'E. Billa Kalina 453350. Coll. WP, RH 
and DB, 3 Sept. 83. A:Pool at top. B:Upper outflow. 
C:Lower outflow. 

760 Spring near Old Billa Kalina ruin. 29°27.66'S, 
136°29.75'E. Billa Kalina 453350. Coll. WP, RH 
and DB, 3 Sept. 83. A:Pool at top. B:Upper outflow. 

761 Billa Kalina, 1.8km S. of ruins. 29°27.98'S, 
136°28.40'E. Billa Kalina 451349. Coll. WP, RH 
and DB, 3 Sept. 83. A:Seep at head. B:Pool at top. 
C:Outflow. 

762 Billa Kalina Springs. 29°27.98'S, 136°28.40'E. 
Billa Kalina 451349. Coll. WP, RH, DB, 3 Sept. 83. 
A:Pool at top. B:Upper outflow. 

763 Billa Kalina Springs. 29°28.53'S, 136°27.22'E. 
Billa Kalina 848348. Coll. WP, RH and DB, 3 Sept. 
83. A:Upper outflow. B:Lower outflow. 

764 Coward Springs. 29°24.78'S, 136°47.28'E. 
Curdimurka 484357 (Cobb, 1975:56). Coll. WP, RH 
and DW, 5 Sept. 83. A:Small seepage on top of 



mound. B:Beginning of outflow. C:Outflow at base 
of mound. 

765 Spring near W. side of Kewson Hill. 
29°22.17'S, 136°46.79'E. Curdimurka 483362. 
Coll. WP, RH and DW, 5 Sept. 83. General. 

766 E. side of Elizabeth Springs mound. 
29°21.36'S, 136°46.30'E. Curdimurka 482363 
(Cobb,1975:59).Coll. WP, RH and DW, 5 Sept. 83. 
A:Head of spring. B:Outflow from top seep. C:Sec- 
ond spring on outflow. D:Outflow, terrace area. E: 
Outflow, lower end of terraces. F:On steep side of 
hill in outflow. G:Base of outflow. 

767 Elizabeth Spring/bore. 29°21.30'S, 
136°47.04'E. Curdimurka 483363 (Cobb, 1975:63). 
Coll. WP, RH and DW, 5 Sept. 83. A:Upper outflow, 
under wood. B:Outflow on sedge. C:Outflow under 
wood. 

768 Jersey Springs. 29°20.81 'S, 136°45.52'E. Cur- 
dimurka 671753. Coll. WP, RH and DW, 5 Sept. 83. 
A:Beginning of outflow. B:End of outflow. 

769 Jersey Springs. 29°20.81 'S, 136°45.52'E. Cur- 
dimurka 481 365. Coll. WP, RH and DW, 5 Sept. 83. 
A:Head of spring. B:Outflow. 

770 Jersey Springs. 29°20.81 'S, 136°45.37'E. Cur- 
dimurka 481364. Coll. WP, RH and DW, 5 Sept. 83. 
A:Top of seepage. B:Outflow. C:Small seep. 

771 Elizabeth Springs, N.W. side of hill. 
29°21.30'S, 136°21.14'E. Curdimurka 483364 
(Cobb, 1975:59). Coll. WP, RH and DB, 7 Sept. 83. 
A:Head of spring. B:Upper outflow. C:Lower out- 
flow. 

772 Julie Springs, S.E. side of hill, between Kew- 
son and Elizabeth springs. 29°21.75'S, 
136°46.67'E. Curdimurka 483363 (Cobb, 1975:63). 
Coll. WP, RH and DB, 7 Sept. 83. A:Pool at head. 
B:Upper outflow. C:On steep fall, upper outflow. 
D:Bottom of hill, lower outflow. 

773 Julie Springs, S.W. side of hill, between Eliza- 
beth and Kewson hills. 29°21.68'S, 136°45.06'W. 
Curdimurka 483363 (Cobb, 1975:63). Coll. WP, RH 
and DB, 7 Sept. 83. A:Upper pool. B:Upper outflow. 
C:Lower outflow. 

785 Seepages in mound S.W. of Little Bubbler 
Spring, Blanche Cup Group. 29°27.36'S, 
136°51.91'E. Curdimurka 491351. Coll. WP and 
WZ, 27 Nov. 83. A and В in two very small seeps on 
mound. 

786 Spring N.W. of Little Bubbler Spring, and N.E. 
of Blanche Cup. 29°27.34'S, 136°51.56'E. Curdi- 
murka 491351 . Coll. WP and WZ, 27 Nov. 83. A:ln 
outlet of spring. B:ln upper part of outflow. C:ln 
smaller outflow on same mound. 

787 Spring N.N.E. of Blanche Cup. 29°27.35'S, 
136°51.57'E. Curdimurka 491351. Coll. WP and 
WZ, 27 Nov. 83. 

1000 Strangways Springs, near Irrapatana, large 
spring on southern end of hill. 29°10'S, 136°33'E. 
Curdimurka 458386. Coll. WP and DW, 31 May 85. 
A:Pool at head. B:Beginning of outflow. C:Lower 
outflow. 



116 



PONDER, HERSHLER & JENKINS 



1001 Big Perry Spring. 28°20.45'S, 136°20.7'E. 
Warrina 438487. Coll. WP and DW, 31 May 85. 
A:Beginning of outflow. B:Middle part of outflow. 
C-D:Small seeps on same mound. 

1002 The Fountain Spring. 28°21.1S, 136°17'E. 
Warrina 431485. Coll. WP and DW, 31 May 85. 
A:Pool at head. B:Beginning of outflow. C:Middle 
part of outflow. D:Lower outflow. 

1003 Twelve Mile Spring. 28°18.5'S, 136°15.4'E. 
Warrina 427490. Coll. WP and DW, 1 June 85. A, 
B:Seeps on same mound as main spring. C:Upper 
outflow, main spring. D:Middle outflow, main 
spring. 

1004 The Vaughan Spring. 28°17.4'S, 136°10.1'E. 
Warrina 426493. Coll. WP and DW, 1 June 85. 
General. 

1 005 Outside Springs (most southern and eastern). 
28°17.39'S, 136°12.69'E. Warrina 422495. Coll. 
WP and DW, 1 June 85. General. 

1006 Outside Springs (middle one of group). 
28°16'S, 136°12.5'E. Warrina 422496. Coll. WP 
and DW, 1 June 85. A:Upper outflow. B:Middle out- 
flow. 

1007-8 Nilpinna Springs (at homestead). 28°13'S, 
135°42'E. Warrina 366502 (Williams, 1979:35). 
Coll. WP and DW, 16 June 85. General. 

Several additional nominal springs were visited 
which proved to be dry and no station numbers 
were allocated. These included: 

Oodnadatta Sheet: 
Unnamed. 365552 (Williams, 1979:50). 

Peake and Denison Geological Map 1:150,000. 
Oodloodiana and Oortooklana Springs. To the 
West of Mt. Denison. Sand Creek, Blind, Coppertop 
and Mud Springs. To the East of Mt. Denison. 

Warrina Sheet 
Kerlatroaboorntallina Springs (Mt. Kingston Bore). 
388527 (Williams, 1979:26). 

List of springs not sampled 

There are several springs that, for various rea- 
sons, have not been sampled. They are grouped in 
the list below according to the 250,000 map sheet 
on which they are found. Springs that are found in 
spring groups that have been subsampled are not 
included in this list. Some of these have been re- 
cently visited by consultants from Social and Envi- 
ronmental Assessment (SEA) while preparing a re- 
port for the South Austrailian Govt, on the mound 
springs. 

Oodnadatta: 
Unnamed spring near Big Cadnaowie Spring 
( = Cadna-owie Springs or MacEllister Springs). 
365546. Williams (1979:52) lists this spring but did 
not visit it. Visited by SEA, no snails reported. Mt. 
Toondina Spring. 330534. Listed by Williams (1979: 
56) but not visited by him. 

Warrina: 
Primrose Spring. 441509. Small spring and seeps: 
described by Williams (1979:5). 
Fanny Springs. 425488. Small seeps and ponds; 
described by Williams (1979:10). 



Little Perry Spring. 440494. Bore on spring, flow 
very small (Williams, 1979:15). 

Several springs West of Lat.135.40'S. on the War- 
rina Sheet have not been visited. The few springs 
sampled in this area did not contain any inverte- 
brates and were mostly just saline pools. Some ex- 
amined only from the air appeared to be very sim- 
ilar to those sampled. Oolgelima Spring was visited 
by SEA, no snails were reported. 

Billa Kalina: 
William spring. 442405. Listed by Williams (1979: 
58), but was not visited by him. Visited by SEA, no 
snails reported. 

Emily Spring. 443401. Listed by Cobb (1975:3) but 
not visited by him. 

Curdimurka: 
Walcarina Spring. 508346. Cobb (1 975:46) lists this 
"spring" and states that it is a small seepage. At- 
tempts to locate this spring from the air have failed. 

Stations at which no hydrobiids 
were collected 

During the course of the survey of mound springs 
a large number of springs within spring groups were 
examined that, mainly because of time constraints, 
were not allocated station numbers. Some of these 
springs were rejected because they lacked inverte- 
brates. Thus, with the exception of a few stations in 
the Hermit Hill area, the following list of springs that 
were found not to contain hydrobiids applies only to 
isolated springs or whole spring groups. 

Oodnadatta Springs: 
Okenden (660), Little Cadnaowie (662), unnamed 
(659). 

Northern Springs: 
Melon and Milne (048, 668), Levi (049), The 
Vaughan (051), Edith (675), Taitón (676), Brinkley 
(677). 

North Western Springs: 
Tidiamurkuna (667), Nilpinna (050, 1007-8), Carda- 
jalburrana (714), Weedina (715), Eurilyana (716). 

Middle Springs: 
Anna (718, 750, 751). 

Southern Springs: 
Jacobs (738), unnamed in Lake Eyre South (701), 
McLachlans (697, 698-700), Gosses (696), Fred 
(704), Smith (695), Beatrice (688), North West 
(719), Wangianna (757), Hermit Hill area: Old 
Woman Group (731 , 732), Old Finniss Group (734). 

Springs to the East of Marree: 
(Numbers refer to grid references on the 1 :250,000 
sheets) 

Marree Sheet: Hergott Spring (now a bore) 
(620328), Wirringina Springs (650314), Rocky 
(233343) and Reedy Springs (233341). 

Note: Most of the extant springs to the East of 
Marree have been sampled. W. Zeidler has visited 
Lignum Dam and Spring and Four Mile Spring and 
Bore and in both no evidence of the original spring 



AUSTRALIAN SPRING HYDROBIIDS 



117 



remains. In our experience, and from the informa- 
tion provided by Cobb (1975) regarding these 
springs they are all either heavily degraded by 
bores being placed on the springs or they are re- 
duced to very small seeps. The one exception is 
Reedy Springs. 

Callabonna Sheet: Public House Spring (not 
named on map) (241314) and Petermorra Springs 
(246313). 

Note: Springs in the Northern Flinders Ranges 
and east and northeast of the Northern Flinders 
Ranges are not listed here, although many have 
been sampled. None contain the invertebrate fauna 
seen in the Lake Eyre Subgroup. 



Locality maps 

The locations of the informal spring sys- 
tems are given in Fig. 2, the more detailed 
locality maps in Figs. 58-63 and the key to 
the locations of the locality maps in Fig. 57. 
The distributions of the taxa are shown in 
Figs. 13, 26, 31 and 39 in the taxonomic sec- 
tion. In each map the main drainage channels 
and the main points of reference are shown. 
Lake Eyre Is a salt lake that contains water 
only after flooding, filling only once in several 
years (Kotwicki, 1986). The general topogra- 
phy is flat. 




FIG. 57. General location map. Numbered rectangles refer to Figs. 58-63. On each of the following maps 
only sampled springs are indicated. 



118 



PONDER, HERSHLER & JENKINS 




FIG. 58. The North Western Springs and Freeling Springs. 



AUSTRALIAN SPRING HYDROBIIDS 



119 




^ Canegrass Ridge 



FIG. 59. The Northern Springs. 



120 



PONDER, HERSHLER & JENKINS 




FIG. 60. The South Western Springs and the Beresford Spring Complex. 



AUSTRALIAN SPRING HYDROBIIDS 



121 




FIG. 61. The Middle Springs and Emerald Springs. 



122 



PONDER, HERSHLER & JENKINS 




FIG. 62. The western Southern Springs. 



AUSTRALIAN SPRING HYDROBIIDS 



123 




Onngla Hil 

Scot« I 2S0000 



FIG. 63A. The Oodnadatta Springs and one of the Freeling Springs Group (station 666). 




FIG. 63B. The easternmost Southern Springs. 



124 



PONDER, HERSHLER & JENKINS 



APPENDIX 2 



TABLES OF MEASUREMENTS 

The following tables are a summary of the 
measurement data used in the statistical ana- 
lyses. The original data set was analysed at 
the population level but the volume of these 
data is far too large for publication and, con- 
sequently, the data are presented only at the 
species level and for those infraspecific taxa 
recognised here as "forms". Copies of the full 
data set are housed in the Australian Museum 



and may be made available on request to the 
Senior author. 

The means (top figure) and standard devi- 
ation (bottom figure) are given for each char- 
acter for each species or form. The number of 
individuals measured is given in parenthesis in 
the first column, which also indicates the sex 
(F = female, M = male). Where the number of 
individuals measured for any one character is 
less than the number in the first column, this is 
indicated in parenthesis between the mean 
and SD. An explanation of the character codes 
and details concerning methods of measuring 
are given in the methods section. 



AUSTRALIAN SPRING HYDROBIIDS 



125 



о ,_ -D 
-Q Ф ■> 

CQ ■>< ^ 

. . CO C7) 

£ E ф 
■о .-^ 

Ф "^ ^ 

^ ix 

Q. 3 w 
CO о . - 

> Й- о 

< О ^ 
О) j_ о 



■ф со 

-С ф 



ф 



"С со Q 

а. о •^ 
со Ф ° 



го 



ф 

d? -^ -û 
< го с 

-'S ^ 
с о 

Ф cj)Q- 

3 m ¡я 
СО — СП . 
ГО > ф -г- 

g О О- О 

Е CL •- -С 

^ « $ 

™ с ^ >- 
о ^ - 



3 ^2 

о Ф — о 

-г-, "^^ о -с 

с ° ir о 

= °р ? 

^ ti ^ го 
с(з Ф z' w 

- -^ О) о 
<1^ ,- 

^ тз 



Э О 

сз- 
5 



я ф Ф > 

Е .Z « 



о. .5 
S3SÔ 

со фЯ г 

о х Ф ф 



ф 



. о 

< . 

00> 



о 



ш 



о 3 

^ -' я ^ 
< -с 9^ 

Н 5 ŒI- 



CÛ 



со 



■^ 1- о 

t^ со CT) 
со ^^d 



^-.со 

CD .,- Ю 

c\j со о 
d ^d 



r~- .^ со 

I-; со о 

о ^"о 



о у- Oí 
со СО Р 






со 

т- см 
со d 



00 3"^ 

см 



Т- Ср Q 



•^ ■ — -о 



CD 
N. ю 
Ю 1- 



■* о 
•г^ d 



со 



о 



;5- о см 
г~- m '- 
со ^^■.- 



Ю о "Л 

см in о 
d ^d 



h- о см 
•>- m о 
о ~-^d 



о о о> 

со LO CD 

■r^ ""d 



"^ S о 



CVJg-ÇM 



v-4 ,_ v^ 

•^ s CM 



о rr-co 
Ю iS о 



Ю со 05 



CD in Tt 
CM СО о 



Ю i2 CM 
■t- со о 



■,- in [v. 

CM со о 
^ Cd 



CD ,- со 

■* со о 
d c.d 



in P со 5 
it^ m со о 

г~. ■>- n^d 

со" 
со 
"^ _со 

^- 05 00 со 
^ ^ СО О 

N. d cd 



1^ ^^ 

CD 

0)0) со 

to CM 1- 

- CM d 

< 

en 

«= о 

m о о 

s^ -^ 

CD -^ о 



CD 
- CM 



о 

О) 
CD 

^- <^ 
O) OO 

со 

CD ■^ 
(Л 



m 

È IX to 
О 

2 ^~■ 

Q. 00 

55" in 

CJ ,. •>- 

Cj LL ,^ 

(0 



C3) CD о 
^ CM о 



CD CD rj- 

CM <^J о 



;:í~cj) 
ю CD см 
1- см о 



о ср 00 

см <>J о 
т^ Cd 



^^оо 

CD о со 

•* ю о 
d £^d 



■^ Гч. h- 

о in -^ о 
со c^d со 



00 со 
о -^ 
£2.d 



^^со 
00 см m 
•^ ю о 
т^ 2í.d 



00 CD со 
■^ СО О 

d 2í.d 



Ю CD см 
CD СО ^ 



СО СО О 

■■- .^ со 
со ^^d 



in со m 
см ^ о 



N со см 
■^ ^ о 
d ""d 



■* Я^ со 
t ^ о 
d ^d 



о -^ 
со d 



см 

СП ^.vCM 
•.-NO 

d ""d 



N ,^co 

NN о 

■1-^ ""d 



■.- 


ю 


о 


d 


■^ 


d 


со 
со 

d 


N 

о •.- 

N 


N 
CJ) 

о 
d 


СП 

ю 
см 


С 
О 

го со 
■ — -со 


о 
ю 


d 


d 




Ü 




(Л 


о 


(Я 


(О 


й 

& 


о 



см .^ о 



^^<У> 





. — . 


О) 


ю 


^ 




in 


ц 


^ 


см 




■^ 




_ 


со 


о 


т — 


о 


см 


■^ 





0.19 

(11) 
0.021 


о 
d 


•^ m 
ц о 
Cd 


1.26 

(11) 
0.065 


СЗ) 


^N 

Cd 


N 
Ю СМ 
■* О 


ю 
ю 


^-.in 
N in 
о о 


d d 


d 


cd 


со 
(У) -^ 
о ■.- 


со 
см 


,-^00 

$2Я 



ОО 


О 


■^ 


d 


00 


см 
о 


d 


d 


О) 

Р 


см 


•^ 


о 


со 


см 


Ovj 


d 


со 


ю 

^^N 

со о 
""d 


см 
ю 


^^00 

со о 
"^d 


N 
О) 


ю 


•^ 


d 


00 

C\J 


О) 

со 


со 


d 


IX 


со 


2 


in 



со с^о 



о -^ in 
ю о о 



сз> 
00 r^cvj 
'- t:; о 



•* СЗ) r^N 

со о î^ •.- 

N Г; 'D _; 



СО 


о 


ОО 


N 


со 


т^ 


ai 




N 




со 




d 


со 


со 


ОО 


о 


•г^ 


со 




со 




о 




со' 
о 


N 
СО 


о 


СМ 


(Я 




с 




о 








го 


CD 


55 


".— 


— 


•^ 


< 




Е 




о 


IX 



со 



126 



PONDER, HERSHLER & JENKINS 



^ -^ CO 
C4J C--r^ 



o::^ö 



2 ^ in 
'^ -^ о 
о ri^ 



й^§ 



S8g 



ocog 
^N -^ (M 

CO r\r< 



2"^ 

Ч о о 






b) cvJ ^ 



'i in 

CD 1- 

■r^ Ö 



„ о 

00 r^ 



Ю 


Ю 


о 
о 


00 


00 


CVJ 


C\j 




■^ 


in 


in 


s 


CM 


00 


о 


d 




d 






00 


CO 


In 


со 


■^ 


00 


о 


Ö 




d 


CO 


in 


СП 

1^ 


in 


00 


о 


■^ 




d 


Ю 


cvj 


in 


in 


00 


о 


Ö 




d 


о 


in 


in 
in 


CO 


CO 


1— 


CO 




d 






со 




03 


in 


in 


r^ 


о 


'- 




d 


CT) 


CO 


о 
со 


T— 


2; 


о 


d 




d 


^^ 






•* 




Tj- 


CO t^ 
1^ 1— 


S" 




2, 


т— 


Kc^ 




d 


■* 






Г-- 






^'CM 




(3) 

in 


Г-- T 




■*" 


-co 




d 


05 






Г-- 






co 






CJ)' 




in 


CO ¡P 




о 


о 1^ 




■^ 


- T— 




d 


CO 






CO 






о 






CD 




со 


о о 




1 — 


о Oi 




т- 


(Я -^ 




d 


с 






g 












га 










00 


W in 




in 


• ■ -^ 






Ecvi 




d 








о 












ш 






ü 




in 






in 

C\J 


i^-^ 




d 


< 






E 






i; X 




CO 


О 












CO 






.ü 


















CO 




о 


^^^ 




21 


CO 






ll 







CM ri- CO 
о CO -^ 
CO '"^■^ 



in .rt- CD 

CM 00 о 
d ~~"d 



CM Tj- -^ 

■>- 00 о 
о "'о 



со 

■ч- т* 00 

ю œ о 
' ^^d 



■^ т1- 00 
ю ^ í-^ 



■^ СМ '^ 
in г^ О 



со t^ ■* 
см i» '- 
со ^^d 



С35 ю '^ 
•^ |< о 



,-^см 
со |v^ со 
т Ç 9 
о ""о 



со t^ ■.- 

pli ^ 

см "^о 



см со 

о 1- 



1-1^ т- 



■.- со т- 



-^ О) ¡л 
см см с 
■г-^ ""d 



со О) Й 

о см о 
о "о 



^^со 

2: СП й 

о см о 
о "^о 



ÇD CD со 

со см -^ 



1П Ln со 
in см о 

d ""d 



,-^со 
см 00 '- 
о см со 

со ^^d 



О) If) со 
■* см о 








см 


г- 


^_ 


см 


т— 


см 


о 


d 




d 






см 


т — 


.,_ 


■^ 


СЗ) 


см 


1— 


т^ 




d 



со •^ 
1^ ri 



S:: 


m 


« Cvi 


d 


О 








СО 




ä^ 


in 


•* 


<«? 


•* 


со 


d 


Е 








о 








^ IX 


ел 






Ü 




со 




о 












со 


■* 


S-^i- 


со 


со 





Г-- r^ о 
q CM т 



^-^co 
in r^ in 
q CM СЭ 
о "^o 



■* (v. CM 
о CM СЭ 

о "о 



см fv. 00 

со CM q 



^^CD 
T- со '^ 
in CM СЭ 

d "^d 



cj) o> со 
CT) CM со 
см ^^d 



о 00 ■^ 

in CM q 



r- in CM 

■^ CM СЭ 

о ""о 



CT5 m i^ 

h- CM '- 



en 
r- о 



о 
<y> in 



in ,- о 

■^ CM <=o 
со ^^d 



^-^C35 
C3) ^ 00 

со CM eu 
d "^d 



^^00 

■* ^ со 

■r- CM q 
о "^o 



CM T- CD 
in CM СЭ 

T-' ""d 



со 
in ?0o 



со со о 

со i^ CM 

со ~~^d 



со 
in 00 •.- 



CD 

со r;^■^ 
см L£,o 

d d 



со 
Î2in° 



о 
in Tj- 

со ••- 



in 
со 
со 

-CM 
CD со 
g CM 



со "* 

55 



л; '- CVJ 



m 

E 



_,r^ 

о CM f^ 
со ^^o 



í^ 'Л 55 
со CM о 
d ""d 



in Ln 2 
T CM q 
о ^"o 



CJ) LO 00 

•^ cvi q 



CM со со 
in !;-: о 
d ^^d 



^^см 
о 00 см 
со ^ см 
со "^d 



2 со Й 

■rt т- о 

■г-' ^^d 



in 
■^ г;:;- см 
см if2,o 
d d 



ст> 
со ---со 
о 'л.о 
см d 



о 

■^ 00 



00 

^ о 
со ■.- 



CT) CD 



AUSTRALIAN SPRING HYDROBIIDS 127 

TABLE 18B. Fonscochlea accepta and F. aquatica, palliai measurements. AC, distance of gill apex from 
left side of filament; CO, distance between posterior end of ctenidium and posterior end of osphradium; 
DO, shortest distance between osphradium and edge of palliai cavity; FC, number of ctenidial filaments; 
HC, filament height; LC, length of ctenidium; LO, length of osphradium; ML, maximum length of palliai 
cavity; MM, minimum length of pallia! cavity; MW, width of palliai cavity; WC, width of ctenidium; WO, 
width of osphradium. 



Sex and No. LC WC FC AC HC LO WO DO CO ML MM MW 



F. accepta form A (Stations 002, 003, 752, 753) 

F x 1.59 0.50 31.54 0.21 0.20 0.45 0.11 0.34 0.28 1.88 1.09 1.31 

(^) (10) (12) (12) (10) 

(13) S 0.274 0.073 2.222 0.090 0.052 0.084 0.019 0.149 0.073 0.243 0.126 0.113 

M X 1.46 0.52 30.85 0.23 0.20 0.41 0.12 0.33 0.27 1.84 1.09 1.24 

(13) S 0.224 0.069 2.911 0.063 0.027 0.083 0.026 0.150 0.060 0.328 0.122 0.093 



F aquatica cf. form A (Stations 683, 741, 767, 771) 

F X 1.49 0.45 31.07 0.21 0.12 

(14) (14) (13) 

(15) S 0.267 0.074 2.868 0.090 0.029 



F accepta form В (Stations 689, 690, 692, 694, 71 1) 

F X 1.36 0.45 30.28 0.16 0.12 0.38 0.10 0.23 0.27 1.71 0.93 1.24 

(17) (17) (17) 

(18) s 0.121 0.073 2.469 0.046 0.027 0.055 0.013 0.055 0.055 0.161 0.093 0.117 

M X 1.30 0.47 28.83 0.16 0.12 0.37 0.11 0.23 0.23 1.59 0.88 1.20 

(13) S 0.124 0.058 2.368 0.044 0.029 0.033 0.014 0.049 0.064 0.120 0.095 0.099 

F. accepta form С (Station 703) 

F X 1.69 0.49 32.50 0.13 0.26 0.43 0.10 0.12 1.39 1.96 1.01 1.48 

(4) (4) (4) 

(5) S 0.193 0.034 2.380 0.041 0.023 0.026 0.011 0.006 0.12G 0.256 0.073 0.149 



M 


X 


1.70 
(3) 


0.53 


33.00 
(3) 


0.18 


0.26 
(3) 


0.38 


0.10 


0.34 
(3) 


0.29 
(3) 


1.97 


0.93 


1.38 


(4) 


S 


0.172 


0.053 


2.646 


0.035 


0.038 


0.056 


0.013 


0.038 


0.074 


0.086 


0.059 


0.091 



F. aquatica form A (Stations 028, 030, 039, 679, 683, 720, 723, 739, 741, 747, 767, 771) 

F X 1.59 0.45 34.49 0.22 0.15 0.39 0.12 0.35 0.27 1,97 1.07 1.49 

(33) (33) (34) (24) (33) (33) (28) (29) (34) (33) (32) 

(35) S 0.260 0.138 4.529 0.100 0.039 0.103 0.023 0.079 0.090 0.346 0.350 0.367 



M X 


1.42 


0.43 


34.05 


0.22 


0.14 


0.36 


0.12 


0.32 


0.24 


1.81 


0.97 


1.41 




(20) 


(20) 


(20) 


(20) 


(15) 


(20) 


(20) 


(20) 


(21) 


(19) 


(19) 


(20) 


(22) S 


0.247 


0.167 


5.596 


0.075 


0.040 


0.031 


0.025 


0.067 


0.107 


0.273 


0.224 


0.316 



F, aquatica form A (typical form: Stations 028, 030, 039, 679, 720, 723, 739, 747, 771) 

F X 1.68 0.45 37.00 0.22 0.18 0.39 0.12 0.38 0.31 2.02 1.03 1.62 

(18) (19) (11) (17) (16) (19) (18) (18) 

(20) s 0.223 0.173 3.844 0.109 0.030 0.128 0.024 0.073 0.067 0.317 0.291 0.399 



M X 


1.56 


0.43 


37.82 


0.22 


0.17 


0.37 


0.11 


0.35 


0.31 


1.93 


1.02 


1.56 




(11) 


(11) 


(11) 


(11) 


(9) 


(11) 


(11) 


(11) 


(12) 


(10) 


(10) 


(11) 


(13) S 


0.193 


0.220 


3.710 


0.080 


0.028 


0.027 


0.016 


0.060 


0.052 


0.310 


0.283 


0.317 



0.40 


0.12 


0.31 


0.23 


1.91 


1.12 


1.31 


(13) 


(13) 


(11) 


(13) 






(14) 


0.046 


0.024 


0.075 


0.100 


0.381 


0.416 


0.232 



M 


X 


1.25 


0.42 


29.44 


0.21 


0.11 
(8) 


0.35 


0.12 


0.27 


0.15 


1.67 


0.91 


1.20 


(9) 


s 


0.196 


0.073 


3.712 


0.071 


0.027 


0.033 


0.033 


0.045 


0.094 


0.143 


0.128 


0.145 



F aquatica form В (Stations 045, 046, 665) 

F X 1.56 0.49 33.25 0.18 0.16 0.37 0.12 0.29 0.35 1.82 1.10 1.33 

(8) (8) 

(10) S 0.037 0.025 1.581 0.031 0.015 0.008 0.035 0.071 0.054 0.110 0.057 0.139 

M X 1.47 0.53 33.75 0.23 0.15 0.39 0.14 0.33 0.32 1.88 1.05 1.40 

(8) s 0.146 0.019 2.315 0.032 0.020 0.045 0.032 0.086 0.039 0.224 0.080 0.070 



128 PONDER, HERSHLER & JENKINS 

TABLE 18C. Fonscochlea accepta and F. aquatica, miscellaneous measurements. BM, length of buccal 
mass; CA, distance between ctenidium and anus; DG, length of digestive gland anterior to gonad; LD, 
length of digestive gland; LG, length of gonad; LS, length of snout; LT, length of cephalic tentacles; MA, 
shortest distance of anus from mantle edge; RS, length of radular sac behind buccal mass. 

Sex and No. LS LT LD DG LG BM RS CA MA 

F. accepta form A (Stations 002, 003, 752, 753) 
F X 0.57 0.48 3.42 

(10) (10) (10) 

(13) s 0.128 0.078 1.026 

M X 0.59 0.47 3.56 0.34 2.44 0.58 1.42 0.56 0.79 

(10) (10) (10) (5) (10) (7) (10) (10) 

(13) s 0.118 0.072 0.687 0.142 0.411 0.232 0.144 0.096 0.171 

F. accepta form В (Stations 689, 690, 692, 694, 71 1 ) 
F X 0.41 0.41 2.29 0.35 

(5) (5) (5) (4) 

(16) S 0.053 0.090 0.401 0.092 



0.66 


1.71 


0.76 


1.42 


0.57 


0.76 


(6) 


(10) 


(4) 


(10) 


(10) 




0.360 


0.472 


0.058 


0.193 


0.132 


0.217 



1.28 


0.69 


1.10 


0.53 


0.53 


(5) 


(5) 


(5) 




(15) 


0.119 


0.104 


0.124 


0.102 


0.090 



M 


X 


0.46 


0.41 


2.76 


0.19 


2.01 


0.70 


1.18 


0.60 


0.55 






(4) 


(4) 


(4) 


(4) 


(4) 


(4) 


(4) 






(11) 


S 


0.154 


0.085 


0.272 


0.057 


0.178 


0.047 


0.173 


0.110 


0.109 



F. accepta form С (Station 703) 

F X 0.48 0.53 2.72 0.31 1.60 0.66 1.30 0.88 0.68 

(4) (4) (4) (4) (4) (4) (4) 

(5) s 0.072 0.046 0.350 0.087 0.158 0.075 0.095 0.094 0.144 

M X 0.47 0.57 3.20 0.33 2.23 0.65 1.26 0.83 0.82 

(4) s 0.078 0.078 0.304 0.047 0.209 0.063 0.279 0.138 0.142 

F. aquatica form A (Stations 028, 030, 039, 679, 683, 720, 723, 739, 741, 747, 767, 771) 

F X 0.63 0.55 3.23 0.46 1.75 0.99 1.61 0.65 0.74 

(23) (24) (26) (26) (28) (24) (30) (22) (22) 

(35) S 0.108 0.160 0.631 0.138 0.660 0.067 0.244 0.163 0.180 

M X 0.58 0.48 2.93 0.41 1.98 0.93 1.53 0.60 0.77 

(16) (17) (17) (17) (18) (17) (19) (18) (18) 

(22) S 0.095 0.109 0.723 0.100 0.649 0.079 0.193 0.185 0.222 

F. aquatica form A (typical form: Stations 028, 030, 039, 679, 720, 723, 739, 747) 

F X 0.65 0.62 3.47 0.47 2.00 0.95 1.61 0.58 0.74 

(10) (12) (17) (17) (19) (12) (19) (10) (9) 

(20) S 0.143 0.180 0.498 0.109 0.642 0.042 0.235 0.105 0.103 



M 


X 


0.61 


0.52 


3.47 


0.45 


2.43 


0.95 


1.57 


0.65 


0.86 






(7) 


(8) 


(8) 


(8) 


(9) 


(8) 


(10) 


(9) 


(9) 


(11) 


s 


0.099 


0.099 


0.476 


0.099 


0.518 


0.031 


0.199 


0.209 


0.269 



F. aquatica cf. form A (Stations 683, 741, 767, 771) 
F X 0.61 0.49 2.79 0.45 

(13) (12) (9) (9) 

(15) s 0.072 0.109 0.635 0.188 



1.23 


1.03 


1.61 


0.71 


0.75 


(9) 


(12) 


(12) 


(12) 


(13) 


0.299 


0.062 


0.127 


0.183 


0.223 



M X 0.56 0.43 2.45 0.38 1.53 0.91 1.49 0.55 0.69 

(9) s 0.092 0.106 0.546 0.096 0.412 0.104 0.189 0.153 0.125 

F. aquatica form В (Stations 045, 046, 665) 

F X 0.55 0.56 3.19 0.60 1.29 0.89 1.71 0.74 0.59 



(8) (8) (8) 



0.60 


1.29 


(6) 


(6) 


0.036 


0.165 



(10) 


s 


0.075 


0.117 


0.408 


0.036 


0.165 


0.052 


0.188 


0.056 


0.084 


M 

(8) 


X 

s 


0.54 
0.085 


0.54 
0.148 


3.11 
0.509 


0.52 
0.111 


1.86 
0.191 


0.83 
0.085 


1.48 
0.082 


0.73 
0.113 


0.70 
0.093 



AUSTRALIAN SPRING HYDROBIIDS 



129 



TABLE 18D. Fonscochlea accepta and F. aquatica, stomach and male genital measurements. AS, height 
of anterior stomach chamber; PL, length of penis; PP, length of palliai portion of prostate gland; PR, 
length of prostate gland; PS, height of posterior stomach chamber; PW, width of prostate gland; Sl! 
length of stomach; SS, length of style sac. 



Sex and No. 



SL 



SS 



AS 



PS 



PL 



PR 



PW 



PP 



F. accepta form A (Stations 002, 003, 752) 

F X 1.05 0.61 0.72 0.63 

(10) s 0.333 0.085 0.124 0.066 



M 



(10) 



0.95 



0.55 



0.383 0.102 



0.64 



0.081 



0.53 



0.046 



2.66 
0.593 



0.52 
0.108 



0.33 

(7) 

0.046 



0.10 
0.086 



F. accepta form В (Stations 692, 71 1 ) 

F X 0.73 0.50 0.65 0.51 

(5) s 0.073 0.024 0.085 0.060 



M 
(4) 


X 

s 


0.69 
0.082 


0.47 
0.061 


0.61 
0.094 


0.53 
0.095 


1.73 
0.311 


0.46 
0.079 


0.29 
0.054 


0.08 
0.021 


F. accepta form С (Station 703) 
F X 0.98 
(4) s 0.052 


0.49 
0.064 


0.66 
0.053 


0.65 
0.032 










M 

(4) 


X 

s 


0.90 
0.102 


0.49 

(3) 

0.012 


0.66 
0.048 


0.64 
0.078 


2.32 
0.184 


0.55 
0.089 


0.25 

(3) 

0.035 


0.10 
0.042 



F. aquatica form A (Stations 028, 030, 039, 679, 683, 
720, 723, 739, 741, 747, 767, 771) 
F X 1.14 0.69 0.85 0.69 

(31) s 0.402 0.108 0.156 0.153 



(Stations 039, 683, 720, 723, 
741, 747, 767, 771) 



739, 



(18) 



X 0.96 
(16) 
s 0.286 



0.61 
0.114 



0.77 0.65 

(18) 
0.126 0.124 



2.33 
(13) 
0.743 



0.51 
(14) 
0.114 



0.35 
(12) 
0.065 



0.11 
(12) 
0.089 



F. aquatica form A (typical form: Stations 028, 030, 039, 
679, 720, 723, 739, 747, 771) 



(Stations 039, 720, 723, 739, 747, 771) 



F 
(18) 


X 

s 


1.32 
0.439 


0.74 
0.078 


0.91 
0.160 


0.75 
0.148 










M 

(15) 


X 

s 


1.13 
0.262 


0.66 
0.093 


0.85 

(7) 

0.150 


0.73 
0.063 


2.52 

(8) 

0.874 


0.55 

(9) 

0.115 


0.38 

(7) 

0.055 


0.11 

(8) 

0.108 


F. aquatica cf. form A (Stations 683, 741, 
F X 0.89 0.61 
(13) s 0.120 0.093 


767) 
0.77 
0.106 


0.60 
0.117 










M 
(9) 


X 

s 


0.79 
0.193 


0.56 
0.113 


0.70 
0.049 


0.57 
0.122 


2.04 

(5) 

0.320 


0.43 

(5) 

0.073 


0.30 

(5) 

0.048 


0.10 

(4) 

0.038 


F. aquatica form В (Stations 045, 046, 665) 

F X 1.09 0.69 0.77 
(10) s 0.111 0.052 0.072 


0.70 
0.068 










M 
(8) 


X 

s 


1.12 
0.044 


0.70 
0.034 


0.79 
0.013 


0.74 
0.051 


2.87 
0.178 


0.56 
0.019 


0.34 
0.071 


0.11 
0.012 



130 



PONDER, HERSHLER & JENKINS 



CO Ь CO о 

■= - « S 
"5 Q ^ - 
о Ü J- > 

(o:9 ~" 

СЛ > 



с <D 



f °o 

о с о 
£■1^- 



-5 « 



CÛ O O 



« ^ -c ^ 

en в SB) 
с с °- с 
55 _Q) -о oj 

§>ÔCC 
zQ О ü со 
03 • -.^ . . 

о ^ - §- ■ 

"5, "^ 0) > о 

с ^ Е о га 

ш 3 га тг Q. 

~ ^Ъ ° <^ 
" о. ,- о 

< о 2 ° "- 
•^ Е -с -= 

ф О) > О = 
^ .d о •^ 



_ о 
о £ 



га f •) 

Е ж " Oí -^ 

_ Ё 2 ф 5 

.^ m g- Е СП 

с 3 



Ф "О =* 



Ф Q. 

О) О _ ^ 

О ^ га .?< 
Е ^ 



Ф "О 

га 

С.ь га 
Е.£ 
9i Е 



Е "^ 

■ä га о ^ 

CD :; *- 

Э ф 



Cr 71 



СО 



га 



C3)ID 

_ со 

^^ &° ^£ 
тз о х: о "О 

со го ^ с СО 

oí" Е СП <="^ 

о ^ о ^ -5 
со : . - га с 

'ч с с s га 



о 
о 
<л 
с *- 



_ га 

о SZ i: 

о i: с 

О) пч 



Q о ^^' s- s 

Ц. 3 « Ф 
"О 3 "о 

ш > -Qq. с 

со о ^ о 

^ о ■ - -с 

CÛ ОПЭ "О ф 

H ^ о o¿ 



у- LO 



;гг г^ 1- 



c\í о ■^ о 

СЗ) -1-^ ' 'о 



об'" SS 



,^ ■>- о 



со О) t^ 



(О :р •.- 
р с\1 р 



со 



о -— о 



"^ CV! й Р 
tC о ^ — ^ о 
со 



сю 



^. р '^ 

CJî 

m CD 

f^ г- ;:сг^ 



CM 



ö 

00 о ^ — ^ о 

CD 



^- 5 2Í ^' 



о 

о (О 

coS«í5 



со "^ й 



Cvj 

< 

E 

o IX (Я 



СЛ 2 ■.- 
Ö ^ Ö 



;: ^n o 



r^ со sr "^ 
со CM °' o 

f^ o — СЭ 



o ■* 

oi m f^ -^ 

r^ CN ^ o 

o ^ d 
со 

o CM 2 o 
cb 



o CM <" 

d o ^ 
со 



CM '" 



со 1^ 

■ ■ a¡ ^ t~- 

£ a> '^ T- 

vr d ^- cb 



Ь ■ 
< '" 

E 

o IX 
50 



— . со 

CD o '- 

O il O 



^ CM o 
Ö ^ Ö 



T^ ^ d 



s ^5 

Ö ^ Ö 



CM 52 o 



■,- c2 o 

Ö ^ Ö 



!:: Ǥ 



со ,- '-^ со 

1^ i^ c2 " 

^" d " d 

«o § 

(O <3 'T 



s 

55 o 

~ o ^ t 

< со " (M 

с '- — d 





1Л 

о 
о 


со о 




со 
со 
d 


со 

да о 
d 




со 
р 


см 




''" 


d 




CD 
О 

d 


да о 




1- 


да 
о 
о 




d 


d 




ю 


да о 
"" d 




ю 
о 
d 


^ eg 
да о 
~ d 




см 
d 


да 
Ó0 о 

~ d 




см 
см 
d 


1Л 

^ см 
да о 
"" d 




C\J 

о 
d 


S ° 




см 
d 


,-. см 
да о 
"^ d 




in 
сч 
d 


да 

— CvJ 

да о 

"" d 




с» 
d 


со 
„ см 
со о 
"" о 


in 

CD 
СО 

со' 

О 


CJ) 

d 


см 

— С35 
CD О 

"" d 


С 

о 
те 

ш 


с» 
h-. 


со о 
""' d 


Ê 
о 


IX 


СЛ 


5 
1 

Cï- 

со 




Ô 


iL 







AUSTRALIAN SPRING HYDROBIIDS 



131 



- Ч- • - 1 






tv. 


^^О) 




tv. 




00 




см 




t^ 


^ 


th; BW 
ngth с 
II width 


Z 


■* 


о 


CD CD ID 


см оГ 


о 


in CM 


о 


in 


■* 


00 


о 


S 


û_ 




1^ 


q -5- t^ 


Ч -ï^ 


CD 


h-- со 


in 


q 


•^ 


00 


CD 


с 
с 




со 


d 


CM ~^d 


■^ 


d 


■^ 


d 


■^ 


■^ 


со 


d 


























о 


•O OJ Q) 


























u 


2E^- 






^_ 


,-^00 




oo 




00 




СМ 




■^ 


■^^ 


О 


■^ 


■* 


■^ CD '^ 


о OÎ" 


■* 


CM 


со 


1— 


1^ 


00 


in 




C\J 


о 


CM -^ о 




о 


■^ со 


о 


со 


о 


см 


о 






Ö 


d 


d ""d 


d "" 


d 


о "" 


d 


d 


d 


d 


d 






























S.E£ 


I 


in 


CD 
CM 


■5Í CD <^' 


со оГ 


CD 
CM 


CD CM 


CM 


^ 


см 


о 


CM 

со 




^ Q- ^ 




q 


T-^q 


о 1Í- 


о 


о со 


о 


■^ 


о 


■^ 


о 






d 


d 


о ""о 


d "" 


d 


d "" 


d 


d 


d 


d 


d 




< ..- 




























., .tr (Ц 






























_J 


CD 


Ю 
CD 


•^ CD со 


1^ Ci 


CM 
Ю 


•^ CM 


in 


in 


о 


00 


CD 

in 




Ш 0) x" 


о 


00 


о 


00 ^ О 


ю ^ 


о 


Ю со 


о 


CD 


■^ 


00 


о 




ö 


d 


d ^d 


d ^ 


d 


d "" 


d 


d 


d 


d 


d 




Q) CO . - 




























Ь 3 w 




























apertL 
Icáreo 
whorl 


Q 
Q. 


о ^^ 


CD 

со 
о 
d 


•^ CD ^ 

•^ -^ о 
d ""d 


5 S 
d 


со 

ID 

о 
d 


en ^ 

d 


о 
d 


■^ ,- о 
d "^d 


T- 5^ см 
■^ !i- о 
d ^d 




. ra ^ 




























X о о 




























^.в§ 






CM 


00 




ю 




T— 




1- 




со 




^ 


о с: 


CD 


со r^-^ 


00 ^ 


.CD 


in ^ 


.r- 


00 


00 


CD 


CD 




to ^ о 


œ 00 


CM 


CM — 'O 


CD Tf 


о 


•^ CM 


o 


CD 


т— 


со 


^ 




c^o 


1— 


cvjS 


-d 


T-I ^^ 


'd 


CM "" 


d 


CM 


d 


см 


d 




0) CT •- 




























m 0^ — 






















































£ — О 




























to ü ш 


Q. 




со 
о 


О (D 1^ 
Ю -^ О 


85 


^o 




:ê 


о 
q 


^о 


о 
ю 


^o 








'd 


T^ ^d 


d 




t— 


d 


■^ 




■^ 
































га^ . 








^ 




CD 




CD 




см 




t^ 




IP, 


> 


oo :^ 


~5- 
о 


^ю§ 


1^ 


О 


(J> 


О 


CD 


с^со 
1^ о 


CD 


'- о 




О 


d2 


■ 

'О 


d^d 


d 


d 


d 


d 


d 


^d 


d 


-'d 




о. о (Л 
о ° CT . 






















































„ •.- 0) -^ 




























and 
ith 
lar p 
whoi 


CO 


CTO ^ 
CM II 


со 

-CT) 

о 


CD 


CD 
00 


5 
о 


Ю 

oo 


о 


CD 

q 


со 
Р2 


CD 
CM 


CD 




= CT 3 ^ 
Ф с Ü -^ 


^Sid 


^S^d 


d 


d 


d 


d 


•r^ 


^о 


•r^ 


^d 




-С tu " ^ 




























w — Ш о 




























- Q-J3 




























<nica, 
OL 
of 

irlof 


^ 




о 
со 


о CM 


CD 


oo 


со 


о 


q 


о 
см 
см 


q 


со 




ü í; Ш ^ 


CO 




d 


T^ d 


T-^ 


d 


T-^ 


d 


CM 


d 


■^ 


d 
































and 
ate ' 

st ha 


^ 


со 
q 


о 
о 


CD 
00 CD 
CD О 


in 


CM 
00 

о 


CD 
CD 


00 
CD 

о 


in 


о 
со 


00 

q 


q 




to E CL ¿ 


< 




d 


d d 


CD d 


d 


d 


d 


4- 


d 


•^ 


d 




с -^^ *^ 










со 








co 










Ilakali 
penul 
r peg; 
dthof 




O) r^ 


со 
со 


ID 
T- CD 


c 
о 
•,;= 00 


со 

Ю 


in 


CM 
CD 


CD 

со T- 
o 


5- 


CM 


q 




^ o|> 


X 


со о 


q 


о о 


га CD 


о 


CD 


о 


'^. 


T 




< 


h~ 


d 


т-^ d 


со d 


d 


d 


d 




о 


■^ 


d 




s, F 
xity 
perc 
WB 




od 
о 






E 








со 
о 










vahabili 

conve 

allest 

whorls; 




со '^ 


со 
d 


со 
■* со 

со -г- 

т^ d 


о 


о 
d 


о 

CD 

d 


со 
I-- 
о 
d 


^ 'S- 

со '^ 


CD 
00 

d 


со 


oo 
со 

d 




m> Z. SZ 




с 
о 














о 










LE 19A, Fonscochle. 

th of body whorl; С 

oconch; PH, height о 

number of teleoconc 


I 

ó 


га 

<C4Í 

Е 

о IX 


О) 
О) 

d 
со 


со см 
см см 
см см 

X со 


О 

га •^ 

ra 

E 

-ÏÏ-IX 

•Su 


ю 
d 

eo 


о 

IX 


oo 
d 

со 


га 

^^ 
m см 

Е 

о IX 

со 


r- 
ю 
со 
d 

со 


00 

in 

CM 

IX 


00 

d 

(0 




z 

■o 

с 
га 


5 

га 


s 
^ 


2 S 




о 

CM 


s 


in 
со 


«0 
1" 


о 
2: 


s 


CM 

со 




H ií Q.I- 


X 


ь 






















ш 
с/5 


LU 






u: 








LL 











132 



PONDER, HERSHLER & JENKINS 



<0 h- 



CO 

о 



52 ^ c^ 

O) о 
6 Ö 



о ^C\J 

■^ ,- о 



s ьо 



CD 
CO ;r-0 
C4J £2-0 



CO CO О 



CD -^ 

■.- d 



in 1^ 
■.- о 



T-: о -- о 



in 
со 

со 

in со 
•* г- 
о ^ 



о 

Е 
о IX 



о 
1- см 
со -^ 
т^ d 



05 о i2 
in (v. о 



CD 
со íT-CO 

•^ Р о 



CD со Çii 
со in р 



о 
î2io§ 



со 



о 
о ^'О о ^--о 



oOp pSP 



со 

CD Р Р 



3 m c^¡ 
•^ c\j р 

d ^о 





со 




iñ 




О 


■^ 


^ 


d 


■^ 


оГ 




т— 


¡п 


о 


d 




d 
in 


CD 


^~- 


О) 


см 


О) 
m 


о 



Sfs? 



со Рем 
Tt см Р 
d ^о 



■^ со см 
m ^ о 



m t^ ^ 
•^ ^ °. 
о ^о 






СП т- 

т^ d 



<^J т^ 



CD 

tv. 
. in 

СП t^ 
CM ^ 

о 
со' 



со (П 

о см 



со 

о , _^ 
О) 1 •^ 
d 



in 
со ■* 
-.- о 

d d 



? о 



со 
т- со 
IV. о 



о см Tí 

■* -^ о 



о 
о см 
■^ о 



о о со 
m ■^ о 



(D Tf -^ 
-- Ю Р 

d ^о 



Я1 г^>^ 



CD 



h- Ч 1Л "" 
. т- ■ — -о 
со 
со 



йё 



^-.со 

CD Й со 

со о о 
d ^d 



^со 

CD 2^ 1^ 

•* о о 
•г^ ^d 



,^о 
со со '* 

-^ ^ р 
d ^о 



со 



со 
т- d 



CD О 
П СМ 



со IV. 
О О 

°см 
со 

с 
о 

я IX 

со 



и: 



AUSTRALIAN SPRING HYDROBIIDS 



133 



< 


ЕЁ 
го и 

05T3 




E 


^1 


" + 


3 

Ъ 

с 

О) 

ü 


о о 








о 


*- СП 

о с 


х: « 


£ 


^^ 


fo 








^ 


05,С 


— £ 


5 


§у 


ГО с 


О 
< 

¡á 


of 

Q ф 




с 


Е я 


^' и 


E 


3 ;^ 


«г СО 
^ ГО 


С ГО 


>;Е 


^- 


<D ТЗ 




3 
(Л 

та 
ф 
Е 


t5 с 
о Ü 
.9-0 

II 

W 3 


> го 
я о 
оо 

го -О 


Е 
о 


11 

£ о 


•о 

с 
го 


о с 
о. . 


О) го 
с со 

0) ^ 

_ го 


го 

го 
о. 


го .- 

IÎ 

тз о 

го _ 


i 'S 
>< 2 

го ^ 
Е о 


(Л 


î= га 


г ^ 



itt|t| 

го -í: TS E — .2 
J)!^ ^ О) ГО í^ ^ 

Е о <^ о.Ф « 

. '^ тз W JD " 
СС 0) с О Е *г 

.о "й та ^ го ° 
с о ^ о ^ ^ 

8 "^if ":§ 

^^ ф го <ц та _- 
5¿- еО 
Q-0 2 5 



■о > 

та XJ 



(Л 



ее ф 

с Ü 

-= с 

та та 



Ч! .52 



.„ S 2 о 1 



(Л Ä 



ф 



05 о Ä 



о. о 



,- - та 

Ф .Ü 



-о -Q тз ' ^ . 

со Е *- СО , о 

с та ся _1 (Г) > 

та ^ ф ..а'> 

^ о -с тз "- . . 

гг, г- о с - (J 

1^ w " -55 5 « 

§ 5 ^ О) та о 

Ш С О) О О С 

О) го Q _^ ^ ф 

у s g ê 5 ^ 

m О) Ф ->';,- 

< Ф с Q > та 

I— ^ та _1 ^ (Л 



1 1 

О) 
LT) о 


о о 


О (М 

со т- 


о о 


со 5 


о о 


CvJ ..- 


■г^ Ö 


см 

сю т- 

с£) í^J 


о Ö 


со 
со со 
t- см 


'- Ö 


1 1 

ел со 

у- о 


CD о 


о .^ 


CD Ö 


см о 


Ö Ö 


о ю 
см о 


Ö Ö 


о о 


Ö Ö 

см 


о ю 
см о 


Ö Ь 


О) 
CD С35 

г- о 


Ö Ö 
1 1 


1 1 

о сЬ 



о со 

< см 
Е ^ 



СП IX 



Си m 



о со 

со 

Е ^ 



со 



о о 



^ 00 

сч см 



со 
см 


ê§ 


см 


ID 

о> о 


Ö 


сЬ 


о 


о 


со 


g§ 


^ 


^ со 


сЬ 


со 


о 


"^ Ö 


см 


см 

CD о 


см 


СП р 



со о ,^ h; 



СМ t^ '- 



О 

<Л СТ> 

С C\J 

О .^ 

S 
СЯ IX 



§ô5£ 



СО S' '" 
со ° '- 

сЬ — о 



т- о о 

о — CD 







•* 


г^ 






о 




о 


Ö 




CD 

in 


t 




■* 






о 


CD 




CD 

in 


СО 


о 


in 
о 


сЬ 




d 

CO 


in 




■* 


in 




.* 


к 




c6 






•^ 


CD 




CD 






О 


CD 




d 


СО 




5 


CD 




T~ 


d 




d 
in 


CO 


^ , 


CD 


in 


2- 




CD 




d 






CO 


CM 


t^ 




CD 




d 

CO 


in 


, ^ 


T— 


CD 


œ 


CO 


CD 




d 
in 


CO 


, , 


t^ 




2Î- 


CM 


-^ 




d 


IX 




CO 


s 




^"' 



134 



PONDER, HERSHLER & JENKINS 



о oo 
x: с о 

^ ^ ts 
Ш о ^ 

О "o > 
m ^ о 

-b-'S о 

i5 О § 
^s: -с 

с О) g 
Е ^ -с 

5 - <1^ 
5 > • - 
rao Я 



ОС о 

Q- JZ 



га . 
спСС 

2 X 

с о 
о я 

я ^ 



о -о о га 



|t5 
щ Ф ^ 

~ га 



< л _ 

. о со 

« ч- с 

■Hob 

^£ га 



га £ 
о.^ 

о 5 

£cû 

ш ^- 
~ ш 

Û- га 
. - х: 
ел о 

с -= 
Ф га 



^ О 

га .g у -с >- 



га 

Q. 

О) ¿ Ü 
0) О Ï 



Ç хз 



л «-5 
Ф Ъ "5 



с с 
ф g 
. С 
_1 О 
CL Q- 
• - О) 

э -^ 

■g 

> 

° S 
о g' 

с О) 

g — 
Я> 



Е 

ф v^ „_ 

W 5 о 



0) £ 
О) 



Ll га 

га фСС 
стз «Q 

^ с X 
со ф с 

il 15 Q. 

no 

О _. 
. т Я 

■й -о Í2 р i: 
^ '^ -I ^ гт 



■> 3 



со О О 

СО CJ) 3 
-Ï с "О 

о~ ° 
о W J- 
со m Ь 

CT) clQ 
■^ о 



Ф Ш 

82 

.^ Q. 

11 

ТЗ ф 

га : 

О) 
_ С 
О О) 



СС 
Ü W 

э ■ . 
■о тз 

о i? - ■ 

^ CT ш 

га ф о 
-1 -2- га 
^2о. 

с (Я ф 

rapo 



CT о. 



ф 



"tí 



LU w 

_| га 3 
m w -о 
< з> 
I— хз о 



CT CT s; 
с с Q> 
Ф Ф « 



^ о 



5 -^ c\j о 

. о "^ о 

00 

о 

о ^ 

(Л т- . .IV. 

£ Г^ (М о 

в Ö — Ö 

со 

^ (О .^ 

< '- c\j о 

Е - ^ 



со 

Е - 



со 

f -, (35 

U со 

Е - 



,-^ •■- •* 



?5 ■^ о 



^ 1^ 



о ■* ^ 



AUSTRALIAN SPRING HYDROBIIDS 



135 



> (Л (Я 

• - «^ 

¿ o^ 

$ *- о 

^ с 

•^± В 

I" 2 

O) о <i> 
с с -Q 
ií О E 

> 2 с 

3 — "05 
Q. E . 

Ф M = 

^ о 2 

Q.-C 
QJ Q) с/) 

ф Г) W 

°- о • - 
га ш « 

- га о 
X о ^ 

< го г 
g£ 8 

с СП о 

О) с "5 
ф 2 

Q. 

'^ .. Ф 
»- Е -Q 
« D Е 

^ "5 3 
е Ü с 

° о О. 



3 ^ 



?В 



= ™q.-ç 



ф i5 



ф 



«^3 5 
тэ ^- о -Q 



о го 5 ^ 



Е с ^ 



8.Е 

i£ go- Sí 

c\j >, i- £ 
.t= ra -o 

_i Ф ü 5 
5 с Фш 
I— o o > 



o q 



Ю -1- o 

d C- o 



ID CT) O) 
Ю Ц C\J 

со ^ d 



^ o ?í X- 



a> Й ¡^ 
со o o 

d ^ d 



c\j c^ -* 
со o o 
d — d 



LO •■- 

■r^ d 



o JO 

""' d 



O) o íO 

o í: ••- 



Ю ° o 



со T- 
T-' d 



to ■* 7- 



^ „ со 
со ■,- o 
■r^ ^^ d 



C«5 --. T}- 

ОТ S£. O 
7- d 



136 PONDER, HERSHLER & JENKINS 

TABLE 20B. Fonscochlea zeidleri, palliai measurements. AC, width of ctenidium from left side to position 
of filament apex; CO, distance between posterior tip of osphradium and posterior tip of ctenidium; DO, 
shortest distance between osphradium and edge of palliai cavity; FC, number of ctenidial filaments; HC, 
filament height; LC, length of ctenidium; LO, length of osphradium; ML, maximal length of palliai cavity; 
MM, minimal length of pallia! cavity; MW, width of palliai cavity; WC, width of ctenidium; WO, width of 
osphradium. 

Sex and No. LC WC FC AC HC LO WO DO CO ML MM MW 

F. ze/d/er/form A (Stations Oil, 013, 018, 024, 026, 028, 030, 034, 039, 046, 694, 742, 766, 771) 

F x 1.43 0.43 26.68 0.19 0.11 0.44 0.12 0.28 0.30 2.05 1.02 1.49 

(32) (33) (31) (16) (31) (32) (26) (23) (25) (29) (27) 

(34) s 0.365 0.136 3.042 0.088 0.039 0.125 0.034 0.088 0.094 0.427 0.364 0.438 



M 
(25) 


X 1.28 
(23) 
s 0.355 


0.43 
(23) 
0.164 


25.42 
(24) 
3.175 


0.20 
(23) 
0.105 


0.08 
(12) 
0.023 


0.38 

(22) 
0.139 


0.13 
(22) 
0.086 


0.27 
(18) 
0.095 


0.29 

(22) 

0.124 


1.66 

(21) 

0.414 


0.93 
(19) 
0.223 


1.47 

(21) 

0.368 


F. zeidleri form В (Statior 
F X 1 .26 
(5) s 0.171 


1661) 
0.37 
0.069 


23.20 
0.837 


0.14 
0.048 


0.12 
0.038 


0.36 
0.038 


0.09 
0.007 


0.31 
0.082 


0.28 
0.024 


1.59 
0.225 


0.91 
0.084 


1.29 
0.142 


M 
(4) 


X 1.23 
s 0.115 


0.33 
0.039 


24.00 
1.633 


0.15 
0.012 


0.14 
0.035 


0.35 
0.021 


0.11 
0.025 


0.30 
0.030 


0.28 
0.057 


1.51 
0.183 


0.95 
0.197 


1.31 
0.138 



TABLE 20C. Fonscochlea zeidleri, miscellaneous measurements. ВМ, length of buccal mass; CA, 
distance between ctenidium and anus; DG, length of digestive gland anterior to gonad; LD, length of 
digestive gland; LG, length of gonad; LS, length of snout; LT, length of tentacles; MA, distance of anus 
from mantle edge; RS, length of radular sac behind buccal mass. 

Sex and No. LS LT LD DG LG ВМ RS CA MA 

F. zeidleri zeidleri (Stations 011, 013, 018, 024, 026, 028, 030, 034, 039, 046, 694, 742, 766, 771) 

F X 0.53 0.38 3.27 0.45 1.50 0.74 0.87 0.47 0.40 

(18) (18) (27) (24) (26) (18) (27) (16) (15) 

(34) s 0.100 0.076 1.050 0.123 0.640 0.105 0.136 0.137 0.150 



M 
(27) 


X 

s 


0.48 

(13) 

0.078 


0.35 

(13) 

0.080 


3.52 

(22) 

0.771 


0.33 

(22) 

0.100 


2.46 

(22) 

0.852 


0.63 

(13) 

0.124 


0.83 

(21) 

0.164 


0.57 

(17) 

0.234 


0.33 

(12) 

0.083 


F. zeidleri form В (Station 661) 
F X 0.45 
(5) s 0.065 


0.41 
0.075 


3.00 
0.245 


0.38 
0.065 


1.55 
0.136 


0.63 
0.046 


0.84 
0.061 


0.49 
0.080 


0.43 
0.071 


M 

(4) 


X 

s 


0.45 
0.021 


0.38 
0.055 


2.51 
0.255 


0.32 
0.116 


1.48 
0.219 


0.63 
0.067 


0.90 
0.104 


0.48 
0.118 


0.40 
0.096 



AUSTRALIAN SPRING HYDROBIIDS 137 

TABLE 20D. Fonscochlea zeidleri, stomach and male genital measurements. AS, height ot anterior 
stomach chamber; PL, length of penis; PP, length of palliai portion of prostate gland; PR, length of 
prostate gland; PS, height of posterior stomach chamber; PW, width of prostate gland; SL, length of 
stomach + style sac; SS, length of style sac. 

Sex and No. SL SS AS PS PL PR PW PP 

F. zeidleri form A (Stations 01 1 , 013, 018, 024, 026, 030, (Stations 01 1 , 018, 024, 034, 039, 

034, 039, 046, 694, 742, 766, 771) 694, 742, 766, 771) 

F X 1.02 0.71 0.81 0.68 

(17) (27) (20) (25) 



(28) 


S 


0.320 


0.151 


0.143 


0.138 










M 
(24) 


X 

s 


0.93 

(14) 

0.377 


0.70 

(22) 

0.168 


0.74 

(19) 

0.177 


0.64 

(19) 

0.133 


1.83 

(23) 
0.537 


0.65 
(24) 
0.290 


0.37 
(17) 
0.129 


0.14 
(23) 
0.119 


F. zeidleri form В (Station 661 ) 
F X 0.83 
(5) s 0.040 


0.69 
0.072 


0.68 
0.050 


0.64 
0.028 










M 

(4) 


X 

s 


0.82 
0.025 


0.69 
0.064 


0.68 
0.062 


0.63 
0.044 


1.59 
0.113 


0.56 
0.047 


0.32 
0.060 


0.14 

(3) 

0.082 



TABLE 20E. Fonscochlea zeidleri, female genital measurements. AG, length of albumen gland; ВС, 
length of bursa copulatrix; BS, length of oviduct between "seminal receptacle" and bursa copulathx; CG, 
length of capsule gland; CV, length of coiled portion of oviduct; DB, length of duct of bursa copulatrix; 
DR, length of duct of "seminal receptacle"; DV, maximal diameter of coiled portion of oviduct; GO, length 
of glandular oviduct; GP, length of genital opening; MO, minimal diameter of coiled portion of oviduct; 
SR, length of "seminal receptacle"; VC, length of free portion of ventral channel; WB, width of bursa 
copulatrix; WR, width of "seminal receptacle". 

Sex and No. GO CG AG ВС WB DB SR WR DR CV DV MO VC BS GP 

F. ze/d /ел/ form A (Stations 011, 013, 018, 024, 026, 030, 046, 694, 742, 766, 771) 

F x 1.55 0.86 0.72 0.24 0.22 0.099 0.31 0.24 0.10 1.56 0.11 0.06 0.47 0.15 0.07 

(25) (25) (26) (27) (27) (27) (26) (22) (25) (27) (27) (26) (17) 

(28) s 0.467 0.246 0.220 0.071 0.048 0.031 0.122 0.079 0.057 0.261 0.022 0.011 0.104 0.053 0.020 

F. ze/d /еп form В (Station 661) 

F X 1.49 0.80 0.68 0.20 0.18 0.04 0.23 0.24 0.09 1.26 0.09 0.05 0.40 0.15 0.05 

(4) (4) (4) 

(5) S 0.132 0.082 0.055 0.017 0.006 0.011 0.023 0.041 0.020 0.071 0.011 0.005 0.028 0.023 0.005 



138 



PONDER, HERSHLER & JENKINS 



TABLE 21 A. Trochidrobia species, shell measurements. AH, aperture height; AW, aperture width; BW, 
length of body whorl; PD, maximal length of protoconch; PW, number of protoconch whorls; SH, shell 
height; SW, shell width; TW, number of teleoconch whorls. 



Sex and No. 



SH 



SW 



AH 



AW 



BW 



PW 



TW 



PD 



T. punicea (Stations 002, 007, 008, 022, 025, 027) 
F X 1.43 1.74 0.80 



(95) 



0.136 



0.263 



0.085 



0.80 
0.068 



1.26 
0.136 



1.48 

(83) 

0.073 



1.97 
0.145 



0.37 

(90) 

0.038 



M 

(36) 


X 

s 


1.35 
0.145 


1.64 
0.104 


0.77 
0.068 


0.75 
0.059 


1.18 
0.128 


1.48 

(34) 

0.059 


1.86 
0.143 


0.36 
0.041 


T. smithi (Stations 033, 
F X 
(26) s 


038) 
1.48 
0.167 


1.80 
0.145 


0.86 
0.062 


0.85 
0.061 


1.30 
0.165 


1.50 
0.072 


1.92 
0.117 


0.41 
0.035 


M 

(19) 


X 

s 


1.48 
0.153 


1.80 
0.122 


0.85 
0.058 


0.83 
0.056 


1.30 
0.113 


1.50 
0.046 


1.92 
0.119 


0.40 
0.030 


T. minuta (Station 045) 
F X 
(11) s 


0.69 
0.061 


1.11 
0.052 


0.44 
0.036 


0.47 
0.026 


0.61 
0.054 


1.46 
0.081 


1.43 
0.085 


0.33 
0.029 


M 

(12) 


X 

s 


0.72 
0.092 


1.10 
0.070 


0.44 
0.035 


0.47 
0.033 


0.64 
0.077 


1.43 
0.098 


1.47 
0.054 


0.34 
0.033 


T. inflata (Station 043) 
F X 
(11) s 


1.51 
0.172 


1.53 
0.170 


0.86 
0.105 


0.81 
0.084 


1.26 
0.155 


1.50 



1.95 
0.204 


0.41 
0.017 


M 

(11) 


X 

s 


1.43 
0.140 


1.45 
0.140 


0.80 
0.056 


0.75 
0.064 


1.18 
0.128 


1.53 
0.090 


1.89 
0.140 


0.42 
0.024 



AUSTRALIAN SPRING HYDROBIIDS 



139 



-g о £ о о 

t CT) Ф 05 CT) 

2 с^ ""- F, £= 

W Ф (J ^ ф 

о о' "^' 1 -Í 



(О 



Ei 



=i ffl - ''^ 

£ в 5 - ^ 
О) о ъ >< В 

х: о ш л -О 

-.9-t3 "-о 

^Г о «I 

Е 2^ ^^ 

D to Р 'S о 

■5 О § ™ 

m ТЗ - С7) >= 

г; С О С « 

" ПЗ U. О) Г5 
»_ г- ТЗ 

О t >, (О 2 

*i "^ Я Х О 

!5 2 о 03 

- g-1 ' ^^ Е 
Ü о го d ё 3 
< *- clS - ^ 
о ^ •-,_'« 

о ЕУ2_с 



(О 



■•= Q) .2 



Е -8 ■§ 2 ш ° 
™ Q. __ -г; ^ 

Е ф .э ^ -g ^- 

i^t^-^E 
2 Ф о О ^ :з 
<л ^ с-" о1 

с ■ - с ^ ■?= ^ 
J ^2 CT Ol 5 

9íd 



Ф с 41 

ü го тз ^ £ 



Е ^ Ф «^ С/3 Ь' 

"гё Е Я -о ь;, '" 
= 2w S.^ií 

Ü . - .2 го 

« 5 §,-0^ ф 

1 1 2 о ^ сл 

^2|2^-Е^ 



Ч -I 



m Ü - Е ^ 

Ш ^ • 2 с _ 



_J с « Ф го с 

CÛ m <и о = я 

< й .^Z. л 2 

I- о -о " ~ " 



о 




140 



PONDER, HERSHLER & JENKINS 



0) .. ■ - 
-о -g 

ü -^ О) 
• - о О) 

•| о« 
Œa о. 



га 



с с 



ÍÍ? ш.о га 

im 



£°^"га 
BS га Í!? 

E I ^ 5 
-S, « Q." m 

с . . > 

Ф > со - 

Е Q с ^ 
3 .. 0) с 

™ -fc "î:; ^ 

^ га о о 

0) «_^-> 
CD ^°- ° 

ш о о ^ 
W é о о 
2 ?€£ 
E^g.f 
га CD хз ^ 

О) Ü о > 
^ "D "о "Ь' 

"3 > "- s 
с о ш i5 

?°EÄ 
га Q га га 

«1 Е£ 

Ш о - "D 

О ч- О '5 
ф о ^ ^ 

W £ • - > 

■Q a)-o-b' 
2 . > с 
■о > о я 
-с О .- СП 

Р XJ 3 ♦- 

к: Е -р 

Ü 
с\1 
ш ^_^ 

m SB 

< ^ ^ 
но О) 



га с «5 
CT)i5 о 
О) о. 

о о 



ф 







^ 


CT) 




со 




г^ 


о 


о 




о 
о 


to 


, , 


со 


см 


2, 


о 


о 




сЬ 


___ 






00 










m 

05 


со "^ 




о 


ел о 




о 


со 






■* 






см 
о 




(О 


to 




см 


oí Р 


г^ 




о т- 




о 


о 






Г-" 






о 






о 






(Л 




, , 


о^ 




о 














га 












œ 










о 


Ti- 




со 


co 


(О 




CD 




о 


LO 


, ^ 




о 


fV;^ 


о 


ö 






со 


^.^ 


со 


т— 


со 


о 


d 




о 
со 


со 


^ 


1Л 


•^ 


00 


см 






о 


г^ 




ел 


см 


2. 




о 




о 


о 




00 


см 




о 


Ö 




Ö 

(О 


со 




о 


см 






о 




о 


s со 
. о 




о 
со 


g 


см 

Ö 


о 






см 






о 






О) 




о 


о со 




00 


о m 






к° 




Ö 


о 






о 






см 




о 


о со 


^ 


■* 


о ,— 


2Î. 


см 


и ^ 




о 


с 






о 












(0 






^1Х 




(Л 








(Ч 






s 




s 


§1^ 




3- 


о. 

К' 







со Ol 



00 — CT) 

r^- со s о 
со о о 



;l ^ CD 

d — о 



со íT см 
,-" -^ ^ о 
00 о -^ о 



00 ю -.- 



о о 



1^' ж 

см С35 

о о 






Е и. 









.^ 




о 


^— V 


eg 






см 


о 




о 
1 




d 

1 




ю 




íñ 




см 




о 




d 




d 




г^ 








00 


s 


о 




о 








2 








о 










s 


о 




d 




d 




■^ 








о 


s 


о 




о 




in 




со 


, 






о 


со 


о 




d 




d 
о 




о 




05 




m 








d 




d 




см 








о 


s 


о 




d 




со 




■* 




см 








о 




d 




d 




оо 




OsJ 








о 




d 




d 

Lf) 




CJ 




(О 




■^ 








d 




d 

со 




LD 


, 






ю 


со 


■^ 




d 




о 










•* 








■* 








о 






со 


со" 


со 


^_^ 


г^ 


■* 


о 


со 


о 


о 


т^ 




d 


(0 








с 








о 
















га 


IX 




со 










W 
















Я 
















■Э 


и. 




^ 


^ 
















к^ 









MALACOLOGIA, 1989, 31(1): 141-156 

ULTRASTRUCTURAL CHANGES IN THE DIGESTIVE TRACT OF 

DEROGERAS RETICULATUM (MÜLLER) INDUCED BY A CARBAMATE 

MOLLUSCICIDE AND BY METALDEHYDE 

R. Thebskom 

Zoologisches Institut I, Universität Heidelberg, 
Im Neuenheimer Feld 230 
D-6900 Heidelberg, F.R.G. 

ABSTRACT 

Electron microscope investigations reveal different reactions of cells in the digestive tract of 
Deroceras reticulatum to intoxication with carbamate or metaldehyde molluscicides. All entero- 
cytes are more strongly attacked by the carbamate compound Mesurol than by metaldehyde. 
The better efficiency of Mesurol is primarily attributed to its severe impact on nuclei, leading to 
other cell damage and finally to an increased macrophage reaction. 

Metaldehyde leaves the enterocyte functions more or less intact except for that of mucus cells. 
It activates mucus extrusion immediately after the onset of intoxication. This mucus serves to 
dilute the toxin, which passes through the digestive tract and is voided. The severe attack of 
metaldehyde on the immature mucus cells results in cessation of mucus production, leading to 
a fatal mucus deficiency in the digestive tract. 

Keywords: Gastropoda; molluscicides; carbamate; metaldehyde; digestive tract; ultra- 
structure 



INTRODUCTION 

To date, the most efficient pesticides 
against slugs are carbamate compounds, 
such as Mesurol, which act as nerve toxins by 
inhibition of the Cholinesterase activity (Getzin 
& Cole, 1 964; Pessah & Sokolove, 1 983). Dur- 
ing the last decade, Mesurol has replaced met- 
aldehyde as the primary commercial mollus- 
cicide, because metaldehyde loses most of its 
efficiency in humid climates (Martin & Forrest, 
1969). In the literature (Pappas et al., 1973), 
however, it is not only Mesurol but also such 
other carbamate compounds as Carzol, 
Furadan and Zectran that are mentioned as 
having an increased efficiency compared to 
metaldehyde (Getzin, 1965; Prystupa et al., 
1987). Whereas in most investigations LD50 
tests are used (Bakhtawar & Mahendru, 
1987), there are only a few publications con- 
cerning cellular mechanisms induced by mol- 
luscicides (Ishak et al., 1970; Banna, 1977, 
1 980a, b; Pessah & Sokolove, 1 983). Up to the 
present, little attention has been paid to the 
fact that both carbamate compounds and met- 
aldehyde are in use as oral toxins (cf. Hend- 
erson, 1969), and, as a consequence, the first 
possible targets for molluscicidal action might 
be the cells of the intestinal epithelia. 



In fact, only one study covers the influence 
of molluscicidal agents on the cells and tis- 
sues of the alimentary tract of slugs after in- 
toxication (Manna & Ghose, 1972). To the 
best of my knowledge, ultrastructural investi- 
gations are completely lacking. Thus, the 
present electron microscope study was de- 
signed to investigate the different cellular re- 
sponses to molluscicidal intoxication in the di- 
gestive tract of the grey garden slug, 
Deroceras reticulatum. A further purpose of 
the paper is to elucidate the reasons for the 
superior efficiency of carbamate mollusci- 
cides by comparing the ultrastructural dam- 
age after oral application of carbamate and 
metaldehyde. 



MATERIALS AND METHODS 

Laboratory-reared specimens of Deroceras 
reticulatum were fed pellets containing 4% of 
the carbamate compound Mercaptodimethur 
(4- (methylthio) - 3,5 - xylyl - methyl -carbamate; 
Mesurol; Bayer) or 4% metaldehyde (Spiess 
Urania 2000). The pellets were weighed be- 
fore and after the slugs had fed, and the 
amount of toxicant effectively ingested was 
calculated. On an average, the animals took 



141 



142 



TRIEBSKORN 



up 200 |л.д Mesurai or 9 mg metaldehyde/g 
wet weight. Animals fed carbamate were dis- 
sected after one, five and 16 hours. The met- 
aldehyde group was fixed after five hours. For 
primary fixation a 2% glutaraldehyde solution 
in cacodylate buffer (0.01 M, pH 7.4) was in- 
jected into the body cavity. Then oesophagus, 
crop, stomach, intestine and digestive gland 
were isolated in fixative and fixed for two 
hours in 2% glutaraldehyde at 4°C. The tis- 
sues were rinsed in cacodylate buffer and 
postfixed in 1% osmium-ferrocyanide (Kar- 
novsky, 1971) for two hours. After rinsing in 
cacodylate and 0.05 M maléate buffer (pH 
5.2), the specimens were stained en bloc in 
1% uranylacetate in 0.05 M maléate buffer 
overnight at 4°C. The samples were dehy- 
drated and embedded either in Araldite or in 
Spurr's medium (Spurr, 1969). 

Ultrathin sections cut on a Reichert ultrami- 
crotome were counter-stained with lead ci- 
trate for 30 minutes and finally examined in a 
Zeiss EM 9. 



RESULTS 

Macroscopic observations 

The macroscopic reactions of the animals af- 
ter molluscicide application correspond to the 
reactions described as typical for carbamate 
or metaldehyde intoxication by Godan (1979). 
By 30 minutes after ingestion of Mesurol pel- 
lets, the animals show violent muscle convul- 
sions. The anterior body begins to swell while 
the posterior flattens. The tentacles are re- 
laxed, the animals release a lucid mucus and 
take up liquid from the environment. After 
three hours they lie almost motionless on one 
side. Usually they die 20 to 30 hours later, but 
recovery is also possible. 

After the application of metaldehyde, the 
animals lose much more slime than after car- 
bamate ingestion. In this case, muscle con- 
vulsions and relaxation of the tentacles could 
not be observed. 

Electron microscopical investigations 

Histology of the epithelia in control animals 

Oesophagus: The oesophageal epithelium 
consists of four cell types, three of which 
reach the lumen (Fig. la): 

Type I: Columnar storage cells (Figs. 2a, 5) 
characterized by high amounts of lipid and 
storage carbohydrate (glycogen or galacto- 



gen) (Fig. 34). In the central cytoplasm, the 
nucleus, small Golgi complexes, mitochon- 
dria and a few peroxisome-like vesicles are 
located, while smooth and granular endoplas- 
mic reticulum occasionally appear in basal re- 
gions of the cells. Under the microvillous bor- 
der a band of mitochondria can be found (Fig. 
10). 

Type II: Columnar secretory cells of an ec- 
crine type (Fig. la), with basally situated 
granular endoplasmic reticulum and an elab- 
orate Golgi apparatus (Fig. 29) producing 
electron-lucent secretory vesicles. Mitochon- 
dria and small amounts of lipid and glycogen 
are dispersed over the cytoplasm. The nu- 
cleus is located in the center of the cell. 

Type III: Secretory cells of a holocrine type 
(mucus producing goblet cells, in the follow- 
ing called "mucus cells") (Figs, la, 3a, 5, 25), 
with conspicious granular endoplasmic retic- 
ulum characterized by a spacious lumen, 
large Golgi apparatus and mucus vacuoles 
that merge in the apical part of the cells. The 
nuclei of these cells are situated in the basal, 
dilated regions. Young mucus cells (Fig. 3a) 
do not contain high amounts of mucus vacu- 
oles and are characterized by a conical cell 
shape. 

Type IV: Small electron-lucent cells (Fig. 5), 
conical in shape, that are dispersed amongst 
the other cells. Their apices do not reach the 
lumen. Containing characteristic lysosomes, 
dictyosomes and a prominent nucleus, they 
resemble the haemolymph macrophages. 

The basal surfaces of all cell types have no 
infoldings (Fig. 5). In addition to numerous mi- 
crovilli, the luminal surface of cell types I and II 
may bear cilia (Fig. 2a). The microvilli of the 
mucus cells are smaller than those of the 
other cell types (Fig. 3a). 

A strong muscle layer, connective tissue 
cells and nerves with different neurosecretory 
vesicles can be found subtending the epithe- 
lium (Fig. 2a, 40, 42). In longitudinal section, 
the muscle filaments are all roughly parallel, 
while in transverse section there is a quasi- 
lattice of thick and thin filaments (Fig. 40). In 
the haemolymph space some macrophages 
can be observed. They are characterized by a 
large nucleus, small Golgi apparatus and a 
few small vesicles of various electron-density 
(Fig. 2a). 

Crop: Apart from a few mucus (Type III) 
and small electron-lucent cells (Type IV), the 
cylindrical epithelium of the crop is dominated 
by a single cell type, resembling the storage 



DIGESTIVE TRACT OF DEROGERAS 



143 





CROP 



ч5^ OESOPHAGUS 




mm 
Ш- 
Ш 

Ш 
■ Щ 



STOMACH/lflTESTINE 



V VII VI 

FIG. 1 . Diagram of the digestive tract of Deroceras reticulatum illustrating the cells investigated in the present 

study. 

1a. Oesophagus: Storage cell (I), secretory cell of an eccrine type (II), secretory cell of a holocrine type, 

called mucus cell (III), and small electron-lucent cell (IV) 
1b. Crop: Storage cell 

1c. Stomach and adjacent intestine: Storage (I) and mucus cell (III) 
Id. Mid-gut gland: Digestive cell (V), crypt cell (VI), and excretory cell (VII) 



cell (Type I) of the oesophagus (Fig. lb). Only 
in regions of the crop adjacent to the stomach 
do these cells bear cilia. 

A snnall muscle layer with associated con- 
nective tissue and nerves underlies the epi- 
thelium. 



Stomach and adjacent intestine: Half of 
the stomach epithelium is made up by cells 
resembling the storage cells of the oesopha- 
gus with respect to their ultrastructural organ- 
isation and storage products (Fig. 1c). The 
cells always bear microvilli and cilia (cf. 



144 



TRIEBSKORN 



METALDEHYDE 




FIG. 2. Reconstruction of an unciliated or ciliated storage cell (2a) and its habit after carbamate (2b) and 
metaldehyde (2c) intoxication 



Walker, 1972). In the stomach crypts, the cilia 
are longer and more numerous (cf. Häffner, 
1924). The rest of the epithelium is made up 
by mucus cells (Type III). In the adjacent in- 
testine, half of the cells are mucus cells (Type 
III), and only a quarter are storage cells (Type 
I). The other quarter of the cells are secretory 
(Type II). An underlying muscle layer is well 
developed. It can be compared with that of 
the oesophagus (Fig. 40). Many nerve fibres 
can be detected. 

Mid-gut gland: The epithelium of the mid- 
gut gland is arranged in tubules that are 
bound together by a meshwork of connective 
tissue. An underlying muscle layer is lacking. 
Three cell types can be distinguished (Fig. 1d, 
4a): 

Type V: The columnar digestive cells, 
highly vacuolized absorptive cells, that domi- 
nate the epithelium. The vacuoles vary in size 
and are generally largest towards the basal 
regions of the cells. Pinocytotic vesicles de- 
velop along the apical plasma membrane, 
where endocytotic channels can also be 
found. The absorptive area is increased by 
numerous microvilli. The digestive cell cyto- 
plasm contains a little granular endoplasmic 



reticulum, a few mitochondria and an occa- 
sional small Golgi apparatus. Lipid and gly- 
cogen storage can be found. The nuclei of 
these cells are basally located. 

Type VI: The crypt cells are conical in 
shape with broad bases abutting on to the 
haemolymph space. Serving secretory func- 
tions, they are characterized by a large 
amount of granular endoplasmic reticulum 
(Fig. 22), a great number of Golgi stacks and 
secretory vesicles in the perinuclear cyto- 
plasm. The nuclei are basally situated, pos- 
sess a large nucleolus and have scattered 
patches of heterochromatin. Mitochondria are 
located near the apical and the basal surfaces 
of the cells. Lipid and carbohydrate storage, 
as well as membrane-bound spherites are 
present (Fig. 4a). The microvilli are longer 
than those of the digestive cells, and the 
basal labyrinth is well developed. 

Type VII: The goblet-like excretory cells are 
characterized by large and small vacuoles 
containing electron-dense material (Fig. 27). 
The membrane of the large vacuole shows 
numerous infoldings. In the cytoplasm a small 
Golgi apparatus, a small amount of smooth 
and granular endoplasmic reticulum and lipid, 
and a few mitochondria can be found. The 



DIGESTIVE TRACT OF DEROGERAS 



145 



Ь , CARBAMATE 



(^ IMETALOEHYDE 




fully differentiated mucus cell young mucus cell 



FIG. 3. Reconstruction of a young and a fully differentiated mucus cell in control animals (3a), after car- 
bamate (3b), and after metaldehyde intoxication. 



microvilli of these cells are as long as those of 
the crypt cells. 

To compare the different theories concern- 
ing the genealogy of these cells and the dif- 
ferent nomenclatures, see David & Götze 
(1963) and Walker (1970). 

Histopathological alterations (Figs. 2, 3, 4) 

Generally speaking, the cytological reac- 
tions in the digestive tract originate in isolated 
cells and spread over the epithelium during 
the following hours. Sixteen hours after inges- 
tion of the molluscicides, a high percentage of 
the cells are significantly damaged. 

The reactions observed after five and 16 
hours generally resemble those after one 
hour, but they are more intense. 

Reactions that appear in the anterior part of 
the digestive tract immediately after the mol- 
luscicide treatment became apparent in cells 
of the posterior part with a time lapse corre- 
sponding to the transport rate of toxic food- 
stuff. 

Sixteen hours after carbamate ingestion, a 
lot of cells have been extruded from the epi- 
thelium. 

Reactions of the basal and apical cell 
surfaces 

MESUROL: Immediately after the applica- 
tion of Mesurol, the basal surfaces of storage 
and secretory cells (Type I and II) are slightly 



stretched (Comp. Figs. 5 and 6). After five 
and 16 hours, the cells exhibit considerably 
extended basal cell extensions (Fig. 7) that 
sometimes contact nerve or muscle cells (Fig. 
8). In the mid-gut gland the basal cell exten- 
sions are less distinct than in the alimentary 
tract. However, the basal labyrinth of crypt 
cells is distended, and the intercellular spaces 
are enlarged (Fig. 35). 

Comparable to the reactions of the cell 
bases, the apical surfaces of the cells react 
very quickly with a reduction of microvilli and 
cytoplasmic protrusions in the anterior, and 
after five or 16 hours in the posterior parts of 
the digestive tract (Figs. 10, 11, 12). 

An intensified vacuolization in the digestive 
cells often leads to a breakdown of the apical 
membrane. 

METALDEHYDE: After intoxication with 
metaldehyde, basal cell extensions are lack- 
ing; the basement membrane is thickened 
and becomes more electron-dense (Fig. 9). 

Protrusions of the apical cytoplasm and re- 
duction of microvilli can occasionally be 
found. 

After intoxication with carbamate and met- 
aldehyde, the shape of all cell types becomes 
more irregular (Figs. 2, 3, 4). 

Reactions of the cytoplasm 

MESUROL: After carbamate intoxication, 
the cytoplasm of storage and secretory cells 
appears slightly condensed (Fig. 7) or elec- 



146 



TRIEBSKORN 



METALDEHYDE 




digestive cell 



crypt cell 



FIG. 4. Reconstruction of a digestive and a crypt cell of the mid-gut gland (4a) and its variable habit after 
carbamate (4b) and metaldehyde (4c) ingestion. 



tron-lucent (Fig. 13). In the digestive cells of 
the mid-gut gland, it is either extremely elec- 
tron-lucent, or totally electron-dense, depend- 
ing on the degree of vacuolization. Electron- 
dense cytoplasmic areas often surround the 
nuclei. 

METALDEHYDE: In mucus cells the cyto- 
plasm is displaced by the enlarged mucus 
vacuoles (Fig. 33). In all cell types it appears 
less electron-dense. 

Reactions of the nuclei 

MESUROL: One hour after intoxication the 
nuclei are severely damaged. 

The karyoplasm becomes less electron- 
dense (Fig. 13, 14, 15), lipid droplets can be 
detected in it (Fig. 15), the nucleoli are irreg- 
ularly deformed (Fig. 16), and the amount of 
heterochromatin is reduced. In some cases, 
the karyoplasm appears totally condensed 
(Fig. 17), or, especially in crop cells, small 
vesicles can be found in it (Fig. 16). Mitotic 
processes are evident. However, even after 
16 hours, there are still some unafflicted nu- 
clei next to totally damaged ones (Fig. 17), 
emphasizing the heterogeneity in cellular re- 
action. 

METALDEHYDE: After metaldehyde in- 
gestion, damage to the nuclei is less intense 
than after carbamate application. The karyo- 
plasm becomes less electron-dense, and in a 
few cases it bears lipid droplets (comp. Fig. 
1 4). Especially in the crypt cells of the mid-gut 



gland, the amount of heterochromatin is re- 
duced. 

Reactions of the mitochondria 

MESUROL: After carbamate intoxication, 
mitochondrial effects originate in the oesoph- 
agus and crop cells from the cell apex, while 
in the posterior parts of the digestive tract mi- 
tochondria located in the basal cytoplasm are 
afflicted earlier. 

Especially in the stomach and the digestive 
gland, electron-dense granules different from 
the common intramitochondrial granules can 
be found in mitochondria, located in mem- 
brane-bound compartiments (Fig. 18). Fur- 
thermore, the organelles are heavily inflated 
and their cristae are reduced (Fig. 19). 

METALDEHYDE: After metaldehyde in- 
toxication, the mitochondria are less afflicted 
than after carbamate ingestion. The common 
intra-mitochondrial granules are often en- 
larged (Fig. 20), and only in a few cases the 
organelles are swollen. 

Reactions of the endoplasmic reticulum 

MESUROL: After Mesurol application, the 
smooth and granular endoplasmic reticulum 
proliferates in basal regions of storage, diges- 
tive and crypt cells. In most cases, the smooth 
endoplasmic reticulum is heavily distended 
(Fig. 21). Degranulation of granular endoplas- 
mic reticulum can be observed in basal re- 



DIGESTIVE TRACT OF DEROGERAS 



147 



'^^^ ^J i 




/ 



® ($ ':f?i 



í -я 



/ 






4v 



mv 






-y- 







FIG 5 Oesophagus, control: Section through the basal part of the oesophagus epitheliurn showing a глисиз 

(mue), storage (sc) and small light cell (sic). There are no infoldings of the cell basis (arrows). 

FIG. 6. Oesophagus, Mesurol, 1 h: Slight basal extensions (arrows). 

FIG. 7. Oesophagus, Mesurol, 5 hs: Strong basal extensions (arrows). 

FIG. 8. Oesophagus, Mesurol, 1 h: Contact of a basal cell extension (be) to a nerve cell (nc.arrow). 

FIG 9. Crop, metaldehyde, 5 hs: Thickening of the basal membrane (bm). 

FIG 10 Crop control: Apex of a storage cell, showing regularly orientated microvilli (mv). 

FIG 11 Oesophagus, Mesurol, 1 h, storage cell: Slight cytoplasm extrusions (ce, arrows). 

FIG 12 Oesophagus, Mesurol, 1 h, storage cell: Intensified cytoplasm extrusions (ce, arrows). 

FIG. 13. Oesophagus, Mesurol, 16 hs, secretory cell: Electron-lucent cytoplasm (cyt) and nucleus (n) with 

the heterochromatin reduced (arrows). 



148 



TRIEBSKORN 









{i 



n 



-^**-^ 
^•*< 






^ 



n VV. 






» о- 











-1^ 










FIG. 14. Oesophagus, Mesurol, 5 hs, storage cell: Nucleus (n) with a totally dissolved karyoplasm. 

FIG. 15. Stomach, Mesurol, 5 hs: Nuclei with lipid inclusions (li) in an electron-light karyoplasm. 

FIG. 16. Crop, Mesurol, 1 h: Nucleus with an irregularly formed nucleolus (nu) and small vesicles (v) in the 

karyoplasm (inset, x6). 

FIG. 17. Crop, Mesurol, 16 hs: Totally damaged nucleus next to an unafflicted one in the adjacent crop cell. 

FIG. 18. Mid-gut gland, Mesurol, 16 hs, crypt cell: Mitochondrion (m) with inclusions in membrane-bound 

areas (arrows); common intramitochondrial granules (gm) are visible. 

FIG. 19. Oesophagus, Mesurol, 1 h, secretory cell: Destruction of cristae in mitochondria (arrows). 

FIG. 20. Stomach, metaldehyde, 5 hs, storage cell: Mitochondrion with enlarged intramitochondriai granules 

(arrows). 

FIG. 21. Intestine, Mesurol, 5 hs, storage cell: Dilation of smooth endoplasmic reticulum (ser) in the basal 

parts of the cell. 

FIG. 22. Mid-gut gland, control, crypt cell: Cisternae of the granular endoplasmic reticulum (ger). 



DIGESTIVE TRACT OF DEROGERAS 



149 



gions of the crypt cells (comp. Figs. 22 and 
23). 

In mucus cells the granular endoplasmic re- 
ticulum is dilated and disorientated. After sev- 
eral hours, the membranes become frag- 
mented. 

METALDEHYDE: The damage to the 
granular endoplasmic reticulum in mucus 
cells is more intense after metaldehyde than 
after carbamate intoxication. The cisterna are 
heavily dilated, and the membranes are 
disarranged, ruptured and sometimes coiled 
to form myéline figures (comp. Figs. 25 and 
26). 

In the crypt and excretory cells of the mid- 
gut gland, the granular endoplasmic reticulum 
disintegrates into short cisternae with frag- 
mented membranes, frequently devoid of h- 
bosomes (Fig. 28). The cisternae of the en- 
doplasmic reticulum often form fingerprint-like 
structures (Fig. 24). In many instances, the 
membranes are ruptured. 

Reactions of the Golgi apparatus 

MESUROL: In the oesophagus, the Golgi 
complexes of the secretory cells (Type II) are 
heavily damaged. Especially the trans-face 
cisternae are either compressed or highly in- 
flated (comp. Figs. 29 and 30). Associated 
with this is a reduction in the number of secre- 
tory vesicles. 

In the mucus cells, the cisternae of the 
Golgi apparatus are strongly dilated (Fig. 31), 
and the regular arrangement of the Golgi 
stacks is often lost. 

METALDEHYDE: Except for the mucus 
cells, the reaction of the Golgi apparatus is 
less intense after metaldehyde than after car- 
bamate ingestion. The trans-faces of the cis- 
ternae are slightly dilated. In the mucus cells, 
the Golgi cisternae are swollen, often the 
membranes are arranged as concentric 
whorls (Fig. 32), and the mucus containing 
vacuoles are enlarged (Fig. 33). 

Alteration of storage products 

MESUROL: After carbamate ingestion, 
compact areas containing glycogen or galac- 
togen can be found between lipid droplets, 
especially in central and basal parts of diges- 
tive cells and in crypt cells of the mid-gut 
gland (Fig. 35). 



METALDEHYDE: In storage, secretory, di- 
gestive, crypt and excretory cells, the 
amounts of lipid and glycogen are reduced 
after metaldehyde poisoning (comp. Figs. 34 
and 36). This reduction is related to the pres- 
ence of vesicles containing electron-dense 
material with a typical lamellar fine-structure 
(Fig. 37). Furthermore, an increased number 
of peroxisome-like structures appears. 

Reaction of the cytoskeleton 

MESUROL: After Mesurol intoxication, no 
reaction of the cytoskeleton is visible. 

METALDEHYDE: In the center of storage 
and secretory cells (Type I and II), condensed 
actin-like microfilaments appear (Figs. 38, 
39). 

Reactions of the underlying muscle, 
connective and nerve tissues 

MESUROL: After the application of the 
carbamate molluscicide, the muscle tissue is 
fragmented, and the regular arrangement of 
the muscle filaments is disturbed (comp. Figs. 
40 and 41). Granules similar to peroxisomes 
(with regard to their size and their electron- 
density) appear in connective tissue cells 
showing intensive contact to smooth muscle 
and nerve cells, as well as in nerve cells 
themselves (Figs. 42, 43). 

METALDEHYDE: After application of me- 
taldehyde, no reactions of muscle, nerve and 
connective tissue could be found. 

Reactions of the macrophages 

MESUROL: After ingestion of Mesurol, the 
number of macrophages in the haemolymph 
space increases. In many cases, mitotic pro- 
cesses can be observed (Fig. 44). Particularly 
16 hours after intoxication, many of these 
cells penetrate the epithelium (Fig. 45). Other 
macrophages contain membrane fragments 
incorporated into vacuoles (Fig. 46). 

METALDEHYDE: Reactions of the macro- 
phages are lacking. 



DISCUSSION 

The present study reveals the impact of the 
carbamate compound Mesurol and of metal- 



150 



TRIEBSKORN 



@ 



.V • ,v • i:; 





















© 




@с^ 


5» * 


г 


^1¿- 


у' 




- > muv 






* 



^ 



да 






ger 




' ( 
л 

ser. <3 



1//m 



да 






о, Sum 



1 4." '' ^. * 



FIG. 23. Mid-gut gland, Mesurai, 1 h, crypt cell: Degranulation of granular ER in the basal part of the cell 

(arrows). 

FIG. 24. Mid-gut gland, metaldehyde, 5 hs, crypt cell: Granular ER (ger) forming fingerprint-like structures. 

FIG. 25. Oesophagus, control, mucus cell: Elaborate Golgi system (ga), granular ER (ger) and mucus 

vacuoles (muv) in a regular arrangement. 

FIG. 26. Oesophagus, metaldehyde, 5 hs, mucus cell: Dilated, electron-lucent cisterae of the granular ER 

and Golgi membranes (ga) forming concentric circles in a cellular disarrangement; nucleus (n) with an 

electron-lucent caryoplasm. 

FIG. 27. Mid-gut gland, control, Excretory cell: excretory cell with excretory granule (eg). 

FIG. 28. Mid-gut gland, Mesurol, 5 hs, crypt cell: Short, fragmentary cisternae of the granular ER (ger) 

frequently devoid of ribosomes. 

FIG. 29. Oesophagus, control, secretory cell: Trans- and cis-face of the Golgi apparatus (ga) can clearly be 

distinguished. 

FIG. 30. Oesophagus, Mesurol, 1 h, secretory cell: Heavily inflated cisternae of a Golgi apparatus (ga). 



DIGESTIVE TRACT OF DEROGERAS 



151 




FIG. 31. Stomach, Mesurai, 5 hs, mucus cell: Golgi apparatus (ga) with dilated cisternae (arrows). 

FIG. 32. Oesophagus, metaldehyde, 5 hs, mucus cell: Membranes of the Golgi apparatus (ga) are arranged 

as concentric whorls surrounding Golgi vesicles. 

FIG. 33. Stomach, metaldehyde, 5 hs: General view of the mucus cells after metaldehyde application; dilated 

cisternae of the Golgi apparatus and the granular ER, large mucus vacuoles (muv) and nuclei (n) with an 

electron-lucent karyoplasm. 

FIG. 34. Oesophagus, control, storage cell: High amount of stored lipid (li) and glycogen or galactogen (gl). 

FIG. 35. Mid-gut gland, Mesurol, 5 hs, digestive cell: Condensation of glycogen (gl) and dilation of the sER 

and the basal labyrinth (bl). 

FIG. 36. Oesophagus, metaldehyde, 5 hs, storage cell: Lipid and carbohydrate reduction; electron-dense 

vesicles (ev) and ER membranes forming a fingerprint-like structure (arrow). 

FIG. 37. Oesophagus, metaldehyde, 5 hs, storage cell: Electron-dense vesicles (ev) characterized by a 

typical lamellar fine-structure (inset, x 4) and by peroxisome-like organelles (p) in the center of the cell. 

FIG. 38. Crop, metaldehyde, 5 hs: Reduction of storage products and aggregation of actin-like filaments 

(arrows). 

FIG. 39. Crop, metaldehyde, 5 hs: Actin-like filaments (af) in the center of the enterocytes. 



152 



TRIEBSKORN 



^"^^^r.'éi 



ъ^: 






^-ä^""***""' 




е. . ♦ 











-^.. 



#i. 






- -f 

2 um 



FIG. 40. Oesophagus, control: Muscle cell (me) with regularly orientated muscle fibres. 

FIG. 41. Oesophagus, Mesurol, 5 hs: Fragmentation of muscle tissue in isolated portions and irregular 

orientation of muscle fibres. Intense contact of nerve (nc) and muscle cell (mc). 

FIG. 42. Oesophagus, Mesurol, 1 h: Nerve (nc) and connective tissue cells (cc) containing electron-dense, 

peroxisome-like structures. 

FIG. 43. Oesophagus, Mesurol, 5 hs: Nerve with peroxisome-like structures (p). 

FIG. 44. Oesophagus, Mesurol, 5 hs: Mitosis taking place in a macrophage (ma). 

FIG. 45. Oesophagus, Mesurol, 16 hs: Macrophage bearing membrane fragments in vacuoles (arrows). 

FIG. 46. Oesophagus, Mesurol, 1 h: Two macrophages that penetrate the epithelium. 



DIGESTIVE TRACT OF DEROGERAS 



153 



dehyde on the ultrastructure of the digestive 
tract of Deroceras reticulatum. In addition to 
the results of Tegelsstrom & Wahren (1972), 
Godan (1979) and Pessah & Sokolove 
(1983), who attributed the molluscicidal ef- 
fects to influences on Cholinesterase activity 
and water regulation, it can clearly be dem- 
onstrated that both chemicals interact with 
several different types of enterocytes. 

As evident by the present cytological find- 
ings, the carbamate compound Mesurol is ab- 
sorbed immediately after ingestion in the an- 
terior parts of the digestive tract, i.e. in the 
oesophagus and the crop. The rapid reac- 
tions in these two regions and the quick and 
intense reaction of the whole animal one hour 
after feeding support the findings of Fretter 
(1 952), Walker (1 972) and Horst et al. (1 986), 
who describe the high absorptive activity of 
the crop using radioactive labelling and bio- 
chemical methods. With regard to the time 
lapse resulting from the passage of food 
through the intestinal tract, the cellular re- 
sponses after one hour in the anterior parts 
can be compared with those in the posterior 
parts of the digestive system after five hours. 
Analysis of cellular responses to metaldehyde 
and carbamate poisoning after five hours al- 
lows three types of reaction to be distin- 
guished: carbamate-specific reactions, metal- 
dehyde-specific reactions, and cell responses 
that appear in both experiments but with dif- 
ferent intensity. 

Reactions such as cytoplasm condensa- 
tion, cytoplasmic protrusions also called 
"surface blebs" (Whyllie, 1981; Réz, 1986), 
reduction of microvilli, mitochondrial swelling 
(Goyer & Rhyne, 1975;Triebskorn, 1988) and 
dilation of Golgi cisterna (Triebskorn, 1988), 
endoplasmic reticulum and intercellular 
spaces (Smuckler & Arcasoy, 1969), as well 
as ER-membrane proliferation or destruction 
(Moore, 1979, 1985; Nott & Moore, 1987) re- 
semble those described in bivalves and ver- 
tebrates as cellular stress symptoms after in- 
toxication with different xenobiotics. Likewise, 
the degranulation of the granular ER and the 
formation of membrane whorls or myelin-like 
membranes by ER are discussed as general 
changes of the cell in response to toxicants 
(Réz, 1986). Most of these reactions are at- 
tributed to membrane destabilization and in- 
creased membrane permeability to ions un- 
der the influence of toxicants, followed by 
osmotic effects and finally cell death (Sparks, 
1972). 

Swelling of mitochondria is suggested to be 



the result of an increased Ca^+ influx (Packer 
et al., 1967; Smuckler & Arcasoy, 1969). 
Bayne et al. (1985), Moore (1985) and Nott & 
Moore (1987) relate the sER proliferation to 
an increase of sER-bound detoxification en- 
zymes, such as the NADPH-neotetrazolium- 
reductase and many others. 

In the digestive tract of Deroceras reticula- 
tum, such unspecific reactions are more in- 
tense after carbamate than after metaldehyde 
treatment except for the mucus cells of oe- 
sophagus, stomach and intestine. This might 
arise from the fact that the amount of metal- 
dehyde taken up by the animals is supposed 
to be closer to the sublethal dose than that of 
carbamate. Therefore, after Mesurol intoxica- 
tion, many more cellular reactions associated 
with cell death are involved. However, the 
comparison of reactions to different mollusci- 
cides in several regions of the digestive tract, 
as well as in several cell types at three times 
allows reactions of general nature to be dis- 
tinguished from specific ones. Thus, for ex- 
ample, in mucus cells the destruction of Golgi 
cisternae, rER and mitochondria is more 
prominent after metaldehyde than after car- 
bamate ingestion. This severe impact of met- 
aldehyde on the mucus producing cells cor- 
relates well with the known influence of this 
molluscicide on water regulation. Metalde- 
hyde enhances the extrusion of available mu- 
cus immediately after intoxication. This mu- 
cus might serve to dilute the toxin but may 
also have the capacity to detoxify it (Trieb- 
skorn, 1988). It passes the digestive tract and 
is voided quickly due to the intensified mucus 
extrusion of the whole animal. Furthermore, 
the replacement of the necessary mucus is 
blocked by destruction of the cellular secre- 
tory apparatus, especially in immature cells. If 
the resulting loss of liquid cannot be compen- 
sated, the animal will desiccate. Therefore, 
the effects of metaldehyde are reversible in a 
humid climate. 

These considerations are in line with other 
investigations studying the advantages of car- 
bamates compared with metaldehyde using 
LD50 tests (Riemschneider & Heckel, 1979; 
Prystupa et al., 1987). They support the find- 
ings of Getzin & Cole (1964), who postulate 
the effect of metaldehyde to be the result of 
water loss by stimulation of mucus secretion. 

Furthermore, metaldehyde poisoning re- 
duces cellular lipids, increases the number of 
electron-dense vesicles and peroxisome-like 
particles, and leads to a thickening of the 
basement membrane and the condensation 



154 



TRIEBSKORN 



of actin-like filaments In the cytoplasm. Until 
now, there are no targets known for the attack 
of toxins in the cytoskeletal system. Nor is 
there any intelligible explanation for the thick- 
ening of the basement membrane. 

With regard to lipid reduction, my own en- 
zyme-histochemical studies have shown that 
catalase activity can be found in the periphery 
of lipid droplets after molluscicide intoxication 
(Triebskorn, in prep.). One might speculate 
that the observed lipid reduction is correlated 
with lipid peroxidation (cf. Tappel, 1975). To 
reinforce this idea, the presence of detoxifica- 
tion products resulting from such reactions as 
well as the nature of the electron-dense ves- 
icles and the peroxisome-like structures 
should be investigated. Beyond this, the re- 
action of macrophages and connective tissue 
cells after intoxication with Mesurol could be 
of interest. While the present study demon- 
strates an increased number of peroxisome- 
like structures in connective tissue cells after 
Mesurol intoxication, Sminia (1972) was able 
to demonstrate peroxidase activity in hae- 
molymph cells, localized in similar per- 
oxisome-like vesicles. In addition, my own 
light-microscope investigations reveal an in- 
tensified catalase activity in the connective 
tissue underlining the epithelium of the diges- 
tive tract and in the haemolymph, where mac- 
rophages can be found (Triebskorn, in prep.). 
Because peroxidative reactions are known to 
be involved in detoxification (Belding et al., 
1970; Recknagel, 1967), I assume that sim- 
ilar processes are of importance in the diges- 
tive tract of slugs after molluscicide intoxica- 
tion. Whereas they might be found after 
metaldehyde treatment in the enterocytes 
themselves, carbamate poisoning leads to a 
disturbance of essential functions of these 
cells so quickly that detoxification processes 
cannot be established in them in time. This 
deficiency might be compensated by mac- 
rophage activity. The function of these cells in 
detoxification is also verified by the results of 
fVloore (1979). He could demonstrate the ac- 
tivation of MFO-enzymes in the haemolymph 
cells of Mytilus edulis after polycyclic hydro- 
carbon poisoning. Furthermore, the haemo- 
cytes are known to be able to penetrate the 
gut epithelium and to phagocytise decaying 
cells (Sminia, 1972). 

The macrophage reaction, nucleic damage 
and a changed glycogen metabolism appear 
one hour after the ingestion of the carbamate. 
The Mesurol attack on the nucleus seems to 
be the most important effect of this chemical 



and accounts for its better molluscicidal activ- 
ity compared with metaldehyde. While kary- 
olysis is often described as a late reaction to 
intoxication in vertebrates and invertebrates 
(Bayne et al., 1985), the reaction of nuclei in 
the present study does not seem to be an 
ultimate one revealing cell death, but a pri- 
mary cell response that leads to cell death. 
Since, in most cells with damaged nuclei, 
other cell death symptoms are lacking, the 
heterochromatin disorganization and in- 
crease in mitotic activity seem to reflect a 
central process in intoxication. The damage 
to the nuclei could explain the reinforcement 
of other cellular responses, such as the dis- 
turbances in glycogen metabolism or the de- 
struction of Golgi apparatus, already de- 
scribed by Flickinger (1971). 

In contrast to these reactions, the formation 
of basal cell extensions requires several 
hours. This is probably due to the inhibition of 
Cholinesterase activity by carbamates. The 
blockade of the esterase center of the en- 
zyme (Wegler, 1970) interrupts nerve stimuli 
conduction, leading to uncontrolled muscle 
contraction and finally to muscle atony. Un- 
controlled contractions in the muscle layer 
that underlies the gut epithelium are respon- 
sible for the basal cell deformation. 

In conclusion, the present results suggest 
that the cells might respond to environmental 
stress in different ways. Lipid mobilization in 
storage cells, for example, only takes place 
after intoxication with less effective mollusci- 
cides, such as metaldehyde. Carbamate in- 
gestion stimulates the activity of the macro- 
phages. Furthermore, peroxidative reactions 
have been suggested to be the biochemical 
pathway of detoxification of carbamate and 
metaldehyde. According to this opinion, the 
detoxification processes in molluscs resemble 
those described for insects by Wegler (1970). 

The present electron microscope study was 
able to extend former observations on whole 
animal behavior following intoxication. Sug- 
gestions about cellular mechanisms induced 
by different molluscicides are developed. 
However, other techniques, such as enzyme 
histochemistry, biochemistry and autoradiog- 
raphy, are necessary to further specify these 
results and to give more detailed information 
about the function of the structures described. 

ACNOWLEDGEMENTS 

This work was partly supported by the Ger- 
man Research Council (DFG Sto 75/9; Ja 407/ 



DIGESTIVE TRACT OF DEROGERAS 



155 



1-1). Personal thanks go to Günter Vogt, 
Thomas Braunbeck and Ulrich Bielefeld for 
the revision of the paper and for help with the 
English version. Thanks are also due to Dr. 
Janssen, to Prof. Storch and to Dr. Künast 
(BASF), who encouraged the present investi- 
gation and provided laboratory facilities and 
chemicals. 



LITERATURE CITED 

BAKHTAWAR, P. M. & V. K. MAHENDRU, 1987, 
Toxicity of organophosphorus molluscicides on 
the snail pest Helix vittata (Hornell). Journal of 
Advanced Zoology, 8(2): 84-87. 

BANNA, H. B. M., 1977, The general histology and 
histopathology of Bulinus truncatus (Andouin) 
with special reference to action of a molluscicide. 
Ph.D. Thesis, London University. 

BANNA, H. B. M. 1980a, Histochemical studies of 
some enzymes in the tissues of the schistosome 
vector snail Bulinus truncatus (Andouin) with 
special reference to the effects of a molluscicide. 

I. Dehydrogenases. Histociiemical Journal, 12: 
139-144. 

BANNA, H. B. M., 1980b, Histochemical studies of 
some enzymes in the tissues of the schistosome 
vector snail Bulinus truncatus (Andouin) with 
special reference to the effects of a molluscicide. 

II. Hydrolases. Histochemical Journal, 12: 145- 
152. 

BAYNE, B. L., D. A. BROWN, K. BURNS, D. R. 
DIXON, A. IVANOVICI, D. R. LIVINGSTONE, 
D. M. LOWE, M. N. MOORE, A. R. D. STEBBING 
& J. WIDDOWS, 1985, The effects of stress and 
pollution on mahne animals. Praeger Scientific, 
New York. 

BELDING, M. E., S. J. KLEBANOFF & G. G. RAY, 
1970, Peroxidase-mediated virucidal systems. 
Science, 167: 195-196. 

DAVID, H. & J. GÖTZE, 1963, Elektronenmikrosko- 
pische Befunde an der Mitteldarmdrüse von 
Schnecken. Zeitschrift für mikroskopische und 
anatomische Forschung, 70: 252-272. 

FLICKINGER, C. J., 1971, Alterations in the Golgi 
apparatus of amoebae in the presence of an in- 
hibitor of protein synthesis. Experimental Cell 
Research, 68: 381-387. 

FRETTER, V. 1952, Experiments with P^^ and V^' 
on species of Helix, Arion and Agholimax. Quar- 
terly Journal of microscopical Science, 93: 135- 
146. 

GETZIN, L. W. & S. G. COLE, 1964, Evaluation of 
potential molluscicides for slug control. Washing- 
ton agricultural experiment stations bulletin, 658: 
1-9. 

GETZIN, L. W., 1965, Control of the grey garden 
slug with bait formulations of a carbamate mol- 
luscicide. Journal of Economic Entomology, 58: 
158-159. 



GODAN, D., 1979, Schadschnecken. Ulmer Ver- 
lag, Stuttgart. 

GOYER, R. A. & B. С RHYME, 1975, Toxic 
changes in mitochondrial membranes and mito- 
chondrial function, in: Trump, B. F. & A. U. Ar- 
stila, eds., Pathology of all membranes, Vol.1. 
New York, 383-428. 

HAFFNER, K., 1924, Über den Darmkanal von He- 
lix pomatia (L.). Zeitschrift für wissenschaftliche 
Zoologie, 121: 126-169. 

HENDERSON, I. F., 1969, A laboratory method for 
assessing the toxicity of stomach poisons to 
slugs. Annals of Applied Biology, 63: 167-171. 

HORST, С, W. BECKER & A. KEMPER, 1986, 
Short-term alterations of the ketone body content 
in the hemolymph of Biomphalaria glabrata (Gas- 
tropoda; Pulmonata). Comparative Biochemistry 
and Physiology, 84B: 555-557. 

ISHAK, M., A. SHARAF, A. MOHAMED & A. 
MOUSA, 1970, Studies on the mode of action of 
some molluscicides on the snail Biomphalaria 
elxandrina. Comparative genetic Pharmacology, 
1:201-208. 

KARNOVSKY, M. J., 1971, Use of ferrocyanide- 
reduced osmium tetroxide in electron micros- 
copy. Journal of Cell Biology, 51 : Abstr. 284. 

MANNA, B. & K. С CHOSE, 1972, Histo-patholog- 
ical changes in the gut of Achatina fúlica caused 
by Endrin, a molluscicide. Indian Journal of ex- 
perimental Biology, 10(6): 461-463. 

MARTIN, T J. & J. D. FORREST, 1969, Zur Ent- 
wicklung von Schneckenkorn Mesurol in 
Großbritannien. Pflanzenschutz-Nachrichten 
Bayer, 22(2): 212-253. 

MOORE, M. N., 1979, Cellular responses to poly- 
cyclic aromatic hydrocarbons and phénobarbital 
in Mytilus edulis. (Marine Environmental Re- 
search, 2: 255-263. 

MOORE, M. N., 1985, Cellular responses to pollu- 
tants. Marine pollution bulletin, 16(4): 134-139. 

NOTT, J. A. & M. N. MOORE, 1987, Effects of poly- 
cyclic aromatic hydrocarbons on molluscan ly- 
sosomes and endoplasmatic reticulum. His- 
tochemical Journal, 19: 357-368. 

PACKER, L., D. W. DEAMER & R. L HEATH, 
1967, Regulation and deterioration of structure in 
membranes. In: Strehler, ed.. Advances in ger- 
ontological research, Vol.2, 77-120. 

PAPPAS, J. L., G. E. CARMAN & G. F. WOOD, 
1973, Weather effects on baits on controlling eu- 
ropean brown garden snails in citrus. California 
Agriculture, 27: 13-15. 

PESSAR, J. N. & P. G. SOKOLOVE, 1983, The 
interaction of organophosphate and carbamate 
insecticides with cholinesterases in the terrestrial 
pulmonate Umax maximus. Comparative Bio- 
chemistry and Physiology, 74c(2): 291-297. 

PRYSTUPA, B. D., N. J. HOLLIDAY & R. B. WEB- 
STER, 1987, Molluscicide efficacy against the 
march slug Dereceras laeve (Stylommatophora; 
Limacidae) on strawberries in Manitoba. Journal 
of Economic Entomology, 80(4): 935-943. 

RECKNAGEL, R. O., 1967, Carbon tetrachloride 



156 



TRIEBSKORN 



hepatotoxicity. Pharmacological Review, 19: 
145. 

RÉZ, G., 1986, Electron microscopic approaches to 
environmental toxicity. Acta Biológica Hungarica, 
37(1): 31-45. 

RIEMSCHNEIDER, R. & A. MECKEL, 1979, Deter- 
mination of the lethal dose of Metaldehyde and 
evidence of a synergetic effect in snails. 
Zeitschrift für Pflanzenkranl<heiten und Pflanzen- 
schutz, 86(8): 479-482. 

SMINIA, T., 1972, Structure and function of blood 
and connective tissue cells of the freshwater pul- 
monate Lymnaea stagnalis studied by electron 
microscopy and enzyme histochemistry. 
Zeitschrift für Zellforschung, 130: 497-526. 

SMUCKLER, E. A. & M. ARCASOY, 1969, Struc- 
tural and functional changes of the endoplasmic 
reticulum of hepatic parenchymal cells. Interna- 
tional Review of Experimental Pathology, 7: 
305-418. 

SPARKS, A. K., 1972, Invertebrate pathology, non- 
communicable diseases. Academic Press, New 
York, London. 

SPURR, A. R., 1969, A low viscosity embedding 
medium for electron microscopy. Journal of Ul- 
trastructural Research, 26: 31-43. 

TAPPEL, A. L., 1975, Lipid peroxidation and fluo- 
rescent molecular damage to membranes. In: 
Trump, B. F. & A. Arstila, eds., Pathobiology of all 
membranes. Vol. I Academic Press, New York. 

TEGELSSTROM, H. & H. WAHREN, 1972, The ef- 
fect of an N-methyl carbamate on esterases from 
snail, mouse and man studied by starch-gel elec- 
trophoresis. Comparative Biochemistry and 
Physiology, 43B: 339-343. 

TRIEBSKORN, R., 1988, Molluskizid-induzierte 
Reaktionen im Verdauungstrakt von Dereceras 
reticulatum (Müller), Verhandlungen der deut- 
schen Zoologischen Gesellschan, 81 : 332. 

WALKER, G., 1970, The cytology, histochemistry 
and ultrastructure of the cell types found in the 
digestive gland of the slug Agriolimax reticulatus 
(Müller). Protoplasma, 71: 91-109. 

WALKER, G., 1972, The digestive system of the 
slug Agriolimax reticulatus (Müller): Experiments 
on phagocytosis and nutrient absorption. Pro- 
ceedings of the malacological Society of London, 
40: 33-43. 

WEGLER, R., 1970, Chemie der Pflanzenschutz- 
und Schädlingsbekämpfungsmittel. Bd 1. 
Springer Verlag, Heidelberg, New York. 

WHYLLIE, A. H., 1981, Cell death: A new classifi- 
cation separating apoptosis from necrosis. In: 



Bowen, I. D. & R. A. Lochshin, eds., Cell death in 
biology and pathology. Capman & Hall, London, 
New York. 

Revised Ms. accepted 20 June 1989 



ABBREVIATIONS 



af 


actin-like filament 


be 


basal extension 


bl 


basal labyrinth 


bm 


basal membrane 


с 


cilia 


ce 


connective tissue cell 


ce 


cytoplasmic extrusion 


eg 


calcium granule 


cyt 


cytoplasm 


dv 


digestive vacuole 


ec 


endocytotic channel 


eg 


excretory granule 


env 


endocytotic vesicle 


ev 


electron-dense vesicle 


ga 


golgi apparatus 


ger 


granular endoplasmic reticulum 


gl 


glycogen 


gm 


intramitochondrial granule 


imv 


intramitochondrial vesicle 


II 


lipid 


ly 


lysosome 


ma 


macrophage 


mc 


muscle cell 


mf 


membrane fragments 


mi 


mitochondria 


mit 


mitosis 


mue 


mucus cell 


muv 


mucus vacuole 


mv 


microvilli 


n 


nucleus 


no 


nerve cell 


nu 


nucleolus 


nv 


neurosecretory vesicle 


P 


peroxisome-like vesicle 


sc 


storage cell 


sec 


small electron-lucent cell 


ser 


smooth endoplasmic reticulum 


SV 


secretory vesicle 



MALACOLOGIA, 1989, 31(1): 157-173 

RETRACTION/EXTENSION AND MEASUREMENT ERROR IN A LAND SNAIL: 
EFFECTS ON SYSTEMATIC CHARACTERS 

Kenneth С Emberton 

Department of Malacology, Academy of Natural Sciences, 19thi and the Parkway, 
Philadelphia, Pennsylvania, U.S.A. 19103 

ABSTRACT 

Multivariate analyses were performed on replicated measurements from a collection of 56 
preserved Ningbingia dentiens Solem, 1985 (Gastropoda: Stylommatophora: Camaenidae), that 
ranged from full extension to full retraction. The positions of body landmarks during retraction/ 
extension vary complexly such that the only reliable indicator is the position of the foot tip relative 
to the remaining body wall. Shell size is no predictor of retraction/extension state. The nerve ring 
dilates, then compresses, as the buccal mass passes through it. From fully extended, to partially 
retracted, to fully retracted specimens, vagina length decreases 25%, then increases 10%; 
spermatheca length remains constant, then decreases 20%; and penis length decreases 15%, 
then increases 5%. 

Counting shell whorls (mean = 5.0) to the nearest 0.1 whorl was exactly as precise as 
measuring shell height (mean = 8.4 mm) to the nearest 0.1 mm; both were 1/7 as precise as 
measuring shell diameter (mean = 17.1 mm) to the nearest 0.1 mm. Distances among body 
landmarks within the shell had measurement errors 30 x to 340 x greater ihan for shell diam- 
eter. Measurement error was about 10% of total variance for vagina length, 20% for sperma- 
theca length, and 30% for penis length. The effects of measurement error and retraction/ 
extension equaled or surpassed individual variation for all three of these measurements. 

Key words: Gastropoda, Pulmonata, Camaenidae, Ningbingia dentiens, retraction, measure- 
ment error, anatomy, systematics. 



INTRODUCTION 

When a pulmonale land snail retracts into 
its shell, it invaginates the anterior part of the 
body, during which air is vented from the 
mantle cavity and blood redistributes among 
sinuses of the hemocoel (Jones, 1975). This 
process alters the relative positions, lengths, 
and shapes of the snail's organs. Since no 
fully reliable method has yet been found for 
always killing and fixing pulmonale land snails 
in an extended state (the success rate of the 
standard procedure of drowning in water or in 
a weak solution of chloryl hydrate or nicotine 
varies widely, depending on the taxonomic 
group and on field conditions), land-snail sys- 
tematlsts must often compare specimens dif- 
fering widely in their degree of body retraction/ 
extension. 

The effect of retraction/extension on body 
organs is poorly understood. Studies on body 
retraction in pulmonales (reviewed by Jones, 
1975) have so far been physiological in ap- 
proach, with little information on anatomical 
variation relevant to syslematisls. The most 
relevant study to dale (Dale, 1974) noted only 



that "the genital and digestive organs In the 
retracted snail are located [between] the man- 
tle cavity floor [and the retracted head-foot]," 
resulting in an unspecified degree of distor- 
tion of these organs. 

The role of measurement error in quantita- 
tively assessing both shell and soft-parts has 
never been examined in great detail. For ex- 
ample, various gastropod systematists mea- 
sure whorl-count to the nearest 1/4-whorl, 
1/8-whorl, 1/10-whorl, and 1/16-whorl, rarely 
with a published demonstration that that is the 
limit of achievable accuracy. Soft-part mea- 
surements, primarily of the lower reproductive 
tracts, are frequently presented in the taxo- 
nomic literature, sometimes with caveats 
about measurement error, but seldom with 
explicit assessments of its effect. 

Furthermore, when an investigator must 
choose which specimens to dissect from a 
preserved lot in which all are retracted within 
the shell, it would help to know whether the 
depth of retraction can be predicted from ei- 
ther shell size or the positions of body land- 
marks visible through the shell. 

The purposes of this paper are, for a single 



157 



158 



EMBERTON 



collection from a panmictic population of a 
pulmonate land snail, to (1) investigate the 
relative positions of body landmarks through- 
out a range of retraction/extension states; (2) 
test whether shell size is any predictor of re- 
traction/extension state; (3) qualitatively and 
quantitatively assess the effects of retraction/ 
extension on organ systems of systematic 
value; (4) compare the precisions of shell and 
soft-parts measurements; and (5) determine 
the relative contributions of retraction/exten- 
sion, individual variation, and measurement 
error to the total variance in the lengths of the 
vagina, spermatheca, and penis. 



MATERIALS AND METHODS 

This study made use of a single collection of 
Ninbingia dentiens Solem, 1985, a camaenid 
snail endemic to the northern Ningbing 
Ranges, north of Kununurra, the Kimberley, 
Western Australia. Sixty live, paratopotypic 
adults (Western Australian Museum 14.84 
and Field Museum of Natural History 205270) 
had been collected from a small pocket of ta- 
lus on a limestone dome that was shaded by 
a 2.5-meter boulder by A. Solem, L. Price, 
and B. Duckworth on 15 June 1980. The total 
area of the colony was much less than one 
cubic meter, and other colonies were sepa- 
rated by at least 50 meters of barren rock. 
Because of these conditions, the specimens 
almost surely belonged to a single panmictic 
population. 

The collection was made about two months 
into the dry season. All specimens therefore 
were in a state of estivation when collected. 
All were at least third-year adults, as evi- 
denced by their full-sized albumen glands 
(see Solem & Christensen, 1984). 

After drowning overnight in two or three wa- 
ter-filled, small jars, to each of which a few 
crystals of chloryl hydrate had been added, 
the specimens were fixed the next morning in 
95% ethanol. On reaching Chicago two 
months later, they were transfered to 70% 
ethanol. The preserved specimens exhibited 
a complete range of retraction states, from 
fully extended to tightly retracted. They af- 
forded the best opportunity I have yet encoun- 
tered to investigate variation in body retraction/ 
extension as it occurs in typical alcohol- 
preserved museum specimens. 

From the 56 intact specimens — three had 
previously been dissected (Solem, 1985), and 
the shell of one was accidently broken — I 



measured shell height and diameter to the 
nearest tenth of a millimeter, and whorl count 
to the nearest twentieth of a whorl. Also to the 
nearest twentieth of a whorl, I measured the 
location, as seen from the ventral side, of the 
shell's basal lip, the mantle color, the tip of 
the foot, the auricular-ventricular junction (a-v 
junction) of the heart, and the base of the om- 
matophore (upper, eye-bearing antenna). 
The heart's position was occasionally difficult 
to locate through the shell; changing the an- 
gle of illumination helped to pinpoint it. The 
mantle collar and the tip of the foot of re- 
tracted snails were clearly visible through the 
shell. The base of the ommatophore of re- 
tracted snails, however, was impossible to ac- 
curately locate, so instead I measured the 
most apical point of the invaginated left om- 
matophore, which was visible as a black tube 
through the ventral shell. From the dorsal sur- 
face of the shell I also measured the position 
of the apex of the liver (posterior digestive 
gland), to the nearest tenth of a whorl. This 
was sometimes obscured by opacity of the 
shell apex or by the presence of denatured 
fluid in the empty apical whorls, but could be 
detected by moving the narrow-beam illumi- 
nator, particularly by reflecting the light off the 
table surface into the umbilicus of the shell. 
All whorl-increment measurements were 
taken under magnification over a subdivided 
circle; shell height and diameter were taken 
manually with dial calipers. 

From these nine soft-part measurements, I 
calculated the following distances, all ex- 
pressed to the nearest twentieth of a whorl: (1 ) 
MANTLP, from the mantle collar to the shell's 
basal lip; (2) FOOTIP, from the mantle collar to 
the tip of the foot; (3) ANTENN, from the base 
of the everted left ommatophore or the ante- 
riad extent of the inverted left ommatophore to 
the heart a-v junction; (4) PALCAV, from the 
mantle collar to the heart a-v junction (this was 
used as an index of the length of the palliai 
cavity, the actual apex of which was not reli- 
ably discernable) ; (5) VICM AS, an index of the 
length of the visceral mass, from the heart a-v 
junction to the apex of the liver; and (6) 
EMPAPX, the empty apex of the shell, from 
the zero-whorl apical notch to the apex of the 
liver. 

To determine their precision, I took all mea- 
surements and performed all calculations 
three separate times. For each of the three 
shell measurements and the six calculated 
soft-part variables, I calculated two different 
indices of precision. The first was a mean со- 



LAND-SNAIL RETRACTION/EXTENSION 



159 



TABLE 1 . Loadings of principal components extracted from soft-part measurements associated with retraction/ 
extension. See text and Fig. 1 for explanation of variables. 



Variable 


PCI 


PC2 


PC3 


PC4 


MANTLIP 


-0.38 


-0.09 


0.47 


0.78 


FOOTIP 


0.30 


0.65 


0.30 


0.01 


ANTENN 


0.46 


0.46 


-0.05 


0.33 


PALCAV 


0.31 


-0.20 


-0.67 


0.51 


VICMAS 


-0.47 


0.43 


-0.32 


-0.06 


EMPAPX 


0.49 


0.36 


0.35 


-0.09 


% Variance 


44% 


24% 


19% 


9% 



efficient of variation. For each specimen, I cal- 
culated thie coefficient of variation (standard 
deviation divided by the mean) of its three 
replicate measurements, and averaged these 
coefficients over all specimens. The second 
index was the percentage of specimens with 
a coefficient of variation of zero, i.e. with iden- 
tical replicate measurements. For the shell 
variables only, I performed a third analysis of 
precision: making pairwise comparisons 
among the three sets of measurements, I cat- 
egorized all deviations as to size, then aver- 
aged the percentage of specimens falling 
within each category of deviation. 

Principal components analysis was used to 
explore the interrelationships among the soft- 
part distances during retraction/extension. To 
determine the influence of shell size and 
shape on this process, I computed the canon- 
ical correlations between the set of shell mea- 
surements and the set of soft-part distances. 
For these analyses, I used only the third set of 
measurements rather than the average of all 
three sets because of my belief that the ac- 
curacy of soft-part measurements increased 
with practice. 

After measuring the shells and body land- 
marks, I cracked and removed the shells, dis- 
sected off the mantle collar and diaphragm 
( = floor of the mantle cavity) from each spec- 
imen, sewed on an identification tag, then 
ranked the specimens from 1 to 57 according 
to how much of the head-foot was showing 
beneath the folded body wall — from all of the 
head and foot to just the tip of the foot. I called 
this variable the retraction/extension rank 
(RETRAN). I calculated non-parametric cor- 
relation ceoffficients between RETRAN 
(which was non-normally distributed) and 
each of the first four principal components of 
soft-part distances. 

I took five snails from each of four stages of 
the head-foot retraction series: "stage a," 



complete extension, ranks 1-5; "stage b," 
head invagination, ranks 20-24; "stage c," 
half-way foot retraction, ranks 35-39; and 
"stage d," full foot retraction, ranks 53-57. 
From each of these 20 snails I took the fol- 
lowing genital measurements: (1) vaginal 
length, from its beginning where the free ovi- 
duct and the spermatheca unite in a V-junc- 
tion to where it joins with the penis in a V- 
junction (the penioviducal angle) to form the 
atrium; (2) spermathecal length, from the free 
oviduct-spermathecal junction to its tip; (3) 
penis-plus-sheath length, from the peniovidu- 
cal angle to the insertion of the penial retrac- 
tor muscle; and (4) penis-minus-sheath 
length, from the penioviducal angle to the 
basal attachment of the sheath as best 
judged without dissection. After taking these 
measurements from the 20 snails and return- 
ing them to their vials, I repeated this mea- 
surement process twice. For each of the four 
genital measurements, I used two-way 
nested analysis of variance (ANOVA) to test 
for significant differences among retraction/ 
extension stages and among snails within a 
stage; and to partition the variance among re- 
traction/extension stage, individual variation, 
and measurement error. 

I next slit and pinned open the uneverted 
penial tube and sheath of each of the 20 
snails, and took five measurements of penial 
sculpture. These were: (1) left pilaster mid- 
width, (2) right pilaster mid-width, (3) number 
of wall ridges at the apical penis, (4) number 
of wall ridges at the mid-penis, and (5) central 
wall ridge mid-width. Means and standard de- 
viations for each extension/retraction stage 
were calculated. 

From each extension/retraction stage (a- 
d), I chose one well-dissected representative 
specimen (those with RETRAN ranks 1 , 24, 
38, and 54) for detailed dissection and illus- 
tration. 



160 



EMBERTON 



RESULTS 

Four principal components accounted for 
96% of the variation in distances among body 
landmarks during retraction/extension. The 
structures of these components are listed in 
Table 1 and are presented diagrammatically 
In Fig. 1, in which the relative contribution of 
each distance to each principal component is 
expressed as the width of its line, and the 
direction of its contribution is indicated by one 
or more arrowheads. The first two principal 
components (PC 1 and PC 2), accounting for 
44% and 24% of the total variance, involve 
simultaneous eversión of the head and foot, 
but differ in the other changes that take place 
during this process. In PC 1 the viceral mass 
shortens, thereby emptying the shell apex, 
whereas the palliai cavity lengthens as the 
mantle collar slides toward the shell lip. In PC 
2, on the other hand, the viceral mass length- 
ens, thereby partially filling the empty shell 
apex and slightly shortening the palliai cavity, 
with no shift in the mantle collar. The third 
principal component (PC 3) accounts for 19% 
of the total variance and primarily concerns a 
shortening of the palliai cavity as the mantle 
collar slides inward from the shell lip, and is 
weakly associated with shortening of the vis- 
ceral mass, emptying of the shell apex, and 
protrusion of the foot. PC 4, which accounts 
for 9% of the total variance, also involves re- 
traction of the mantle collar, but this time as- 
sociated with a lengthening of the palliai cav- 
ity and a slight evagination of the head. 

PC 1 and PC 2 were each significantly cor- 
related with the retraction/extension rank ( = 
RETRAN) (Spearman coefficients -.61 and 
-.69, p = 0.0001 for both). PC 3 was very 
weakly correlated (-.28, p = 0.03), and PC 4 
was uncorrelated (-.07, p = 0.63). 

Canonical correlation analysis between the 
set of shell variables and the set of soft-part 
distances yielded a single, highly significant 
(Wilks' Lambda = 0, p = 0) canonical varí- 
ate, with a canonical correlation of 1 .000. It is 
clear from the structure of this canonical varí- 
ate (Table 2) that it is an artifact due to the fact 
that EMPAPX + VICMASS + PALCAV + 
MANTLP = WHORLS. 

The true relationship between shell size 
and degree of retraction/extension is best 
shown by the squared multiple correlations 
between each soft-part distance and the ca- 
nonical variate (Table 2): they are negligible, 
ranging from 0.00 to 0.07, except for EM- 
PAPX (0.14). Thus, except for a very slight 



correlation between whorl count and the 
amount of empty apex, there is no real effect 
of shell size on the degree of retraction/ 
extension. 

External aspects of the four chosen stages 
of retraction/extension rank are shown in Fig. 
2. Stage a (Fig. 2a) is fully extended, b (Fig. 
2b) has just the head invaginated, с (Fig. 2c) 
has about half of the foot retracted within the 
body wall, and d (Fig. 2d) has most of the foot 
retracted within the body wall. 

Fig. 3 shows changes in the retractor mus- 
cle system, the head-foot, and the anterior 
digestive system during the four stages of re- 
traction/extension shown for the same four 
snails of Fig 2, in the same positions but in 
semi-diagrammatic mid-sagittal section. Fig. 
4 shows changes in the nerve ring ( = circum- 
esophageal ganglion = brain) during om- 
matophoral and head retraction. Figures 5-7 
show changes in the reproductive system 
during the four stages of head-foot retraction, 
illustrated by the same specimens shown in 
Figs. 2 and 3. 

The sequence of muscular contractions 
during retraction (Fig. 3) is the same as that 
observed in Helix pomatia (Trappman, 1916; 
Jones, 1975): rhinophoral, ommatophoral, 
buccal, and then pedal. 

In stage a of retraction/extension (Fig. 3a), 
the rhinophores are invaginated. In stage b, 
the ommatophores are also invaginated 
within the body cavity, where they are 
pressed folded against the body wall by the 
head-foot. 

The head is nearly filled by the buccal 
mass, which contains the radular ribbon (with 
its generative sac), the odontophore, the jaw, 
and the highly complex musculature for ma- 
nipulating these structures through the mouth 
during a feeding stroke (Carriker, 1946; Run- 
ham, 1975). During retraction/extension, the 
positions of the buccal mass, the mouth, and 
the foot relative to each other remain fairly 
constant (Fig. 3). At full extension (Fig. 3a), 
the mouth (BO) is bounded by upper and 
lower lips (dark stippling) which lie above the 
anterior lobe of the foot (light stippling). The 
upper lip shields the sharp, chitonous jaw. As 
the head-foot is retracted (Fig. 3b-d), the up- 
per lip is overlapped by the lower lip, which in 
turn is overlapped by the anterior lobe of the 
foot. At extreme contraction (Fig. 3d), the 
mouth region is distorted: the anterior foot 
lobe and the lower lip, enclosed within the 
body wall, compress and stretch the upper lip. 

The buccal mass (B), because of its thick 



LAND-SNAIL RETRACTION/EXTENSION 



161 







FIG. 1 . Components of retraction/extension in preserved snails. The structures of four principal components 
(PC 1 to PC 4) are shown that explain 96% of the total variance in six soft-part measurements between 
homologous landmarks. The relative contribution of each measurement is indicated by the width of its line, 
and the direction of its contribution by the arrowhead(s). 



musculature, remains relatively undistorted 
throughout retraction/extension (Fig. 3). The 
esophagus (O) is straight in stages a and b, 
but in stage с it has a kink and by stage d it Is 
folded double, with the buccal mass in a pos- 
terior position. The stomach (IZ) is a simple 
sac in retraction/extension stages a and b, but 



the compression of stages с and d makes it 
appear to have two lobes. 

The cerebral nerve ring and its connectives 
are highly elastic. Fig. 4 shows in dorsal view 
how the nerve ring (N) is deformed during the 
earliest stages of retraction/extension of the 
buccal mass. During fullest extension (Fig. 



162 



EMBERTON 



TABLE 2. Canonical correlation between shell measurements and soft-part measurements associated 
with retraction/extension. Loadings of the variables are given on each of three canonical variâtes (CV), 
as well as the squared multiple correlation (R^) of each variable with all variables in the other set. 



Variable 


CV1 


CV2 


CV3 


R^^ 


R22 


R23 


Shell: Diameter 


0.00 


0.30 


-1.25 


0.09 


0.14 


0.17 


Height 


0.00 


0.90 


1.05 


0.21 


0.29 


0.29 


Whorls 


1.00 


-0.50 


-0.11 


1.00 


1.00 


1.00 


Body: MANTLP 


0.43 


0.56 


-0.39 


0.00 


0.03 


0.04 


FOOTIP 


0.00 


-1.35 


-0.76 


0.00 


0.03 


0.03 


ANTENN 


0.00 


1.29 


0.97 


0.05 


0.05 


0.05 


PALCAV 


0.57 


-0.54 


-0.25 


0.07 


0.07 


0.07 


VICMAS 


2.35 


-0.37 


0.69 


0.00 


0.01 


0.04 


EMPAPX 


2.54 


-0.14 


-0.23 


0.14 


0.14 


0.17 







10 mm 



FIG. 2. Four stages (a-d) in the retraction/extension process: shell and head-foot. F = foot, MC = mantle 
collar, SL = shell lip. 



LAND-SNAIL RETRACTION/EXTENSION 



163 







5 mm 



FIG. 3. Four stages (a-d) in the retraction/extension process: digestive and retractor-muscle systems and 
the head-foot. В = buccal mass, BO = mouth, BR = buccal retractor muscle, F = foot, FR = pedal 
retractor muscle, IZ = stomach, О = esophagus, RT = inverted (retracted) ommatophore, T = ommato- 
phore, TER = ommatophoral retractor muscle, TVR = rhinophoral retractor muscle. 



4a), the nerve ring encircles the esophagus 
(O) and the two anterior ducts of the salivary 
gland (OG), just posterior to the buccal mass. 
As the buccal mass begins to retract, how- 
ever, it is pulled all the way back through the 
nerve ring (Figs. 4b and c). This process ini- 
tially stretches the nerve ring in all dimensions 
(Fig. 4b), then compresses it longitudinally 
(Fig. 4c), producing substantial changes in 
both size and shape of the dorsal cerebral 
ganglion (N, dark stippling). 



When retracted, the terminal portion of the 
reproductive system — from the genital pore to 
halfway up the prostate-uterus — is com- 
pressed between the retracted head-foot and 
the floor of the mantle cavity (Fig. 5). In the 
fully extended position (Fig. 5a), the genital 
pore (Y) opens on the right side of the head 
lateral to the anterior region of the buccal 
mass (B, stippled); the penis (P) is slightly 
bent and its retractor muscle (PR) is long and 
stretched; the vagina (V), spermatheca (S), 



164 



EMBERTON 

I 5 mm 






FIG. 4. Changes in the nerve ring (stippled) during early retraction/late extension. В = buccal mass, N 
nerve ring, О = esophagus, OG = salivary glands, RH = retracted head. 



and prostate-uterus (UT) are straight. (In 
Figs. 5-7, I have not differentiated between 
the tightly bound prostate and uterus, but 
have labeled them UT instead of the proper 
DG-UT.) 

When the head and anterior foot are re- 
tracted (Fig. 5b), the genital pore Is pulled 
back within the mantle collar (MC), adjacent 
to the retracted head-foot (RH-F, stippled); 
the buccal mass is pulled back to the level of 
the spermatheca; the penis is loosely con- 
torted; the vagina is bowed and folded at its 
junction with the spermatheca (not visible); 
the prostate-uterus is doubled back on itself 
as it folds against the retracting buccal 
mass. 

When the foot is half retracted (Fig. 5c), the 
genital pore retains its position near the apex 
of the retracted head-foot, the further retrac- 
tion of which pushes the pore back toward the 
penis; the penis therefore is tightly contorted 
and its retractor muscle is contracted; the va- 
gina is folded in half, with the apex of the fold 
held in place by connective tissue (not shown) 
to the anterior body wall; the prostate-uterus 
retains its tight folding adjacent to the buccal 
mass, which also apparently compresses the 
spermatheca (not labeled). 

When the foot is nearly completely re- 
tracted (Fig. 5d), it has slid further past the 
genital pore, pushing the penis inward so that 
its retractor muscle is stretched; the penis re- 



tains its tight contortion; the mid-way fold of 
the vagina remains attached by connective 
tissue to the anterior body wall, so that its 
length between this fold and the genital pore 
is stretched backward by the retracting head- 
foot; the prostate-uterus develops additional 
folds in the region of the further retracted buc- 
cal mass. 

Figures 6 and 7 show the removed repro- 
ductive system and the terminal genitalia at 
stages a through d of retraction/extension. 
Table 3 summarizes the means and standard 
deviations of the lengths of the vagina, sper- 
matheca, penis-plus-sheath, and penis-mi- 
nus-sheath calculated from five measured 
snails, each snail the average of three re- 
peated measurements. Vaginal length at full 
extension (stage a) averaged 4.5 mm, but 
when the head was invaginated (stage b), it 
only averaged 3.3 mm, a decrease of more 
than 25%. Later stages (c and d) of retraction 
actually increased the length of the vagina 
slightly to 3.6 mm (a 10% increase); as men- 
tioned previously, this is due to stretching the 
vagina between its distal ligamental attach- 
ment to the body wall and its proximal attach- 
ment to the retracting genital pore. These in- 
terstadial differences in vaginal length were 
significant (two-way nested ANOVA, F = 
7.18; degrees of freedom = 3, 16; p < 
0.005), despite the high and significant varia- 
tion among replicates at each retraction stage 



LAND-SNAIL RETRACTION/EXTENSION 



165 




a 






5 mm 



FIG. 5. Four stages in the retraction/extension process: reproductive system in relation to the buccal mass 
and retracted foot (both stippled). В = buccal mass, F = foot, MC = mantle collar, P = penis, PR = penial 
retractor muscle, RH-F = retracted head-foot, S = spermatheca, UT = prostate-uterus, V = vagina, Y = 
genital pore. In Fig. 5a, the genital pore opens toward the observer, whereas in 5b-d it opens away from the 
observer, into the space between the retracted foot and invaginated head skin, as indicated by the dashed 
circle around the pore. 



(F = 16.36; degrees of freedom = 16, 40; p 
«0.001). 

The spermatheca ( = bursa copulathx) was 
bound to the free oviduct (UV, Figs. 6 and 7) 
by connective tissue. Changes in the free ovi- 
duct were not quantified, but probably covar- 
ied with those of the spermatheca, the length 



of which was significantly affected by head- 
foot retraction (F = 9.55, p < 0.001), but in a 
different way than vaginal length (Table 3). 
The spermatheca averaged 2.7 mm regard- 
less of whether the snail was fully extended 
(stage a) or had invaginated its head (stage 
b). By the time it retracted half of its foot 




EMBERTON 





UT s 




5 mm 



FIG. 6. Four stages in the retraction/extension process: reproductive system dissected out from the body 
cavity. GD = hermaphroditic duct, GG = albumen gland, PR = penial retractor muscle, PS = penial 
sheath, S = spermatheca, UT = prostate-uterus, V = vagina, Y = genital pore, Z = posterior digestive 
gland plus gonad. 



(stage c), however, its spermatheca was 
shortened to 2.2 mnn (an 18% reduction). Full 
foot retraction (stage d) shortened it even fur- 
ther to 2.0 mm. Within each of these retrac- 
tion stages, however, there was highly signif- 
icant variation among replicate snails (F = 
5.14, p << 0.001), as evidenced by the high 
standard deviations in Table 3. 

The length of the penis and its sheath 
showed a retraction-stage pattern similar to 
that of the vagina (Table 3). Head invagina- 
tion (stage b) reduced its average length from 
6.4 to 5.5 mm (a 14% reduction), which was 
maintained (stage c) until extreme foot retrac- 
tion (stage d) stretched it slightly to 5.7 mm (a 
5% increase). Extreme variation among rep- 
licates (F = 6.77, p << 0.001) kept this re- 
traction effect just short of statistical signifi- 
cance (F = 2.62, 0.05 < p < 0.10). 

The length of the penis without the sheath 
varied insignificantly (F = 1.92, 0.10 < p < 
.25), but in the same manner as penis-plus- 



sheath length (Table 3). There was extreme 
variation among replicates (F = 4.97, p << 
0.001). 

The degree of retraction/extension had no 
effect on penial sculpture or on my ability to 
view it by dissection. Fig. 6 shows the com- 
plete reproductive systems, dissected free 
from the body cavity, of stages a-d. Despite 
increasing contortion of the penis and its 
sheath (PS) from stages a through d, the pe- 
nis could always be stretched out with pins in 
a dissecting dish, slit longitudinally, and 
pinned open to reveal its functional surface 
(Fig. 7). In the dissections shown in Fig. 7, the 
vas deferens (VD) has been cut and the pe- 
nial sheath has been cut near its apex in order 
to stretch out the upper part of the penis, 
which normally lies tightly coiled within the 
sheath (Solem, 1985, fig. 243). 

The sculpture of the lower part of the penis 
consists of two smooth, longitudinal pilasters 
(PP), the left of which — i.e. appearing on the 



LAND-SNAIL RETRACTION/EXTENSION 



167 




5 mm 




Щ PR 




FIG. 7. Four stages in the retraction/extension process: terminal genital tracts, pinned out and dissected 
open. P = penis, PP = penial pilasters, PR = penial retractor muscle, PS = penial sheath, S = sper- 
matheca, UT = prostate-uterus, UV = free oviduct, V = vagina, VD = vas deferens, Y = genital pore 



left side of the dissection, which is actually on 
the right side of the everted penis— is 
narrower than and stands about twice as high 
as the right. The ventral space between 



these two pilasters is smooth, but the dorsal 
penial wall is sculpted with longitudinal 
ridges, which are thin and numerous at the 
penial apex, but which anastomose in various 



168 



EMBERTON 



TABLE 3. Lengths of the terminal genital tracts at four stages of retraction/extension. Means and 
standard deviations are based on five snails per stage. 











Stage of Retraction/Extension 




Measurement 


a 




b 




С 


d 


Vaginal Length 




4.5 




3.3 




3.6 


3.6 






(0.7) 




(0.4) 




(0.3) 


(0.2) 


Spermathecal Length 


2.7 




2.7 




2.2 


2.0 






(0.2) 




(0.4) 




(0.2) 


(0.3) 


Penis + Sheath Length 


6.4 




5.5 




5.4 


5.7 






(0.4) 




(0.9) 




(0.5) 


(0.6) 


Penis - Sheath Length 


2.1 




1.7 




1.8 


1.9 






(0.3) 




(0.2) 




(0.3) 


(0.2) 


TABLE 4. Penial 


sculpture as measi 


ired at four stages 


of retraction/extension. 


















Dorsal Wall 


Ridges 


Retraction/ 


Left 
Pilaster 




Right 
Pilaster 




Number 




Width of 


Extension 






Central 


Stage 


Width (mm) 


Width (mm) 




Apical 


Mid 


Mid (mm) 


a 


0.18 




0.20 




13.6 


6.4 


0.09 




(0.05) 




(0.06) 




(2.9) 


(1.3) 


(0.03) 


b 


0.17 




0.20 




12.4 


5.4 


0.08 




(0.03) 




(0.04) 




(1.1) 


(0.5) 


(0.03) 


с 


0.18 




0.23 




12.3 


5.8 


0.07 




(0.05) 




(0.03) 




(2.2) 


(2.2) 


(0.02) 


d 


0.17 




0.22 




12.6 


6.0 


0.08 




(0.03) 




(0.04) 




(3.2) 


(0.7) 


(0.03) 



patterns proximally to become thicker and 
less numerous. 

Measurements and counts of these fea- 
tures of penial sculpture (Table 4) show that 
they are extremely variable within the popu- 
lation and that there is no effect whatever due 
to the stage of retraction/extension. Averaged 
over all 20 specimens, the left pilaster width 
was 0.17 mm, with a coefficient of variation 
(CV) of 0.2; right pilaster width = 0.21 mm 
(CV = 0.2); number of apical dorsal ridges = 
12.7 (CV = 0.2); number of mid-dorsal ridges 
= 5.9 (CV = 0.2); and central mid-dorsal 
ridge width = 0.08 mm (CV = 0.4). 

Table 5 gives values of the two indices of 
measurement precision, along with the 
means, standard deviations, and ranges of 
the variables as calculated from the third set 
of replicated measurements. The mean coef- 
ficient of variation (CV Mean) was lowest for 
shell diameter (0.001 ), and was 7 x this value 
for shell height and whorl-count, and 31 x to 
344 X this value for the soft-part distances 
between homologous landmarks. 

The second index, the percentage of spec- 
imens with zero CVs, indicates the percent- 
age of specimens for which the three repli- 



cated measurements were identical. Its value 
varied from 9% to 70% and the ranges of val- 
ues were equable between shell and soft-part 
variables (Table 5). 

The precision of the shell measurements is 
analyzed in more detail in Table 6. Re-mea- 
surement of a shell's diameter gave precisely 
the same result 76% of the time, differed by 
0.1 mm in 23% of the cases, and differed by 
0.2 mm in only 1% of the cases. The greater 
imprecision of height measurements is evi- 
denced by both its percentage (55%) and its 
range (up to 0.3 mm) of deviations. Whorl 
counts were 70% repeatable, and of the de- 
viations, 93% (0.28/0.30) were off by only 
one-tenth whorl; no deviation exceeded two- 
tenths whorl. 

The results of variance-partitioning of the 
lengths of the terminal genitalia are presented 
in Table 7. In this table, the effects of retraction/ 
extension (four stages), individual variation (5 
replicates per stage), and measurement error 
(three measurements per replicate) are sum- 
marized from analysis of variance for each of 
four genitalic measurements. For vagina 
length, over half of the total variation was due 
to retraction/extension, less than half was due 



LAND-SNAIL RETRACTION/EXTENSION 



169 



TABLE 5. Shell and soft-part rneasurements: univariate statistics and two indices of measurement 
precision. 





Mean (SD) 


Range 


Precision Indices 




Variable 


CV Mean (SD) 


CV = 


Diameter 


17.1 (0.5) 


15.9-18.3 


0.001 (0.002) 


64% 


Height 


8.4 (0.3) 


7.7-9.2 


0.007 (0.005) 


21% 


Whorls 


5.0(0.1) 


4.85-5.25 


0.007 (0.006) 


32% 


MANTLP 


0.05 (0.06) 


0.00-0.20 


0.344 (0.660) 


70% 


FOOTIP 


0.25(0.15) 


-0.05-0.50 


0.121 (0.139) 


40% 


ANTENN 


0.43(0.16) 


0.15-0.80 


0.133(0.152) 


13% 


PALCAV 


0.54 (0.07) 


0.35-0.65 


0.067 (0.069) 


31% 


VICMAS 


2.72 (0.30) 


2.15-3.35 


0.031 (0.046) 


19% 


EMPAPX 


1.55(0.32) 


0.9-2.3 


0.087(0.128) 


9% 



TABLE 6. Deviations of repeated shell measurements from their grand means. Deviation units are 0.1 
mm for diameter and height, and 0.1 whorls for whorl-count. The proportion of replicates (and its 
standard eviation) is given for each deviation unit. 







Deviation 




Measurement 


0.0 


+ 0.1 


+ 0.2 or 0.3 


Diameter 

Height 

Whorls 


0.76(0.10) 
0.45 (0.03) 
0.70 (0.04) 


0.23 (0.09) 
0.45 (0.05) 
0.28 (0.05) 


0.01 (0.01) 
0.10(0.03) 
0.02 (0.02) 



to individual variation (among the five snails 
measured for each stage of retraction), and 
less than a tenth was due to measurement 
error. For spermatheca length, over half of 
total variation was due to retraction/ex- 
tension, about one-fourth was due to individ- 
ual variation, and about one-fifth was due to 
measurement error. For penis-plus-sheath 
length, individual variation accounted for over 
half the total variation; retraction/extension 
accounted for one fifth, whereas measure- 
ment error accounted for over one fourth, of 
this variation. For penis-minus-sheath length, 
measurement error was high, accounting for 
over one third of the total variation; individual 
variation accounted for half, and retraction/ 
extension accounted for only about one tenth 
of the total variation in the length of the penis 
minus the sheath. 



DISCUSSION AND CONCLUSIONS 

Body Landmarks During Retraction/ 
Extension 

Interpreting the principal components 
(PCs) of retraction/extension (Table 1, Fig. 1) 
must allow for the facts that (1) since the six 
variables used in the analysis differ in their 



measurement precisions, their differences 
may show up as artifacts in the structure of 
one or more PCs; and (2) the variable AN- 
TENN should, in retrospect, have been mea- 
sured from the mantle collar rather than from 
the heart, which causes it to overlap PALCAV 
and therefore makes its interpretation more 
difficult in the context of PC structure. 

Retraction/extension of the head and the 
foot (ANTENN and FOOTIP) are strongly cor- 
related (PCs 1 and 2) because of their phys- 
ical connection, but apparently either end 
may slightly precede the other (PCs 3 and 4). 
During head-foot extension/retraction, the vis- 
ceral mass is either farther or closer to the 
shell apex (PCs 1 and 2), and the mantle col- 
lar is either closer or farther (PCs 1 , 3 and 4) 
from the shell lip (Fig. 1, Table 1). Thus re- 
traction/extension involves a complex inter- 
play of body landmarks. This complexity may 
be due, at least in part, to confounding the two 
separate processes of retraction and exten- 
sion. Thus, for example, the mantle collar 
may be pulled along with the head-foot during 
extension, but may lag behind the head-foot 
during retraction. For this reason, the best 
single criterion for the state of retraction/ 
extension in preserved specimens is the po- 
sition of the tip of the foot relative primarily to 
the body wall, and secondarily to the shell lip. 



170 



EMBERTON 



TABLE 7. Partitioning of the total variance in the lengths of terminal genital tracts. 







Source of Variation 




Measurement 


Retraction/ 
Extension 


Individual 
Variation 


Measurement 
Error 


Vaginal Length 
Spermathecal Length 
Penis + Sheath Length 
Penis - Sheath Length 


52% 
55% 
20% 
12% 


40% 
26% 
53% 
50% 


8% 
19% 
27% 
38% 



Thus the mantle collar, which is often the 
most visible landmark through the shell of re- 
tracted snails, is unreliable as an Indicator 
(compare Fig. 2c and d). 

When I ranked the deshelled specimens 
according to the position of the foot-tip rela- 
tive to the body wall (RETRAN), this Index 
was significantly but not strongly correlated 
with PCs 1 , 2, and 3. Thus, only RETRAN Is a 
reliable general measure of retraction/ 
extension state as It distorts systematically 
important body organs. 

Effect of Shell Size 

Shell size Is no predictor of the state of re- 
traction/extension. This conclusion does not 
come from canonical-correlation analysis be- 
tween the sets of shell variables and soft-part 
variables (Table 2), because the latter were 
measured as whorl Increments and the sum 
of four of them equals the shell variable 
"Whorls" (thus the first, and only significant, 
canonical varíate merely reflects this mathe- 
matical relationship). The conclusion results 
rather from the facts that the second and third 
canonical variâtes were not only non-signlfl- 
cant but also biologically nonsensical, and 
that the multiple correlation of each soft-part 
variable with the set of shell variables was 
zero to negligible (Table 2). 

Retraction/Extension and Systematic 
Characters 

The four stages I chose from the continuum 
of retraction/extension are fairly distinct and 
easily Identified In deshelled specimens: full 
extension, head invagination, half retraction 
of the foot, and full retraction of the foot (Fig. 
2a-d). The latter stages are difficult or Impos- 
sible to Identify without removing the shell, 
however, because the transition between the 
exposed foot and the folded body wall Into 
which It retracts Is seldom discernable 
through the shell and — sometimes — the man- 



tle (see Fig. 2c and d). Since the specimens 
did not fall into discrete categories, the four 
stages were chosen at equal distances (I.e. 
numbers of specimens) along the retraction/ 
extension continuum. For this reason, the five 
"replicates" of each retraction/extension 
"stage" are actually only five adjacent speci- 
mens In one region of the continuum; the sim- 
ilarity of replicates was probably greatest at 
the endpolnts: full extension and full retrac- 
tion. 

Body retraction Is effected by the sequen- 
tial contractions of four muscles that are pos- 
teriorly fused — the rhinophoral, ommatopho- 
ral, buccal, and pedal retractors. The 
sequence Is the same In Helix pomatia (Trapp- 
man, 1916; Jones, 1 975). It appears that suc- 
cessive muscles are not effective throughout 
the process of body retraction, but each 
reaches a limit of contraction, after which It 
folds as the next muscle(s) continue to con- 
tract (Fig. 3). Body retraction is remarkably 
rapid. For example, several species of poly- 
gyrid land snails retract too quickly to be fixed 
In extended condition by Immersion In liquid 
nitrogen (Emberton, unpublished). 

Distortion of Internal organs by the retrac- 
tion process differs widely In both Its degree 
and its nature. The retractor muscles them- 
selves, despite drastic changes in length, 
shape, and both absolute and relative dis- 
tances among each other, retain their basic 
topology throughout retraction. The length of 
the foot Is not greatly altered during body re- 
traction, if measured along the curvature of Its 
sole (Fig. 3). The upper lip is extremely 
stretched during the final stage of retraction. 
The size and shape of the buccal mass is 
virtually unaffected, although this was not 
studied in detail. The esophagus undergoes 
extensive shape changes, but its overall 
length appears stable. Both size and shape of 
the stomach, however, are sensitive to the 
stage of body retraction (Fig. 3). 

One of the most drastic changes during re- 
traction occurs during Its early stages. This is 



LAND-SNAIL RETRACTION/EXTENSION 



171 



when the buccal mass Is pulled back through 
the nerve ring. The effects of the size and 
shape of the nerve ring and its constituent 
ganglia are major (Fig. 4). 

The distal reproductive system undergoes 
a great deal of distortion during body retrac- 
tion as it is displaced by both the buccal mass 
and the foot (Figs. 5 and 6). The lengths of the 
terminal tracts are often given in systematics 
accounts, but the recorded variation com- 
pounds individual variation, variation due to 
retraction state, and measurement error. 

The results of this study indicate clearly that 
body retraction contributes a significant 
amount of variation in the lengths of the ter- 
minalia, even when they are dissected free 
from the body and pinned out straight for 
measurement. In the case of the lengths of 
the vagina and the spermatheca, the effect of 
body retraction actually outweighed individual 
variation (Table 7). Thus, these tracts are not 
only bent and folded by retraction, but are 
also physically shortened. The effect of body 
retraction is nonlinear (Table 3), and the na- 
ture of the nonlinearity is likely to differ among 
species, so excluding the effect of body re- 
traction on the vaginal and spermathecal 
lengths for interspecific comparisons would 
be difficult indeed. This irremedial effect is 
quite profound: one-fourth retraction can re- 
duce the total length of the vagina by 25% 
and the spermatheca by 18%. (This caveat 
probably does not apply, however, to the ma- 
jority of snails having a long, thin-walled sper- 
mathecal duct.) Therefore, interspecific differ- 
ences can easily be rendered undetectable 
by employing retracted specimens. Measure- 
ment precisions for the vagina and sperma- 
theca were good (approximately 1/10 and 1/5 
total variances [Table 6]), so interspecific 
comparisons using only fully extended spec- 
imens should be fairly reliable. 

The land-snail penis is frequently of great 
value in systematics because of its vahability. 
Fortunately, body retraction/extention has no 
substantial effect on penial characters. For 
the length of both the total and the basal 
penis, the effect of retraction/extention was 
well below measurement error (Table 7). 
Also, no matter how distorted the penes were 
in retracted states (Figs. 5 and 6), it was 
always possible to quantitatively compare the 
sculptures of their functional surfaces by dis- 
section (Fig. 7). The widths of the two pilas- 
ters and of the central ridge of the penial wall 
and the numbers of dorsal wall ridges, al- 
though variable among individuals, were un- 



affected by the state of body retraction/ 
extension (Table 4). 

Measurement Precision 

The differences in precision of the shell 
measurements depended in part on the prob- 
lems of orienting the shells for diameter and 
height measurements, and of locating the 
zero-whorl notch for whorl counts. Ningbingia 
dentiens has a relatively low-spired shell ap- 
proximately equal in size to the tip of one of 
the human fingers used to hold and orient it 
for measurement with dial calipers. These 
factors make it difficult to judge the position of 
the coiling axis. Measuring the diameter re- 
quires holding the coiling axis parallel to the 
jaws of the calipers and rotating the shell on 
this axis until the maximum diameter is 
achieved. This rotation — it becomes slight 
and almost unnecessary with practice — helps 
locate the coiling axis and ensures that the 
maximum diameter is measured. The preci- 
sion of this measurement, therefore, is rela- 
tively high: 95% of repeated measurements 
lay within 0.5% of their mean, and no single 
error was greater than 1 % of the grand mean. 

The measurement of shell height is sensi- 
tive to any tilt away from the coiling axis. Such 
tilt is especially difficult to detect and control in 
the plane perpendicular to the jaws of the cal- 
ipers. Rotation in this plane does not affect 
the measurement of the diameter. Because of 
this additional source of error the precision of 
height is lower than that of diameter: 95% of 
repeated measurements lay within 2.8% of 
their mean, and no single error exceeded 4% 
of the grand mean. 

Accurately locating the origin of the suture 
(Schindel, 1989) — the apical notch of zero 
whorls (see Emberton, 1985, fig. 1) — requires 
a perfectly clean apex and an incidental 
source of narrow-beam illumination. The re- 
sult was that 95% of repeated measurements 
lay within 2.8% of their mean, and no single 
error exceeded 4% of the grand mean. Thus, 
for this, on average, 8.4-by-17.1-mm, 5- 
whorled shell, whorl counts to the nearest 0.1 
whorl were just as precise as height measure- 
ments to the nearest 0.1 mm, which were 1/7 
as precise as diameter measurements to the 
nearest 0.1 mm. 

The applicability of this precision analysis 
to other gastropod studies is limited, depend- 
ing both on the size, shape, number of whorls, 
and ease of detection of the suture origin of 
the shell, and on the experience and care of 



172 



EMBERTON 



the investigator. On the other hand, its meth- 
odology should be broadly applicable to other 
studies. The indices (Table 5) and categori- 
zations (Table 6) of precision are useful, can 
be applied to any kind of measurement, and 
provide the kind of backup necessary to, but 
too often lacking in, morphometric studies 
(e.g. Gould & Woodruff, 1986). 

The precision of each soft-part distance de- 
pends on two main factors: the ease of locat- 
ing its landmarks, and its size. The mantle 
collar in this species is easy to see either 
flush with the shell lip or, when it is retracted 
up to 0.2 whorl, through the shell. Its distance 
(MANTLP) from the shell lip, another reliable 
landmark, is calculable with high precision. 
Because MANTLP is such a short distance 
however (mean = 0.05 whorl), the few devi- 
ations among replicated measurements are 
equal to or greater than its mean value. This 
combination of factors explains why, among 
all variables, MANTLP shows the highest pre- 
cision using the CV = index (70%), but the 
lowest precision (= highest value) using the 
CV-mean index (0.34) (Table 1 ). CV = is the 
better index because of its independence 
from the size of the variable; MANTLP is a 
variable calculable with high precision. 

The tip of the foot is as easy to locate as the 
mantle collar, but the distance between these 
two landmarks (FOOTIP) has a lower preci- 
sion (CV = value = 40%) than MANTLP for 
two reasons. First, the mantle collar is a 
slightly less reliable landmark than the shell 
lip. Second, and most importantly, the everted 
foot — unlike in retracted foot — rarely follows 
the contour of the body whorl as it is depicted 
diagrammatically in Fig. 1. Instead, it is more 
or less straight (Fig. 2a and b), so imprecision 
enters in estimating its curved distance. 
Nonetheless, FOOTIP is the second most 
precise soft-part distance. 

Slightly less precise (CV = value = 32%) 
is the distance between the mantle collar and 
the auhculo-ventricular junction of the heart, 
used as a truncated measurement of the pal- 
liai cavity (PALCAV). Although the mantle col- 
lar is a good landmark, the position of the 
heart, as mentioned in the Materials and 
Methods section, can be difficult to pinpoint, 
depending on its chromatic differentiation 
from surrounding tissues, its size (state of 
contraction? degree of distention by haemo- 
coelic pressure changes at death?), and the 
opacity of the overlying shell. 

Finding the retracted ommatophore was al- 
most always easy — it is very dark gray and 



shows well through the shell — but judging its 
anterior-most point involved error. Measuring 
the base of the ommatophore on an extended 
snail suffered from either guessing its position 
on an invaginated head, or estimating its 
curved distance (as for the everted foot), or 
both. Combining this imprecision with that of 
the heart made the distance between these 
two landmarks (ANTENN) fairly imprecise, 
with a CV = index value of 13%. In retro- 
spect, a better measurement of antennal re- 
traction (ANTENN) would have been the dis- 
tance between the ommatophore and the 
mantle collar (instead of the heart); this would 
have been more precise. 

There is apparently no way to improve the 
low precisions (CV = values = 19% and 
9%) of measurements of the visceral mass 
(VICMAS) and the apex of the shell empty of 
(unoccupied by) body tissue (EMPAPX). The 
landmark common to these two variables, the 
apex of the visceral mass, is often quite diffi- 
cult to pinpoint. As mentioned in the Materials 
and Methods section, it can be obscured both 
by denatured fluid within the empty apex and 
by opacity of the shell, the apex of which re- 
ceives the earliest and heaviest abrasion. 
Finding this landmark is aided by thoroughly 
cleaning the umbilical pit and by manipulating 
the light source. The imprecision index of 
VICMAS (the distance between the heart and 
the apex of the viscera) reflects — as do the 
indices of all soft-part distances, to a lesser 
degree — my learning process in locating the 
landmarks. The least precise of all soft-part 
distances, EMPAPX, combines the discrimi- 
natory problem of the visceral apex with that 
of its other landmark, the suture origin, which 
was discussed above. 

In sum, soft-part distances (Table 5, Fig. 1) 
varied greatly in precision depending on their 
size and on the difficulty of locating and mea- 
suring their landmarks. The "CV mean" index 
was an indicator of size-related imprecision, 
and the "CV = 0" index was an indicator of 
landmark-related precision. Both indices also 
include my improvement in discriminating dif- 
ficult landmarks. 

Relative Sources of Variation 

The results indicate clearly that body re- 
traction/extension contributes a significant 
amount of variation to the lengths of the ter- 
minal genital tracts, even when they are dis- 
sected free from the body and pinned oui 
straight for measurement. In the case of the 



LAND-SNAIL RETRACTION/EXTENSION 



173 



lengths of the vagina and the spermatheca 
(including duct), the effect of body retraction 
actually outweighed individual variation 
(Table 7). Thus, these tracts were not only 
bent and folded by retraction, but are also 
physically shortened. 

The penis is frequently of great value in 
systematics because of its variability, proba- 
bly due to sexual selection (e.g. Cain, 1982; 
Eberhard, 1985). Measuring the length of the 
penis of Ningbingia dentiens requires 
straightening the upper region that is tightly 
coiled within the penial sheath (Fig. 7). This 
doubtless contributed to the large error of 
measurement, which accounted for over one 
fourth of total variance in the total length of 
the penis (Table 7). Nevertheless, the mea- 
surement error for the basal penis, exclusive 
of the sheath, was even more imprecise, ac- 
counting for over one third of total variance in 
length. These are substantial components of 
variation that could easily interfere with any 
attempts at meaningful comparisons among 
populations or species. Fortunately, none of 
this measurement error affected the penial 
sculpture, as viewed by dissecting open the 
uneverted penial tube (Fig. 7), which can be a 
valuable source of systematic characters 
(e.g. Schileyko, 1978; Solem, 1985; Ember- 
ton, 1988). However, this study has shown 
quantitatively that intrapopulational variance 
in these characters can be quite high (Table 
4). 



ACKNOWLEDGMENTS 

This work was carried out with the support 
of the following grants from the National Sci- 
ence Foundation: BSR-81-19208 to Alan 
Solem for fieldwork, and BSR-83-12408 to 
Alan Solem and BSR-87-00198 to the author 
for analysis. I am grateful to Alan Solem for 
criticising an early draft of the manuscript. 



LITERATURE CITED 

CAIN, A. J., 1982, On homology and convergence. 
Pp. 1-19 in: JOYSEY, K. A. & A. E. FRIDAY, 
eds.. Problems of phylogenetic reconstruction. 



Systematics Association Special Volume No. 21, 
Academic Press, London & New York. 

CARRIKER, M. R,, 1946, Morphology of the ali- 
mentary system of the snail Lynmaea stagnalis 
appressa Say. Transactions of tfie Wisconsin 
Academy of Sciences, Arts and Letters, 38: 1- 
88. 

DALE, В., 1974, Extrusion, retraction and respira- 
tory movements in Helix pomatia in relation to 
distribution and circulation of the blood. Journal 
of Zoology, London, 173: 427-439. 

EBERHARD, W. G., 1985, Sexual selection and 
animal genitalia. Han/ard University Press, Cam- 
bridge. 

EMBERTON, K. C, 1985, Seasonal changes in the 
reproductive gross anatomy of the land snail Tri- 
odopsis tridentata tridentata (Pulmonta: Polygy- 
ridae). Malacologia, 26: 225-239. 

EMBERTON, K. C, 1988, The genitalic, allozymic, 
and conchological evolution of the eastern North 
American Triodopsinae (Gastropoda: Pulmo- 
nata: Polygyridae). Malacologia, 28: 159-273. 

GOULD, S. J. & D. S. WOODRUFF, 1986, Evolu- 
tion and systematics of Cerion (Mollusca: Pulmo- 
nata) on New Providence Island: a radical revi- 
sion. Bulletin of ttie American ¡\Auseum of Natural 
History, 182: 389-490. 

JONES, H. D., 1975, Locomotion. Pp. 1-32 in: 
FRETTER, V. & J. PEAKE, eds., Pulmonates, 
Volume 1, Functional anatomy and pfiyslology. 
Academic Press, London & New York. 

RUNHAM, N. W., 1975, Alimentary canal. Pp. 53- 
104 in: FRETTER, V. & J. PEAKE, eds., Pulmo- 
nates, Volume 1, Functional anatomy and pfiys- 
lology. Academic Press, London & New York. 

SCHILEYKO, A. A., 1978, On the systematics of 
Tricfiia s. lat. (Puimonata: Helicoidea: Hygromi- 
idae). t\/lalacologia, 17: 1-56. 

SCHINDEL, D. E., 1989, Architectural constraint on 
the coiled geometry of gastropod molluscs. In: 
ALLMON, W. & R. ROSS, eds., Biotic and abiotic 
factors in evolution. University of Chicago Press 
(in review). 

SOLEM, A., 1985, Camaenid land snails from 
Western and central Australia (Mollusca: Puimo- 
nata: Camaenidae) V. Remaining Kimberley gen- 
era and addenda to the Kimberley. Records of 
tfie Western Australian (Museum, Supplement 
No. 20, pp. 707-981. 

SOLEM, A. & С CHRISTENSEN, 1984, Camaenid 
land snail reproductive cycle and growth patterns 
in semi-arid areas of north-western Australia. 
Australian Journal of Zoology, 32: 471-491 . 

TRAPPMANN, W., 1916, Die Muskulatur von Helix 
pomatia L. Zeitscfirifft für Wissenscfiafñlicfien 
Zoologie, 115: 460-489. 

Revised Ms. accepted 5 July 1989 



MALACOLOGIA, 1989, 31(1): 175-195 

BIOLOGY AND COMPARATIVE ANATOMY OF DIVARISCINTILLA YOYO AND 

D. TROGLODYTES, TWO NEW SPECIES OF GALEOMMATIDAE (BIVALVIA) 

FROM STOMATOPOD BURROWS IN EASTERN FLORIDA 

Paula M. Mikkelsen^ & Rüdiger Bieler^ 



ABSTRACT 

Two new galeommatid bivalves, Divariscintilla yoyo and D. troglodytes, are described as 
commensals in burrows of the stomatopod Lysiosquilla scabhcauda from central eastern Flor- 
ida. They are remarkable in their snail-like appearance and behavior, due to elaborately orna- 
mented palliai layers enclosing the shell, and their ability to actively crawl on a highly mobile foot. 
Both are simultaneous hermaphrodites, brooding their larvae in the suprabranchial chamber 
prior to release of straight-hinged veligers. The two new species differ from one another in shell 
morphology, the number of secretory "flower-like organs," and the nature and ornamentation of 
the mantle. They differ from the type and only other described species in this genus, D. maoria, 
primarily in shell characters, namely in anterior (rather than posterior) prolongation, and in the 
absence of a ventral cleft. The genus Divariscintilla, previously known only from New Zealand, 
is redefined with the following diagnostic characters: incompletely internalized shell with anterior 
or posterior prolongation, species-specific numbers of palliai tentacles and papillae, a two-part 
foot used in active crawling and "hanging" utilizing both byssus- and byssus adhesive glands, 
secretory "flower-like organs" on the anterior surface of the visceral mass, eulamellibranch 
ctenidia with interlamellar and interfilamentary junctions, and simultaneous hermaphroditism 
with larval brooding. 

Keywords: Divariscintilla, Galeommatidae, Galeommatoidea, systematics, anatomy, Sto- 
matopoda, commensalism, Florida. 



INTRODUCTION 

A wide variety of mollusks are known to 
associate with other invertebrates in symbi- 
otic relationships. Galeommatoidean [= ga- 
leommatacean] bivalves are among the best 
known symbionts (Boss, 1965, as Erycina- 
cea), and are interesting in the anatomical 
and behavioral modifications associated with 
their specialized mode of life. These include 
(1) internalization of the shell by the middle 
palliai fold, (2) elaboration of this pallia! layer 
by tentacles and papillae, (3) snail-like loco- 
motion on a highly extensible foot, and (4) the 
occurrence of hermaphrodites or dwarf 
males. Anatomical data are available for spe- 
cies in less than 30 of the approximately 110 
Recent, presumably valid genera (Vokes, 
1980; Chavan, 1969). 

Within the family Galeommatidae Gray, 
1840, the monospecific genus Divariscintilla 
Powell, 1932, was originally based on empty 
shells of the New Zealand species D. maoria 
Powell, 1932. Distinguishing shell characters 



include a deep ventral notch, a strongly ob- 
lique posterior prolongation, and dentition lim- 
ited to a small conical tooth in each valve. The 
anatomy and biology of D. maoria were sub- 
sequently described by Judd (1971) from 
specimens found living in the burrows of the 
stomatopod crustacean Heterosquilla tricari- 
nata (Claus). 

A study of organisms associated with the 
sand-burrowing stomatopod Lysiosquilla sca- 
bricauda (Lamarck) in shallow waters in 
eastern Florida has yielded a number of 
undescribed or poorly known molluscan 
species. Data on the two species of vitrinellid 
gastropods in the burrows have appeared 
elsewhere (Bieler & Mikkelsen, 1988). Five 
previously undescribed species of galeom- 
matid bivalves were also encountered, and 
two, assignable to Divariscintilla, are here 
described. The data presented here identify 
anatomical characters of value at the generic 
level and represent a step toward clarification 
of the taxonomic disorder in this super- 
family. 



'Indian River Coastal Zone Museum, Harbor Branch Océanographie Institution, 5600 Old Dixie Highway, Ft. Pierce, Florida 
34946, and Dept. of Biological Sciences, Florida Institute of Technology, Melbourne, Florida 32901 U.S.A. 
^Delaware Museum of Natural History, P.O. Box 3937, Wilmington, Delaware 19807 U.S.A. 



175 



176 



MIKKELSEN & BIELER 



MATERIAL AND METHODS 

Stomatopod burrows in shallow-water sand 
flats in the Indian River lagoon just inside the 
Ft. Pierce Inlet, St. Lucie County, eastern 
Florida (27°28.3'N, 80°17.9'W), were sam- 
pled using a stainless steel bait pump ("yabby 
pump") and sieves of 1-2 mm mesh. Depths 
during extreme low water ranged from less 
than 0.5 m to supratidal, wherein the water 
level lay several centimeters below the level 
of the sand. 

Living clams were maintained in finger 
bowls of seawater at room temperature 
(24°C). Behavioral studies were aided by 
video recordings taken of the living animals in 
aquaria using a standard commercial 1/2- 
inch-format video camera equipped with a 
macro lens. 

Carmine and fluorescein sodium particles 
aided observation of ciliary action and cur- 
rents produced by the animals. Relaxation 
prior to dissection or preservation was most 
effectively accomplished with menthol crys- 
tals, floated on the seawater surface, or with 
crystalline magnesium sulfate, added directly 
in small, gradual amounts. Methylene-blue/ 
basic-fuchsin and neutral red were used to 
delineate tissues and organs in gross dissec- 
tions. 

For histological serial sections, animals 
were fixed in either a glutaraldehyde-formalin 
solution (4% formalin, 2.5% glutaraldehyde in 
0.1 M phosphate buffer, pH 7.2) or in 5% buff- 
ered formalin (Humason, 1962: 14). Shells 
were decalcified using either dilute hydrochlo- 
ric acid (complete decalcification within min- 
utes, however, with bubble production pre- 
senting technical histological difficulties) or a 
1% solution of ethylene diamene tetraacetic 
acid (EDTA, adjusted to pH 7.2; decalcifica- 
tion complete over a period of 5-6 days). 
Specimens were embedded in paraffin, sec- 
tioned at 5-7 fxm and stained with alcian blue/ 
periodic-acid-Schiff (PAS), counterstained 
with Harris' hematoxylin/eosin (Humason, 
1962: 125, 269, 298), hereafter referred to as 
APH. Staining reactions described in the text 
refer to this method unless othenwise noted. 
Colors referred to in the text are supplied for 
future use, e.g., to infer homologies of the var- 
ious glands. Other similarly prepared speci- 
mens were stained with hematoxylin/eosin. 
The section in Fig. 23 was fixed in Kar- 
novsky's fixative (Karnovsky, 1965), post- 
fixed in 1% osmium tetroxide in a phosphate 
buffer, dehydrated through an ethanol propy- 



lene oxide series, embedded in Epon-812, 
sectioned at 1 |лт and stained with Richard- 
son's stain (Richardson, et al., 1960). Photo- 
micrographs of sections were taken either 
with a Zeiss Photomicroscope-3 or an Olym- 
pus BH-2 stereomicroscope fitted with an 
Olympus OM-2 camera. 

For scanning electron microscopy (SEM), 
partially dissected preserved specimens were 
passed through an ethanol-to-acetone series 
and critical-point dried. These and air-dried 
shells were coated with gold/palladium and 
scanned using a Zeiss Novascan-30. 

All cited anatomical measurements were 
taken from specimens of average size (ap- 
proximately 10-15 mm mantle length). Be- 
cause of the extreme expansivity and con- 
tractility of the mantle, it is difficult to 
accurately measure "length" of a living animal 
of this type. Approximate mantle length was 
measured along an anteroposterior axis from 
an animal in normal crawling or hanging pos- 
ture; throughout the text, this is expressed as 
"relaxed" and does not refer to any chemical 
treatment of the animals. Measurements of 
type specimens refer to preserved mantle 
lengths. Shell length is expressed as the max- 
imum dimension, i.e. an oblique anteropos- 
terior length. 

Cited institutions are (* indicates location of 
type and other voucher material): 

AMNH — American Museum of Natural His- 
tory, New York 

*DMNH — Delaware Museum of Natural His- 
tory, Wilmington 

HBOI — Harbor Branch Océanographie In- 
stitution, Ft. Pierce, Florida 

*IRCZM — Indian River Coastal Zone Mu- 
seum, HBOI 

*MCZ — Museum of Comparative Zoology, 
Harvard University, Cambridge, 
Massachusetts 

SMSLP — Smithsonian Marine Station at Link 
Port, Ft. Pierce, Florida 

*USNM — National Museum of Natural His- 
tory, Smithsonian Institution, Wash- 
ington, D.C. 



TAXONOMIC DESCRIPTIONS 

Family GALEOMMATIDAE GRAY, 1840 

Gray (1840: 154) introduced the family 
name Galeommidae for Galeomma Turton, 
1825 (erroneously spelled "Galeomidae" by 



BIOLOGY OF DIVARISCINTILLA 



177 



Gray, 1842: 78). It has been used in that form 
by various authors (e.g. H. & A. Adams, 1 857; 
Tyron, 1872; Kisch, 1958). Dall (1899: 875) 
emended the spelling without explanation to 
Galeommatidae, and it is this form that is now 
in common use (e.g. Thiele, 1934; Popham, 
1939; Vokes, 1980; Chavan, 1969; В. Morton, 
1973; Abbott, 1974; Boss, 1982). Dall's action 
was a justified emendation of an incorrect 
original spelling [ICZN, 1985: Art. 29(b)(i), 
32(c)(iii)] because the Greek noun o/xfjca pro- 
vides the stem ommat- for the formation of the 
family name. As a justified emendation, Ga- 
leommatidae bears Gray, 1840, as authority 
and date [ICZN, 1985: Arts. 11(f)(ii), 19(a)(i)]; 
the name of the superfamily is accordingly 
Galeommatoidea [= Galeommatacea] Gray, 
1840. The nominal superfamilies Leptonoidea 
Gray, 1847; Erycinoidea Deshayes, 1850; 
and Chlamydoconchoidea Dall, 1889, are 
here considered junior synonyms. 

DIVARISCINTILLA POWELL, 1932: 66. 

Type species: Divariscintilla maoria Powell, 
1932 (by original designation). Recent, New 
Zealand. 

DIVARISCINTILLA YOYO, SP. NOV. (FIGS. 
1, 3, 5, 6, 8-11, 26) 

Material examined: Holotype: 5.3 mm [pre- 
served mantle length], USNM 860036. Para- 
types (8): 5.8, 4.7 mm, USNM 860037; 8.3, 
7.5 mm, MCZ 297406; 6.0, 5.2 mm, IRCZM 
064:01721; 5.7, 4.7 mm, DMNH 175516. To- 
tal material: 83 specimens: FLORIDA: Ft. 
Pierce Inlet: March 1987, 4; 2-3 May 1987, 
19 (including MCZ paratypes); 24 June 1987, 
4; 03 August 1987, 6; 14 August 1987, 7; 31 
August 1987, 17 (including USNM, IRCZM, 
and DMNH type specimens); 28 December 
1 987, 1 ; 1 1 March 1 988, 2; 1 2 April 1 988, 1 3. 
—Sebastian Inlet: 30 December 1987, 10. 

Type locality: Ft. Pierce Inlet, St. Lucie 
County, Florida, 27°28.3'N, 80°17.9'W, in 
Lysiosquilla scabricauda burrows on intertidal 
sand flats with patches of the seagrass 
Halodule wrightii Ascherson. 

Diagnosis 

Animal translucent white. Mantle thick, sur- 
face granular, falling into large creases. Two 
long "cephalic" palliai tentacles; one very 
short mid-dorsal tentacle just anterior to ex- 



current siphon. Shell wedge-shaped, elon- 
gate-pointed anteriorly, with weak internal 
ribs, length approximately 40% of extended 
mantle length. "Flower-like organs" on ante- 
rior surface of visceral mass ventral to labial 
palps, numbering 3-7 (usually 5). 

Description 

External features and mantle: Living ex- 
tended animal (Fig. 1) 10-15 mm in length, 
globular in general shape, entirely translucent 
white, except for dark upper portion of diges- 
tive gland. Shell nearly completely enclosed 
by thick mantle with granular external surface 
(Fig. 6; formed by middle palliai fold) falling 
into large creases, and with sparse, minute 
papillae; mantle thinner, more translucent, 
and with scattered, relatively larger papillae in 
smaller (approximately 7 mm) specimens. 
Anteropedal palliai opening wide, forming ex- 
tensive anterior cowl (Fig. 1, c). Two long, 
retractable, "cephalic" tentacles (Fig. 1 , cpht) 
anterodorsally just behind cowl. Very short (< 
1 mm) median palliai tentacle (Fig. 1, mpt) on 
dorsal midline anterior to excurrent siphon 
(Fig. 1, exs). Each tentacle with central core 
of longitudinal muscle and nerve fibers, visi- 
ble as an inner thread under low magnifica- 
tion. Mantle fused dorsally from edge of cowl 
to excurrent siphon located posterodorsally 
and often on conspicuous rounded protuber- 
ance (dependent on degree of palliai expan- 
sion); fusion interrupted only by small circular 
(approximately 2 mm diameter) foramen (Fig. 
1, uf) just over umbo of shell. Mantle fused 
posteroventrally from excurrent siphon to 
mid-ventral point (Fig. 1 , mf ) at postenor end 
of anteropedal opening. Inner palliai fold 
highly muscular; fibers continuous with mus- 
cles of central "core" of tentacles. 

Preserved animals characterized by gen- 
eral globular appearance, with retracted 
cephalic tentacles, mantle-covered shell (with 
small foramen over umbo), contracted cowl 
and excurrent siphon. Foot usually com- 
pletely withdrawn into palliai cavity. 

Shell (Figs. 8-11). Thin, transparent to 
translucent white except for yellow prodisso- 
conch and network of opaque white coloration 
on early portion; equivalve, showing fine 
growth lines; oval initially, changing to anteri- 
orly elongate-pointed with angulate corners 
anteriorly and posteriorly near attachment 
points of adductor muscles; weak internal ra- 
dial ribs corresponding in placement to weak 



178 



MIKKELSEN & BIELER 





< ft-bag byg vgr mf 



pef 




cpht uf sh 



mpt 



FIGS. 1-5. External appearance and internal shell morphology. 1. Divahscintilla yoyo in crawling position, 
from left side. 2. D. troglodytes, same as Fig. 1. 3. D. yoyo, internal surface of right valve, showing 
approximate location of muscle insertions. 4. D. troglodytes, same as Fig. 3. 5. D. yoyo, in hanging position, 
from right side. Scale bars: 1 , 2, 5 = 2.0 mm; 3, 4 = 1 .0 mm. (aam, anterior adductor muscle; apr, anterior 
pedal retractor muscle; apt, anterior palliai tentacle; bag, byssus adhesive gland; by, byssus; byg, byssus- 
gland; c, cowl; cpht, cephalic tentacles; ct, ctenidium; exs, excurrent siphon; fl, flower-like organ; ft, foot; mf, 
point of ventral mantle fusion; mpt, median palliai tentacle; pam, posterior adductor muscle; pef, posterior 
extension of foot; ppm, pedal protractor muscle; ppr, posterior pedal retractor muscle; ppt, posterior palliai 
tentacles; sh, shell; uf, umbonal foramen; vgr, ventral groove). 



external grooves. Muscle scars indistinct. 
Hinge (Fig. 10) primarily internal, with vjeak 
external ligament, stronger Internal resilium 
and two rudimentary, non-interlocking cardi- 
nal teeth; lateral teeth absent. Shell nearly 
completely enclosed In chamber formed by 
palliai layers, communicating with exterior via 
small umbonal foramen (Fig. 1, uf). Perio- 
stracal groove lying between Inner and outer 
paillai folds. Size small In relation to body, 
extending only over dorsal portion of visceral 
mass; length approximately 40% of relaxed 
mantle length. Permanently gaping at 110- 
120° angle while relaxed. Incapable of closure 
more than 50°. Periostracum colorless, most 
evident as perlostracal webbing extending 
between valves, anterior and posterior to 
hinge, and at periphery of shell. Shell micro- 
structure (Fig. 1 1) cross-lamellar, with thin ho- 
mogeneous layer on either side. 



Prodlssoconch (Fig. 1 6) approximately 350 
jxm In length, having distinct prodlssoconch I 
and prodissoconch II stages; prodlssoconch I 
approximately 140 fxm in length, with "granu- 
lated" surface sculpture and marginal radial 
ridges; prodlssoconch II stage relatively large, 
sculptured with distinct concentric ridges. 
Abrupt demarcation between prodissoconch 
and dlssoconch. 

Organs of the palliai cavity: Foot (Figs. 1 , 
5, 18-19, ft, pef) highly extensible, with 
hatchet-shaped anterior crawling portion and 
elongated tubular posterior extension. Ante- 
rior portion internally with dorsal haemocoel, 
sparse longitudinal musculature, and with ex- 
tensive lateral and ventral external dilation 
and accompanying mucous glands (staining 
turquoise to dark blue In APH). Ventral groove 
(Figs. 1, 18, 21, vgr) extending from approx- 



BIOLOGY OF DIVARISCINTILLA 



179 




FIGS. 6-7. Exterior palliai surfaces (SEM). 6. Divariscintilla yoyo. 7. D. troglodytes; arrows indicate enlarged 
palliai papillae. Scale bars = 50 (хгл. 



¡mate midpoint of anterior tip to terminus of 
posterior extension, heavily ciliated interiorly 
along entire length. Byssus-gland (Figs. 1,21, 
byg) restricted to very small area on either 
side of ventral groove, a short distance be- 
hind anterior end of groove in vicinity of pedal 
ganglia (see below); closely associated with 
numerous blood spaces; faintly whitish in liv- 
ing animal, staining dark blue in methylene 
blue, purple in APH. Posterior extension ter- 
minally pointed (Fig. 19), consisting internally 
largely of longitudinal muscle fibers and con- 
nective tissue; proximal half with free edges 
of ventral groove heavily ciliated externally; 
external ciliation and accompanying mucous 
glands disappearing abruptly to leave distal 
half of posterior extension with ciliation re- 
stricted to interior of groove. Byssus adhesive 
gland [Figs. 1, 22, bag; see Ecology and Be- 
havior (below) for explanation of term] just 
short of terminus of posterior extension, with 
internal lamellae surrounding a common lu- 
men; opaque white in living animal, staining 
dark blue in methylene blue, light purple in 
APH. 

Anterior and posterior adductor muscles 
subequal, long, of moderate diameter. Attach- 
ment ends oval, subequal. Anterior pedal re- 
tractor slightly smaller in diameter than, and 
attaching to shell anterior to, anterior adduc- 
tor; posterior pedal retractor smaller than, and 
attaching to shell posterodorsal to, posterior 
adductor. Very small pedal protractor attach- 
ing to shell dorsal to anterior adductor. 
Integument of visceral mass with numerous 



longitudinal muscle fibers, most highly con- 
centrated posteriorly; these continuous with 
anterior and posterior pairs of pedal retractor 
muscles, only slightly smaller in diameter than 
adductor muscles. Small anterior pedal pro- 
tractor originating ventral to anterior adductor 
muscle within anterior tissues of digestive 
gland, passing posterodorsally to insert on 
shell, dorsal to anterior adductor. Muscles 
leaving no visible attachment scars on shell; 
approximate insertion locations shown in 
Fig. 3. 

Labial palps (Fig. 28, Ip) large, oval, each 
with 10-14 lamellae each side; each pair 
fused at midline near mouth. Outer palps at- 
tached laterally by elongated strip of tissue to 
inner surface of mantle lining shell; inner 
palps similarly attached to surface of visceral 
mass. Cilial currents moving particles oral- 
ward on inner palp surfaces, laterally toward 
ctenidia on outer palp surfaces. Ctenidia (Fig. 
25) eulamellibranch, homorhabdic, hanging in 
loosely pleated folds, with numerous fila- 
ments; inner and outer demibranchs on each 
side with both ascending and descending 
lamellae. Both demibranchs with well-devel- 
oped, numerous interfilamentary junctions 
(Fig. 24, ifj; Fig. 25, arrow), and evenly dis- 
tributed, albeit few, interlamellar junctions 
(Fig. 24, ilj). Inner demibranch approximately 
50% larger, extending farther ventrally and 
anteriorly than outer, with food groove (Fig. 
25, fg) at free edge; anterior end (with termi- 
nus of food groove) extending between labial 
palps. Ventral tips of anterior filaments of in- 



180 



MIKKELSEN & BIELER 




FIGS. 8-15. Shells and shell structure (SEM). 8. Divariscintilla yoyo, left valve, external view, 3.9 mm 
[maximum dimension]. 9. D. yoyo, right valve, internal view, 3.0 mm. 10. D. yoyo, hinge, anterior to left. 1 1 . 
D. yoyo, microstructure, with internal surface at top. 12. D. troglodytes, same as Fig. 8, 5.9 mm. 13. D. 
troglodytes, same as Fig. 9, 5.3 mm. 14. D. troglodytes, same as Fig. 10. 15. D. troglodytes, same as Fig. 
11. Scale bars: 10 = 100 цт; 11,15 = 5 jim; 14 = 200 |xm. 



BIOLOGY OF DIVARISCINTILLA 



181 




FIGS. 16-17. Prodissoconch and larval shell morphologies (SEM). 16. Prodissoconch (Divahscintilla yoyo); 
arrow indicates boundary between prodissoconch I and II. 17. Shell of newly hatched larva (D. troglodytes); 
arrow indicates zone of initial shell formation. Scale bars: 16 = 50 jim; 17 = 10 |xm. 




FIGS. 18-19. Foot (SEf^). 18. Ventral view of entire foot, showing ventral groove {Divariscintilla troglodytes). 
19. Terminus of posterior foot extension, showing byssus-threads (D. yoyo). Scale bars: 18 = 200 fi,m; 19 
= 50 M-m. (by, byssus; pef, posterior extension; vgr, ventral groove). 



ner demibranch "not inserted into a distal oral 
groove" (Stasek, 1963, Category III); antero- 
ventral margin of inner demibranch fused to 
inner palp lamella. Outer demibranch shorter, 
without food groove; margins inserting be- 
tween inner and outer labial palps, along up- 
per portion of visceral mass, and on inner sur- 
face of mantle to posterior end of palliai 
cavity. Cillai currents as in Fig. 30, moving 
food particles in food groove and in groove 
between demibranchs oralward. Tract on in- 
ner surface of cowl anterior to palps as exit 
point for removal of waste particles. No cilial 



currents evident on inner palliai surface or 
surface of visceral mass. 

Visceral mass of brownish digestive gland 
anterodorsally, whitish granular-appearing 
gonad posteriorly, distally, and, in ripe speci- 
mens, overlapping digestive tissue laterally 
and dorsally; red pedal ganglia visible at distal 
end of gonad in base of foot. 

Cluster of 3-7 (most often 5; x = 4.7, n = 
31) nonretractable "flower-like organs" (Figs. 
23, 26) originating on anterior surface of vis- 
ceral mass just ventral to labial palps. Usually 
arranged as single organ closest to palps, 



182 



MIKKELSEN & BIELER 



20 




mg 



man-— ^ 



mus 





man 



vgr 



4W 



Лкл% . mus 



pn Vp 






mug 





bag 



áP~^'am 










FIGS. 20-24. Internal structure, histological sections. 20. Cross-section at level of esophagus and anterior 
edge of ctenidia (Divariscintilla yoyo). 21 . Cross-section through foot at region of byssus-gland (D. yoyo). 22. 
Longitudinal section through byssus adhesive gland at terminus of posterior foot extension (D. yoyo). 23. 
Longitudinal thin section through two "flower-like organs" (D. yoyo). 24. Cross-section of ctenidium showing 
interlamellar and interfilamentary junctions (D. troglodytes). Scale bars: 20 = 300 ixm; 21, 24 = 100 jim; 
22, 23 = 50 цт. (bag, byssus adhesive gland; byg, byssus-gland; ct, ctenidium; dg, digestive gland; es, 
esophagus; gon, gonad; idb, inner demibranch; ifj, interfilamentary junctions; ilj, interlamellar junctions; lam, 
lamellae of byssus adhesive gland; Ip, labial palp; man, mantle; mg, midgut; mug, mucous glands; mus, 
muscle fibers; odb, outer demibranch; pef, posterior extension of foot; pg, pedal ganglia; pn, pedal nerve; sh, 
shell; ss, style sac; vgr, ventral groove). 



BIOLOGY OF DIVARISCINTILLA 



183 




FIG. 25. Ctenidia of Divariscintilla yoyo (SEM). Ar- 
rows indicate interfilamentary junctions. Scale bar 
= 100 (xm. (fg, food groove of inner demibranch; 
idb, inner demibranch; odb, outer demibranch). 



with successive organs commonly originating 
in pairs, ventral to first; more ventral pairs of- 
ten slightly smaller. Tendency for more nu- 
merous "flower-like organs" in larger speci- 
mens (< 5 in specimens of shell length < 4.0 
mm; > 5 in specimens of shell length > 4.0 
mm), but highly variable (smallest specimen 
= 2.1 mm shell length with 5 organs; largest 
specimen = 5.8 mm shell length with 5 or- 
gans; specimens of 2.6 and 4.7 mm shell 
length with 7 organs each). Each organ 0.3- 
0.7 mm in diameter, 0.5 mm in height (in- 
cluding "head" plus stalk); head composed of 
numerous (80-90), onion-shaped subunits, 
each opening to exterior via large pore (Fig. 
23); homogeneous stalk of blood-filled spongy 
connective tissue, without lumen and without 
obvious nervous connection. 

Digestive system: Mucous glands embed- 
ded in bases of labial palps surrounding 
mouth and anterior esophagus. Short esoph- 



agus leading from mouth into stomach at dor- 
sal center of visceral mass. Stomach (Fig. 28, 
st) round or oval, slightly elongated antero- 
posteriorly. Major openings into stomach in- 
clude those from: (1) esophagus (eso), open- 
ing anteriorly, (2) right and left digestive ducts 
(dd), opening posteroventral to esophageal 
opening, (3) left pouch (Ipch), adjacent to left 
digestive duct, and associated with shallow 
dorsal pocket [dorsal hood (dh)], (4) style sac 
(ss), opening posteroventrally, and (5) midgut 
(mg), opening just anterior to, but morpholog- 
ically separate from, style sac. Major typhlo- 
sole (ty) on mid-ventral surface of stomach as 
wide loop extending from midgut and style 
sac to right digestive duct. Faintly ridged ar- 
eas (?sorting areas) present, adjacent to 
esophageal opening and between edge of 
gastric shield and major typhlosole. Many 
small ducts to digestive gland opening into 
right and left digestive ducts and left pouch. 
Gastric shield (Fig. 28, gs and A) with thick- 
ened knob-shaped dorsal projection (upon 
which crystalline style rotates) attached to tis- 
sue flap between left digestive caecum and 
left pouch; remainder of gastric shield thin, 
wrapped around dorsal end of crystalline style 
with dorsal extension forming narrow channel 
into dorsal hood. 

Style sac extending nearly entire length of 
visceral mass, with ventral tip visible exter- 
nally within gonadal tissue on left side of foot; 
internally with typhlosole on anterior wall con- 
tinuing into stomach and communicating with 
major typhlosole; crystalline style (Fig. 28, cs) 
extending entire length of style sac, sharply 
tapered at distal end. Midgut (mg) with 2-3 
loops within anterior part of visceral mass, of- 
ten visible just below surface of integument 
on either side; typhlosole initially large, de- 
creasing in size rapidly; extending to distal 
end of visceral mass on right side, near tip of 
style sac, where it loops back to pass directly 
dorsal [as hindgut (hg)] near surface along 
posterior right side of gonad, leaving visceral 
mass near region of heart. Rectum (re) pass- 
ing posteriorly through heart and kidney; rec- 
tal glands absent. Anus (Figs. 28, 29, an) po- 
sitioned just inside excurrent siphon. Fecal 
strands of irregular length and varying width. 



Suprabranchial chamber (Fig. 29). Re- 
ceives openings of gonad, kidneys and diges- 
tive system. Two oval, whitish, glandular 
patches [?hypobranchial glands (hgl)] on roof 
of suprabranchial chamber, flanking rectum. 



184 



MIKKELSEN & BIELER 




FIGS. 26-27. Flower-like organs (SEM). 26. of Divariscintilla yoyo, showing typical number of five, of 
subequal size. 27. of D. troglodytes, in lateral view. Scale bars: 26 = 100 ц.т; 27 = 50 \i.m. 



Nervous system (Fig. 31 ). Nervous system 
of typical bivalve organization with three pairs 
of ganglia, red in living animal, connected by 
long commissures. Ganglia relatively large, 
subequal (length approximately 0.4 mm), 
cerebropleural ganglia more elongate than 
others. Cerebropleural ganglia (cpig) lying 
just anteroventral to anterior adductor mus- 
cle, dorsal to esophagus; cerebral commis- 
sure (cc) short; each ganglion giving rise to 
three additional major branches: (1) dorsally, 
common trunk giving rise to cephalic tentac- 
ular nerve (ctn) leading to cephalic tentacle, 
and palliai nerve (pain) leading to ventral shell 
edge as a continuous cord (to visceral gan- 
glion) with numerous smaller nerves extend- 
ing into mantle tissue; (2) ventrally, cere- 
bropleural-pedal commissure (cpc) passing 
between visceral mass and integument along 
anterior face of foot, penetrating gonadal tis- 
sue distally to join with pedal ganglion; and (3) 
laterally, cerebrovisceral commissure (cvc) 
passing through upper portion of visceral 
mass to visceral ganglion. Pedal ganglia 
(Figs. 21, 31, pg) closely fused at midline; 
showing through integument of foot in living 
animal as red organ at extreme distal end of 
gonad; each dorsally receiving cerebropedal 
commissure from cerebropleural ganglion; 



each ventrally giving rise to two anterior and 
one posterior pedal nerves. Statocyst (Fig. 
31 , stc) on posterodorsal face of each pedal 
ganglion; each with one spherical statolith 
(sti) 50 |xm in diameter. Visceral ganglia (vg) 
joined together at midline (but not as closely 
fused as pedal ganglia), lying ventral to heart, 
posteroventral to posterior adductor muscle; 
each receiving cerebrovisceral commissure 
(anterolaterally) and palliai nerve (dorsolater- 
ally) from cerebropleural ganglion; each giv- 
ing rise ventrolaterally to branchial nerve (with 
somewhat swollen base), which extends 
along common axis of inner and outer demi- 
branchs of ctenidium. 



Reproductive system: Simultaneous her- 
maphrodite. Ovotestis white, encompassing 
most of volume of visceral mass, extending 
from small portion in umbonal area, down 
posterior surface of digestive gland, and ex- 
panding to surround pedal ganglia and ventral 
extensions of style sac and intestinal loops. 
"Spent" appearance sparse, with pedal gan- 
glia fully exposed at terminus, and with silvery 
ducts clearly visible, packed with mature 
spermatozoa. Common genital ducts with 
large ciliated openings, emptying into supra- 



BIOLOGY OF DIVARISCINTILLA 

29 



185 




pg apn PP" 



FIGS. 28-31. Anatomical structures in Divariscintilla species. 28. Visceral mass, from right side, showing 
stomach, opened laterally. Midgut (mg) with four cross-sections, showing reduction of typhlosole. (A) = 
gastric shield, excised. 29. Roof of suprabranchial chamber, with anterior end up and inner lamellae of right 
and left inner demibranchs removed. 30. Diagrammatic representation of ctenidial demibranchs in cross- 
section, showing cilial currents. 31. Nervous system (semi-diagrammatic). Tentacles with nerves drawn as 
broken lines present only in D. troglodytes; remainder identical for both species. Sizes of ganglia and lengths 
of nerves not drawn to scale. Scale bars = 1.0 mm. (an, anus; apn, anterior pedal nerves; aptn, anterior 
pallia! tentacular nerve; bn, branchial nerve; cc, cerebropleural commissure; cpc, cerebropleural-pedal 
commissure; cs, crystalline style; cpig, cerebropleural ganglion; ct, ctenidium; ctn, cephalic tentacular nerve; 
cvc, cerebropleural-visceral commissure; dd, right and left digestive ducts; dg, digestive gland; dh, dorsal 
hood; eso, esophageal opening; exs, excurrent siphon; fl, flower-like organs; ft, foot; go, gonadal opening; 
gon, gonad; gs, gastric shield; hg, hindgut; hgl, ?hypobranchial gland; idb, inner ctenidial demibranch; ki, 
kidney; kio, kidney opening; Ip, labial palps; Ipch, left pouch; mg, midgut; mptn, median palliai tentacular 
nerve; odb, outer ctenidial demibranch; pain, palliai nerve; pam, posterior adductor muscle; pg, pedal 
ganglion; ppn, posterior pedal nerve; ppr, posterior pedal retractor muscle; pptn, palliai tentacular nerve; re, 
rectum; sh, shell; ss, style sac; st, stomach; stc, statocyst; sti, statolith; ty, major typhlosole; vg, visceral 
ganglion; vm, visceral mass). 



186 



MIKKELSEN & BIELER 



branchial chamber just anterior to visceral 
ganglia (Fig. 29, go). 

Ova small, approximately 20 fxm in diame- 
ter (maturity not determined, measured in 
paraffin sections, in posterodorsal region of 
gonad). Spermatozoa approximately 7 цт in 
head length (acrosome + nucleus + middle 
piece); nucleus cylindrical, asymethcal; ac- 
rosome subterminal, dish-shaped with central 
"papilla," tilted approximately 45° from long 
axis of nucleus; middle piece short; tail long. 
Gametes and gametogenesis to be described 
in detail in a paper currently in preparation. 

Brooding large number of small larvae for 
an undetermined period (longest time ob- 
served 14 days). Larvae held primarily within 
outer demibranch, and in suprabranchial 
chamber where they are circulated via palliai 
expansions and contractions. During brood- 
ing, excurrent siphonal opening constricted 
by sphincter-like muscles around plug formed 
by free end of rectum (often expanded into 
bulb by haemocoelic pressure), allowing di- 
gestive processes to continue during brood- 
ing and preventing loss of larvae through ex- 
current siphon. Larvae initially white, turning 
pink with shell development; released as 
straight-hinged veligers with apical flagella, 
122-138 |jLm in shell length (x = 131.8 |xm, n 
= 20; Fig. 17). Larval shell with distinct zone 
of initial shell formation (Fig. 17, arrow). Lar- 
vae expelled through excurrent siphon via 
strong contractions of shell and palliai mus- 
cles. Adults brooding larvae collected in May 
and June 1987, and April 1988. 



one other location approximately 45 km north, 
Sebastian Inlet, Brevard County, Florida. 

Etymology: A noun in apposition from the 
English vernacular 'уо-уо. ' a child's toy orig- 
inating in China about 1000 B.C., referring 
here to the periodic up-and-down motion of 
the bivalve as it hangs from its byssus-thread 
The word "yo-yo" is in Tagalog, an Indone- 
sian language, for a similarly constructed, six- 
teenth-century hunter's weapon made of 
large wooden disks and twine. 

DIVARISCINTILLA TROGLODYTES, SP. 
NOV. (FIGS. 2, 4, 7, 12-15, 27). 

Material examined: Holotype: 7.7 mm [pre- 
served mantle length], USNM 860038. Para- 
types (9): 6.7, 4.5 mm, USNM 860039; 6.8, 
6.6 mm, MCZ 297407; 7.0, 5.0 mm, IRCZM 
064:01722; 4.9, 4.8, 4.5 mm, DMNH 175517. 
Total material: 87 specimens: FLORIDA: Ft. 
Pierce Inlet: 10 March 1987, 1; 2-3 May 
1987, 12 (including MCZ paratypes); 24 June 
1 987, 1 2; 03 August 1 987, 4; 1 4 August 1 987, 
7; 31 August 1987, 13 (including USNM and 
IRCZM type specimens); 28 December 1987, 
6; 11 March 1988, 6; 12 April 1988, 13. — 
Sebastian Inlet: 30 December 1987, 13 (in- 
cluding DMNH paratypes). 

Type locality. Ft. Pierce Inlet, St. Lucie 
County, Florida, 27°28.3'N, 80°17.9'W, in 
Lysiosquilla scabricauda burrows on intertidal 
sand flats with patches of the seagrass 
Halodule wrightii Ascherson. 



Circulatory system: Heart just posterior to 
umbones, within pericardium lined by brown- 
ish pericardial gland. Doughnut-shaped ven- 
tricle traversed by intestine; auricles lateral to 
ventricle, inconspicuous. Blood vessels not 
evident; major haemocoelic spaces present 
within foot, tentacles and main axes of demi- 
branchs. 

Excretory system: Kidney yellow, ventral to 
heart, dorsal to visceral ganglia, surrounding 
pedal retractor muscles on roof of suprabran- 
chial chamber. Paired ciliated renopericardial 
apertures opening anteriorly into ventral wall 
of pehcardium. Paired renal openings into 
the suprabranchial chamber large, funnel- 
shaped, adjacent to visceral ganglia. 

Distribution: Known only from the type lo- 
cality. Ft. Pierce Inlet, St. Lucie County, and 



Diagnosis 

Animal translucent yellowish-white. Mantle 
thin, surface granular with numerous evenly 
distributed short papillae. Two long "cephalic" 
tentacles; two short anterior palliai tentacles; 
three long posterior palliai tentacles surround- 
ing posterodorsal excurrent siphon. Shell 
oval, elongate-rounded anteriorly, with weak 
internal ribs, length approximately 50% of ex- 
tended mantle length. A single "flower-like 
organ " on anterior surface of visceral mass 
ventral to labial palps. 

Description 

External features and mantle: Living ex- 
tended animal (Fig. 2) 10-15 mm in length, 
globular to oval in general shape, translucent 
white to yellowish, except for dark upper por- 
tion of digestive gland showing through tis- 



BIOLOGY OF DIVARISCINTILLA 



187 



sues. Shell nearly completely enclosed by rel- 
atively thin mantle, clearly revealing outlines 
of shell and ctenidia. Extended mantle some- 
what posteriorly elongated, with granular sur- 
face and numerous, evenly distributed, short 
papillae (Fig. 7, arrows). Anteropedal pallia! 
opening and cowl as in Divariscintilla yoyo. 
Two long, retractable, "cephalic" tentacles 
anterodorsally just behind cowl; two shorter, 
retractable anterior palliai tentacles originat- 
ing laterally to cephalic pair. Posterodorsal 
excurrent siphon on prominent rounded pro- 
tuberance; palliai fusion as in Divariscintilla 
yoyo. Three long posterior palliai tentacles 
surrounding excurrent siphon: one anterior 
and mid-dorsal, two lateral. Umbonal foramen 
(Fig. 2, uf) a transverse slit-like opening, ca- 
pable of distention during periods of stress to 
expose nearly entire shell. 

Preserved animals with shell nearly com- 
pletely exposed by retraction of umbonal fo- 
ramen, retracted pallia! tentacles, contracted 
cowl and excurrent siphon. Foot often anteri- 
orly protruding from palliai cavity. 

Shell {F\gs. 12-15). Nearly completely en- 
closed in chamber formed by pallia! layers, 
communicating with exterior via slit-like um- 
bonal foramen (Fig. 2, uf). Shell length ap- 
proximately 50% of relaxed mantle length, ex- 
tending over dorsal half of visceral mass, and 
anterior portion of ctenidia. Permanently gap- 
ing at 80-120° angle while relaxed, incapable 
of closure more than 50°. Shell (Figs. 12-13) 
thin, transparent to translucent white except 
for yellow prodissoconch and network of 
opaque white coloration on early portion; 
equivalve, oval, roundly elongated anteriorly, 
showing fine growth lines; weak internal ra- 
dial ribs strongest at periphery, corresponding 
in placement to weak external grooves. 
Slightly scalloped edge formed by ends of ra- 
dial ribs, also evident on former heavier 
growth lines. Periostracum, hinge (Fig. 14), 
internal muscle scars (Fig. 4), prodissoconch, 
and microstructure (Fig. 15) as in Divariscin- 
tilla yoyo. 

Organs of the palliai cavity: Foot, visceral 
mass, and ctenidia as in Divariscintilla yoyo. 
Labial palps of same general structure as 
those of D. yoyo, but each with more numer- 
ous (14-20) lamellae each side. Adductor, 
pedal, and internal palliai musculature similar 
to that in D. yoyo. Core muscle bundles of all 
tentacles continuous with palliai muscle layer; 
single posterior tentacle receiving muscles 



from both sides of midline. "Flower-like 
organ" (Fig. 27) always single, similar in gen- 
eral size and form to those of D. yoyo, head 
with fewer (approximately 25) subunits. 

Nervous system: As described for Divari- 
scintilla yoyo. Anterolateral and paired poste- 
rior tentacles, as well as single posterior ten- 
tacle receiving innervation from branches of 
the palliai nerve (Fig. 31, aptn, mptn, pptn). 

Reproductive system: Simultaneous her- 
maphrodite. Ovotestis, ova and spermatozoa 
as in Divariscintilla yoyo. Gametes and game- 
togenesis to be described in detail in a paper 
currently in preparation. 

Divariscintilla troglodytes broods its larvae 
in the suprabranchial chamber and outer 
demibranch as in D. yoyo for an undeter- 
mined period (longest time observed 29 
days). Adults brooding larvae were collected 
in June and December 1987. White larvae 
with initial stages of shell measured at 68 |xm 
diameter. Newly released straight-hinged 
veligers 120-130 fxm in length (x = 126.1 
ixm, n = 20). 

Digestive, circulatory and excretory sys- 
tems: As described for Divariscintilla yoyo. 

Distribution: Same as that of Divariscintilla 
yoyo. 

Etymology: A noun in apposition from the 
Greek тршуАоотг}^ = troglodytes, a hole- or 
cave-dweller. 



ECOLOGY AND BEHAVIOR 

Neither Divariscintilla yoyo nor D. troglo- 
dytes was ever found actually attached to a 
stomatopod, either in the field or in museum 
specimens (IRCZM). They are assumed to be 
free-living within the burrow near the open- 
ing(s), although specimens were never visible 
at the opening prior to pumping. However, in 
spite of other burrowing invertebrates in the 
area (callianassid shrimps, polychaetes, sip- 
unculans), neither Divariscintilla species has 
ever been collected in any habitat other than 
a Lysiosquilla burrow. The two species were 
often collected together [and often also with 
two species of vitrinellid gastropods (Bieler & 
Mikkelsen, 1988) and two of another galeom- 
matid genus] from a single Lysiosquilla bur- 
row (of 21 burrows with Divariscintilla, 1 2 con- 



188 



MIKKELSEN & BIELER 






Fig. 32. Diagrammatic representation of (A) crawling, (B,C) byssus-thread production, (D) hanging, (E,F) 
"yoyo" response to stimuli, and (G) crawling to break byssus attachment. 



tained both, 9 contained only one species). 
Densities were low, frequently of only one or 
two specimens per species per burrow sam- 
ple; the highest number of specimens in one 
burrow was 13 in the case of either species, 
and not all burrow samples included galeom- 
matoideans. However, it must be noted that in 
no case was an entire stomatopod burrow ex- 
cavated and assessed for mollusks; the 
yabby pump only effectively samples its own 
length (0.5-1 .0 m) of the burrow adjacent to 
an opening. Estimates of occurrence and/or 
density of any clams living in the deeper hor- 
izontal section of the burrow was not possible 
using this method. 

Small (0.5 mm length) parasitic worms 
(Trematoda: ?Dlgenea) were found encased 
in small tissue pockets on the palliai layer lin- 
ing the inner shell surface of a specimen of 
Divariscintilla troglodytes. Density per clam 
and frequency of occurrence were not as- 
sessed. 

Living animals spend much of their time in 
the laboratory "hanging" from byssus-threads 
from the water surface, or, more frequently, 
on the sides of the aquarium or finger bowl. 
The hanging sequence is depicted in Fig. 32. 



The byssus-threads (usually two) are pro- 
duced by the byssus-gland located in the 
crawling portion of the foot (Fig. 19). Strong 
pulsing of the byssus-gland area during 
thread production (Fig. 32, b) is probably 
caused by engorgement of the many blood 
spaces in the vicinity; this area remains 
somewhat swollen for a short time after com- 
pletion. The threads are attached immediately 
to the substrate, usually with a V-shaped at- 
tachment. As they are completed, the threads 
appear to be "picked up" (Fig. 32, c) by the 
byssus adhesive gland at the terminus of the 
posterior elongation so that they are secured 
within the ventral groove between the two 
glands; details of this process are unclear. 
The "hanging" animal thereafter appears to 
be suspended from the posterior tip of the foot 
(Figs. 5; 32, d). While relaxed in this posture, 
the tentacles and posterior siphon are ex- 
tended, the ventral edges of the cowl are 
pursed together forming a functional incurrent 
siphon, and the crawling portion of the foot is 
partially withdrawn into the palliai cavity. Pe- 
riodically, and especially in response to exter- 
nal stimuli, the adductor muscles, and mus- 
cles of the mantle, tentacles and foot contract 



BIOLOGY OF DIVARISCINTILLA 



189 



simultaneously producing rapid movement 
upward toward the byssus attachment point. 
This is followed by gradual relaxation and re- 
turn to the resting/feeding posture. This ac- 
tion (Fig. 32, e,f) resembles the up-and-down 
motion produced with a child's yo-yo toy, and 
suggested the name for one of the species 
described here. 

During normal "hanging," the posterior foot 
extension is capable of stretching to a length 
approximately 1-2 times the mantle length. 
One specimen of Divariscintilla troglodytes, 
whose byssus adhesive gland was accidently 
severed during examination, was able to pro- 
duce a byssus-thread and hang, although the 
threads did not lie within the full extent of the 
ventral groove; the distal half of the posterior 
extension remained curled at the side of the 
animal, unextended and unused. These ob- 
servations lead to the conclusion that the 
glandular structure at the terminus of the pos- 
terior foot extension serves to secure the bys- 
sus-threads within the full length of the ventral 
groove. It is likely that this gland secretes an 
adhesive substance for this purpose, there- 
fore we refer to it here as the "byssus adhe- 
sive gland." The fact that the injured animal 
could still use part of the ventral groove for 
hanging suggests that the extensive ciliation 
and mucous secretion of the proximal part of 
the groove also serve to secure the threads. 
This specimen was maintained and observed 
in the laboratory; two weeks after severing the 
tip of the posterior extension, the animal was 
observed to be hanging normally and the tip 
of the foot with its whitish gland had regener- 
ated and regained function. 

When dislodged, the clams actively crawl 
about, using an even, gliding motion pro- 
duced by ciliary action on the ventral surface 
of the foot. Sudden contractions of the shell 
and palliai muscles frequently occur during 
crawling, and, although not providing any sig- 
nificant forward propulsion, probably assist 
the animal in moving its not-so-streamlined 
body, as well as in cleansing the palliai cavity. 
The anterior unslit tip of the foot is continually 
actively "seeking" appropriate substrate. The 
terminal half of the posterior extension, which 
is unciliated externally, is not active in crawl- 
ing, being carried behind either in a trailing 
curl or with the tip bearly touching the sub- 
strate behind. The clams were observed to 
dislodge themselves voluntarily from labora- 
tory substrate via initiation of crawling activity, 
thus stretching the byssus-threads until 
breaking occurred (Fig. 32, g). It is assumed 



from laboratory observations that the animals 
spend most of their time in the burrows at- 
tached to the smoothly packed walls, and that 
crawling is utilized only when relocation is 
necessary or when dislodged by external 
forces. 



DISCUSSION 

The two new species described here are 
remarkably similar to each other in morphol- 
ogy and behavior. Significant differences are 
found in shell morphology, the number of 
"flower-like organs," and the nature of the 
mantle, including color, thickness, papulation, 
and number of palliai tentacles (Table 1). 
Thus far, both species are known only from 
the specimens studied and cited here; shells 
are unknown from dry collections (AMNH, 
Dr\/INH, IRCZM, USNM), probably because of 
their fragile nature. 

The two new species agree closely in anat- 
omy, habitat, and behavior with those de- 
scribed for Divariscintilla maoria Powell, 
1932, type and sole described species of the 
genus, by Judd (1971). Some of the features 
reported here (e.g. stomach, nervous system, 
reproductive anatomy, shell musculature, cir- 
culatory and excretory systems) were not dis- 
cussed for D. maoria, and others (e.g. 
ctenidia, "flower-like organ," byssus appara- 
tus) were described in less detail (Judd, 
1971). Differences between our two species 
and D. maoria in shell, palliai, and perhaps 
ctenidial characters (see below; Table 1 ) are 
weak when weighed against the many simi- 
larities, and, we accordingly place our new 
species in Divariscintilla. 

Shell and musculature: The most distinct 
difference between the two species described 
here and the type species of Divariscintilla is 
in the shape of the shell. Divariscintilla maoria 
has a ventral notch in each valve, a character 
used at the generic level (Chavan, 1969) to 
treat Divariscintilla as a subgenus under Vas- 
coniella Dall, 1899, whose members possess 
a likewise-notched shell. The shell of Vas- 
coniella jeffreysiana (P. Fischer, 1873), the 
type and sole described species of the genus, 
however, is greatly inequivalve and notched 
in only the right valve (Kisch, 1958). As noted 
by Judd (1971 : 344), the ventral notch of Di- 
variscintilla maoria "does not appear to be 
functionally important," as it is not associated 
with a cleft in the mantle nor with the passage 



190 



MIKKELSEN & BIELER 



TABLE 1. Distinguishing characteristics of the three described species of Divariscintilla (for additional 
information, see text). 





D. maoria 








(from Judd, 1971) 


D. yoyo 


D. troglodytes 


Shell: 








General shape 


oval 


elongate-pointed 


oval 




ventrally notched 


unnotched 


unnotched 


Prolongation 


posterior 


anterior 


anterior 


Sculpture 


unribbed 


unribbed 


internal radial ribs 


Length relative to 








mantle length 


68% 


40% 


50% 


f^antle: 








Color, thickness 


(not given) 


whitish, thick 


yellowish, thin 


Extent covering shell 


margins only 


entire 


entire 


Papillae 


very small 


sparse, very small 


numerous, small, 
evenly-distributed 


Anterior tentacles 


4 long 


2 long 


2 short, 2 long 


Posterior tentacles 


1 long 


1 very short 


3 long 


Defensive appendages 


6-8 present 


absent 


absent 


Flower-like organs: 








number 


1 


3-7 (usually 5) 


1 


Ctenidia: 


smooth 


pleated 


pleated 


Labial palps: 








Lamellae per palp 


(not given) 


10-14 


14-20 


Geographical range: 


New Zealand 


eastern Florida 


eastern Florida 



of byssus-threads. Therefore, we do not con- 
sider the presence of this notch a prerequisite 
to inclusion in Divariscintilla and furthermore 
do not advocate treatment of Divariscintilla as 
a subgenus of Vasconiella on this basis. 
Claims of a "tendency" within the family for 
the development of a concave or indented 
ventral shell margin (Powell, 1932; J.E. Mor- 
ton, 1957) appear overstated; a cursory sur- 
vey of the galeommatid shells illustrated by 
Chavan (1969) show that most are evenly 
rounded at the ventral margin. 

A second conchological difference between 
Divariscintilla maoria and the two new spe- 
cies is the direction of prolongation of the 
shell. All are skewed, but in opposite direc- 
tions; D. maoria is prolonged posteriorly while 
D. yoyo and D. troglodytes are prolonged an- 
teriorly. 

Divariscintilla maoria also has a relatively 
larger shell in relation to its body (maximum 
shell length approximately 68% of the relaxed 
mantle length in fig. 1 of Judd, 1971), which 
also differs by being covered by palliai folds 
only along its margins. The degree of shell 
reduction and internalization in members of 
this superfamily varies widely and is a char- 
acter which deserves further attention at su- 
praspecific levels. 

Divariscintilla yoyo and D. troglodytes both 
possess the full complement of five major 



muscles (e.g. two adductors, two pedal re- 
tractors, and one protractor). As in D. maoria 
(see Powell, 1932), these leave no muscle 
scars on the shell, even to the extent, realized 
during this study, that they do not show under 
scanning electron microscopy. 

Mantle ornamentation: The complement of 
palliai tentacles and papillae is quite different 
in Divariscintilla maoria and the two new spe- 
cies described here. The two pairs of anterior 
tentacles and the single posterior tentacle of 
D. maoria, described by Judd (1971), seem 
homologous to tentacles described in this 
study. However, D. yoyo and D. troglodytes 
do not possess anything resembling the "pos- 
terior appendages" described by Judd 
(1971). They do, however, possess numer- 
ous papillae on the exterior portion of the 
mantle, beyond the shell margins; this area is 
without papillae in D. maoria. 

Ctenidia: As shown in Judd (1971: fig. 3) 
and confirmed here (Fig. 30), the ciliary cur- 
rents of the ctenidia in Divariscintilla can be 
ascribed to type C(1) as defined by Atkins 
(1937). This type, in which only the inner 
demibranch bears a ventral marginal food 
groove, is known from a great number of gen- 
era (e.g., Galeomma: B. Morton, 1973; Cera- 
tobornia Dall, 1899: Narchi, 1966). Also like 



BIOLOGY OF DIVARISCINTILLA 



191 



most other galeommatoideans, the outer 
demibranch is shorter in Divariscintilla, and 
interfiiarnentary junctions are well-developed 
and numerous. The presence of interlamellar 
junctions in Divariscintilla (this study; no data 
available for the type species), contradicting a 
statement by B. Morton (1973: 142) that the 
presence of interfiiamentary and the lack of 
interlamellar junctions "is typical of the Lep- 
tonacea [ = Galeommatoidea] in general and 
can be correlated with the habit of incubating 
their larvae within the ctenidia . . ." (see also 
B. Morton, 1981: 97-99). Divariscintilla does 
show, however, a negative correlation be- 
tween the number of interlamellar junctions 
and the incubatory habit (see also Oldfield, 
1961:290). 

In gross morphology, the gills of Divariscin- 
tilla yoyo and D. troglodytes contrast with 
those of D. maoria and nearly all other ga- 
leommatoideans in being loosely pleated 
rather than smooth. This may be an adapta- 
tion for increasing surface area of the food- 
gathering structures of animals in habitats 
(e.g. burrows) that may present reduced food 
density. Pleating was also observed by 
Popham (1939) in Phlyctaenachlamys lysio- 
squillina Popham, 1939, another burrow- 
dwelling commensal, although it was inter- 
preted as an artifact due to preservation. 

Flower-like organs: The function of the 
"flower-like organs" of Divariscintilla species 
is not evident, although the presence of glan- 
dular tissues in the flower head plus the lack 
of nervous connections point to a secretory 
rather than sensory role. Judd (1 971 : 352) de- 
scribed a single "small median papilla on the 
dorsal edge of the anterior part" of the vis- 
ceral mass of D. maoria which is clearly the 
same structure. No function was suggested 
by Judd, however, reference to a "mucoid se- 
cretion" from subepithelial gland cells agrees 
with this study that the organs are generally 
secretory. The "flower-like organs" could 
possibly emit a pheromone for attracting re- 
productive partners in conditions of low pop- 
ulation densities. If so, their placement at the 
incurrent open'mg, requiring flow of the attrac- 
tant through the animal prior to release, is in- 
deed peculiar. 

"Flower-like organs" have not been re- 
ported in any other galeommatoidean genus. 
However, the pheromone organ and defen- 
sive papillae of Chlamydoconcha orcutti Dall, 
1884 (see B. Morton, 1981), bear at least su- 
perficial resemblance. They cannot be con- 



sidered homologous, because the structures 
described for C. orcutti arise from the middle 
palliai fold, while Divariscintilla s "flower-like 
organs" are from the surface of the visceral 
mass. Also unlike the organs of C. orcutti, the 
"flower-like organs" of Divariscintilla are not 
retractable. 

Stomach and feeding: The structure of the 
stomach was not discussed by Judd ( 1 971 ) in 
the redescription of Divariscintilla maoria. 
Stomach structure in the two species de- 
scribed here agrees well with those of others 
in this superfamily (e.g. Phlyctaenachlamys 
Popham, 1939; Galeomma:B. Morton, 1973), 
defined as type IV by Purchon (1958). Major 
common features include complete separa- 
tion of the midgut from the style sac (excep- 
tions are Pseudopythina P. Fischer, 1884: B. 
Morton; 1972, and Montacutona Yamamoto & 
Habe, 1959: В. Morton, 1980), and an arc- 
shaped major typhlosole leading toward the 
openings to the digective diverticula. Varia- 
tion occurs in the degree of concentration of 
these latter openings into caeca or ducts, and 
in the extent of the typhlosole in the midgut. 

Reproduction: Galeommatoideans are fre- 
quently cited as having "the most complex 
reproductive patterns in the Bivalvia" (Ó Foi- 
ghil & Gibson, 1984: 72). Hermaphroditism, 
the occurrence of dwarf or "complementar' 
males, and ctenidial brooding of larvae are 
common features of this superfamily and 
have been claimed to be trends associated 
with a commensal mode of life (B. Morton, 
1976; Ó Foighil, 1985). No data were pre- 
sented on the reproductive biology of Divari- 
scintilla maoria by Judd (1971: 349) other 
than noting the incubation of larvae in the "ex- 
halant chamber." The two Divariscintilla spe- 
cies described here were both found to brood 
their young within the folds of the outer demi- 
branch as well as in the suprabranchial cham- 
ber. Both species are simultaneous hermaph- 
rodites. The most unusual reproductive 
feature encountered was the morphology of 
the spermatozoa which exhibit rotational 
asymmetry: the dish-shaped acrosome is 
tilted approximately 45° from the long axis of 
the cylindrical nucleus. 

Foot and locomotion: All galeommatoide- 
ans are capable of active, snail-like locomo- 
tion, and most (e.g. Galeomma and Kellia 
Turton, 1822: Popham, 1940) possess a 
blunt, heel-like posterior foot with a ventral 



192 



MIKKELSEN & BIELER 



groove and a more-or-less postero-terminal 
byssus-gland. Divariscintilla differs frorл the 
typical, but is not unique in having an elon- 
gated posterior foot; this feature is also 
present in Phlyctaenachlamys, Rhamphi- 
donta Bernard, 1975, and Ceratobornia. The 
foot of Phlyctaenachlamys (Popham, 1939) is 
nearly identical to that of Divariscintilla, with 
an elongated posterior portion and a whitish 
organ at the terminus. In Phlyctaenachlamys, 
the byssus-gland (although likewise concen- 
trated in the central portion) continues 
throughout the posterior extension (Popham, 
1939); the "opaque white area immediately 
short of the tip" of the posterior extension 
(Popham, 1939: 64) may be similar to the 
byssus adhesive gland of Divariscintilla, al- 
though "hanging" behavior has not been de- 
scribed for Phlyctaenachlamys. 

The byssus-gland in the elongated poste- 
rior foot of Rhamphidonta retifera (Dall, 1899) 
was vaguely described by Bernard (1975) as 
centrally located; this posterior extension is 
very different from those discussed above in 
being dorsoventrally flattened and wider than 
the anterior portion. "Hanging" behavior was 
not mentioned by Bernard (1975) for Rham- 
phidonta. 

Members of both described species of Cer- 
atobornia, С longipes (Stimpson, 1855) and 
С сета Narchi, 1 966, are capable of hanging 
from a highly extensible posterior foot; the bys- 
sus-gland has been described at the extreme 
posterior tip in both [for С longipes, see Dall, 
1899: 889, pl.88, figs. 10, 11, 13 (previously 
unpublished figures of Stimpson); for С сета, 
see Narchi, 1966: 515, 517-518, text-figs. 2, 
5]. However, the byssus apparatus in Cerato- 
bornia may deserve reinvestigation in light of 
the structures found in Divariscintilla. Narchi 
(1966: 518, fig. 5) provided confusing state- 
ments about the midventral "main mucous 
gland' and byssus-gland of Ceratobornia 
cema, with the latter both at the extreme pos- 
terior in a text-figure and "extend[ing] along 
the groove throughout the posterior portion, 
excepting for the tip" in the text. The "bysso- 
genus [sic] lamellae" described for the byssus- 
gland of C. cema bear greater resemblance to 
Divariscintilla s byssus adhesive gland than to 
the byssus-gland proper. Stimpson (1855: 
1 12) reported that, in C. longipes, "there is no 
true byssus" although a "glutinous substance" 
secreted by the "opaque byssal gland" at the 
posterior terminus "may be slightly drawn 
out"; this material was re-interpreted by Dall 
(1899: 889) as a "single byssal thread." 



In Divariscintilla maoria, the white organ at 
the end of the posterior extension was inter- 
preted as the byssus-gland, secreting a single 
thread by which the animal "hangs." In view 
of the present interpretation of the foot of the 
two new species, this seems in need of re- 
investigation. 

The members of Divariscintilla, Phlyctaen- 
achlamys, and Ceratobornia have all been 
described as inhabiting the holes of burrowing 
invertebrates (Stimpson, 1855; Popham, 
1939; Narchi, 1966); although the location of 
the byssus-gland apparently differs, the ability 
to hang from an extensible, thread-like foot 
may be an advantage in attachment to verti- 
cal walls. [However, other galeommatoideans 
that lack this feature also inhabit holes, e.g. 
three species referrable to Scintilla De- 
shayes, 1856, from these same Lysiosquilla 
burrows (Mikkelsen & Bieler, pers. obs.).] 
Ceratobornia cema was also reported to at- 
tach "temporarily to the body of the shrimp 
[Callianassa major Say]" (Narchi, 1966: 522), 
although details of this attachment were not 
given. 

The byssus apparatus of Kellia suborbicu- 
laris (Montague, 1803) and of Montacuta sub- 
striata (Montagu, 1808) described by Oldfield 
(1961 : 269-270, figs. 5, 7), each consist of a 
"subsidiary" byssus-gland dorsal to the ven- 
tral groove of the foot, plus a "main" byssus- 
gland with "byssogenous lamellae" equipped 
with a duct to the posterior "heel." These bear 
superficial resemblance to the foot of Divari- 
scintilla, but need behavioral observations 
and more detailed histological examination 
for proper comparison. 

The supportive "chondroid wedge" de- 
scribed for Ceratobornia cema, Rhamphi- 
donta retifera, and several other galeomma- 
toideans (Narchi, 1966; Bernard, 1975) is not 
present in Divariscintilla. 

Although extremely similar in gross mor- 
phology, the anterior foot of Divariscintilla is 
not the "compact mass of muscle" described 
for Phlyctaenachlamys by Popham (1939). 
Longitudinal muscle fibers are sparse in the 
anterior foot of Divariscintilla, being more con- 
centrated in the posterior extension. Exten- 
sion of the anterior tip is apparently accom- 
plished by haemocoelic pressure. 

The crawling activity of many galeomma- 
toideans (e.g. Montacuta Tuúon, 1822: Gage, 
1968a; Mysella Angas, 1877: Gage, 1968b) 
has been described roughly as an extend-at- 
tach-pull fonward sequence. Divariscintilla, in 
contrast, relies primarily on ciliary action for 



BIOLOGY OF DIVARISCINTILLA 



193 



forward crawling movernent. Propulsive shell 
and palliai muscle contractions assist locomo- 
tion in Phlyctaenachlamys (Popham, 1939) 
and Galeomma (B. Morton, 1973), and at 
least to some extent in Divariscintilla. During 
these contractions in Divariscintilla, the rela- 
tively large adductor muscles partially close 
the shell, reducing the shell angle by about 
50%. This is in contrast to the "poorly devel- 
oped" adductor muscles of Phlyctaenach- 
lamys, contraction of which causes "little 
movement of the shell valves" (Popham, 
1939: 67). Retention of well-developed ad- 
ductor muscles in Divariscintilla may have 
been favored by providing additional water- 
propelling force for locomotion and cleansing 
of the palliai cavity. 

Commensalism: The nature of the associ- 
ation between these bivalves and their sto- 
matopod "host" is unclear. Although appar- 
ently more dependent upon the burrow 
habitat than on the resident stomatopod, col- 
lection records suggest a strongly obligatory 
association with Lysiosquilla for both new Di- 
variscintilla species. The large-diameter bur- 
rows of this particular stomatopod, with their 
smooth, hard-packed walls, are well suited for 
byssus attachment by the clams and provide 
a well-maintained, protective habitat. Further- 
more, strong respiratory currents produced in 
the burrow by the stomatopod are undoubt- 
edly beneficial to these more-or-less con- 
fined, filter-feeding clams. Ockelmann & 
Muus (1978) have suggested that this type of 
association is dependent upon the bivalves' 
responses to chemical/host, rather than phys- 
ical, stimuli. Indeed, neither Divariscintilla 
species described here has ever been found 
in any other habitat, protective or otherwise. 

Redescription of the genus: This work re- 
defines Divariscintilla and identifies the fol- 
lowing generic characters: Shell thin, pro- 
longed anteriorly or posteriorly, incompletely 
internalized within palliai tissues; hinge teeth 
reduced, possessing only small cone-shaped 
cardinals. Mantle ornamented with species- 
specific numbers of tentacles and papillae. 
Byssus-gland communicating by ventral 
groove with posterior byssus adhesive gland. 
Two-part foot, consisting of extensible ante- 
rior crawling portion and tubular posterior 
extension, used in active crawling and 
"hanging." Secretory "flower-like organs" on 
anterior surface of visceral mass, situated 
ventral to labial palps. Eulamellibranch 



ctenidia with interlamellar and interfilamen- 
tary junctions. Simultaneous hermaphroditic 
reproduction with ctendial incubation of lar- 
vae. 

The known range of the genus is extended 
from New Zealand alone (type species, Di- 
variscintilla maoria) into the western Atlantic 
(D. yoyo and D. troglodytes, described here). 



ACKNOWLEDGMENTS 

The following are gratefully acknowledged: 
Dr. Raymond B. Manning (USNM), for bring- 
ing this interesting fauna to our attention; 
William D. ("Woody") Lee (SMSLP), for in- 
valuable field help; Patricia A. Linley (HBOI) 
for SEM assistance and preparation of the 
sections in Fig. 23; Julianne Piraino (SMSLP) 
for SEM assistance; Tom Smoyer (HBOI) for 
photographic work; Dr. Kerry B. Clark (Florida 
Institute of Technology, Melbourne) and John 
E. Miller (HBOI) for the use of and help with 
video recording equipment; Dr. R. Tucker Ab- 
bott for literature; Solene Morris [British Mu- 
seum (Natural History)] for the loan of perti- 
nent museum specimens; Drs. Richard S. 
Houbrick (USNM), Kenneth J. Boss (MCZ), 
and one anonymous reviewer for valuable 
comments on the manuscript. 

This research was supported in part by the 
Smithsonian Marine Station at Link Port; the 
cooperation of Dr. Mary E. Rice and her team 
is gratefully acknowledged. Funding for this 
project was derived in part from a National 
Capital Shell Club scholarship to P.M., and a 
NATO Postdoctoral Fellowship at SMSLP to 
R.B. (administered by the German Academic 
Exchange Service [Deutscher Akademischer 
Austauschdienst (DAAD)], Bonn). This is Har- 
bor Branch Océanographie Institution Contri- 
bution no. 704 and Smithsonian Marine Sta- 
tion Contribution no. 237. 



LITERATURE CITED 

ABBOTT, R. T., 1 974, American Seashells: the Ma- 
rine Mollusca of the Atlantic and Pacific coasts of 
North America, 2nd ed. Van Nostrand Reinhold, 
New York, 663 pp., 24 color pis. 

ADAMS, H., & A. ADAMS, 1857, The genera of 
Recent Mollusca; arranged according to their or- 
ganization. Vol. II. John Van Voorst, London, 661 
pp. 

ATKINS, D., 1937, On the ciliary mechanisms and 
interrelationships of lamellibranchs. Part III. 
Types of lamellibranch gills and their food cur- 



194 



MIKKELSEN & BIELER 



rents. Quarterly Journal of Microscopical Sci- 
ence, 79: 375-421. 

BERNARD, F. R., 1975, Rhamphidonta gen. n. 
from the northeastern Pacific (Bivalvia, Leptona- 
cea). Journal de Conchyliologie, 1 12(3-4): 105- 
115. 

BIELER, R., & P. M. MIKKELSEN, 1988, Anatomy 
and reproductive biology of two western Atlantic 
species of Vitrinellidae, with a case of protan- 
drous hermaphroditism in the Rissoacea. The 
Nautilus, 102(1):1-29. 

BOSS, K. J., 1965, Symbiotic erycinacean bivalves. 
Malacologia, 3(2): 183-195. 

BOSS, K. J., 1982, Mollusca. Pp. 945-1166, in: 
PARKER, S. P. (ed. -in-chief), Synopsis and 
Classification of Living Organisms, vol. 1 . Mc- 
Graw-Hill, New York, xviii + 1166 pp. 

CHAVAN, A., 1969, Superfamily Leptonacea Gray, 
1847. Pp. 518-537, in: MOORE, R.C. (ed.), 
Treatise on Invertebrate Paleontology. Part N. 
h/lollusca 6. Bivalvia 2. Geological Society of 
America (Boulder, Colorado) & University of Kan- 
sas (Lawrence), ii + N491 -N952 pp. 

DALL, W. H., 1899, Synposis of the Recent and 
Teritary Leptonacea of North America and the 
West Indies. Proceedings of the United States 
National Museum, 21 : 873-893. 

GAGE, J., 1968a, The mode of life of Montacuta 
elevata, a bivalve 'commensal' with Clymenella 
torcuata (Polychaeta). Canadian Journal of Zo- 
ology, 46(5): 877-892. 

GAGE, J., 1968b, The mode of life of Mysella cu- 
neata, a bivalve 'commensal' with Phascolion 
strombi (Sipunculoidea). Canadian Journal of 
Zoology, 46(5): 919-934. 

GRAY, J. E., 1840, [Mollusca]. Pp. 86-89, 106- 
156, in: Synopsis of the Contents of the British 
Museum, 42nd ed. British Museum (Natural His- 
tory), London. 

GRAY, J. E., 1842, [Mollusca]. Pp. 48-92, in: Syn- 
opsis of the Contents of the British Museum, 44th 
ed. British Museum (Natural History), London. 

HUMASON, G. L., 1962, Animal tissue tech- 
niques. W.H. Freeman & Co., San Francisco & 
London, XV + 468 pp. 

INTERNATIONAL COMMISSION ON ZOOLOGI- 
CAL NOMENCLATURE, 1985, International 
Code of Zoological Nomenclature [ICZN], 3rd ed. 
International Trust for Zoological Nomenclature, 
London, XX + 338 pp. 

JUDD, W., 1971, The structure and habits of Di- 
variscintilla maoria Powell (Bivalvia: Galeomma- 
tidae). Proceedings of the Malacological Society 
of London, 39: 343-354. 

KARNOVSKY, M. J., 1965, A formaldehyde-glut- 
araldehyde fixative of high osmolarity for use in 
electron microscopy. Journal of Cell Biology, 27: 
137A. 

KISCH, B. S., 1958, Vasconiella jeffreysiana (P. 
Fischer). Proceedings of the Malacological Soci- 
ety of London, 33(1): 21-24, pi. 3. 

MORTON, В., 1972, Some aspects of the func- 
tional morphology and biology of Pseudopythina 



subsinuata (Bivalvia: Leptonacea) commensal 
on stomatopod crustaceans. Journal of Zoology 
(London), 166: 79-96. 

MORTON, В., 1973, The biology and functional 
morphology of Galeomma (Paralepida) takii (Bi- 
valvia: Leptonacea). Journal of Zoology (Lon- 
don), 169(2); 133-150. 

MORTON, В., 1976, Secondary brooding of tem- 
porary dwarf males in Ephippodonta (Ephip- 
podontina) oedipus sp. nov. (Bivalvia: Leptona- 
cea). Journal of Conchology, 29: 31-39. 

MORTON, В., 1980, Some aspects of the biology 
and functional morphology (including the pres- 
ence of a ligamental lithodesma) of Montacutona 
compacta and M. olivácea (Bivalvia: Leptonacea) 
associated with coelenterates in Hong Kong. 
Journal of Zoology (London), 192: 431-455. 

MORTON, В., 1981, The biology and functional 
morphology of Chlamydoconcha orcutti with a 
discussion on the taxonomic status of Chlamy- 
doconchacea (Mollusca: Bivalvia). Journal of Zo- 
ology (London), 195: 81-121. 

MORTON, J. E., 1957, The habits of Scintillona ze- 
landica (Odhner) 1924 (Lamellibranchia: Ga- 
leommatidae). Proceedings of the Malacological 
Society of London, 32: 185-188. 

NARCHI, W., 1966, The functional morphology of 
Ceratobornia cema, new species of the Erycina- 
cea (Mollusca, Eulamellibranchiata). Anais da 
Academia Brasileira de Ciencias, 38(3-4): 513- 
524. 

OCKELMANN, К. W., & К. MUUS, 1978, The biol- 
ogy, ecology and behaviour of the bivalve My- 
sella bidentata (Montagu). Ophelia, 17(1): 1-93. 

Ó FOIGHIL, D., 1985, Form, function, and origin of 
temporary dwarf males in Pseudopythina rugifera 
(Carpenter, 1864) (Bivalvia: Galeommatacea). 
The Veliger, 27(3): 245-252. 

Ó FOIGHIL, D., & A. GIBSON, 1984, The morphol- 
ogy, reproduction and ecology of the commensal 
bivalve Scintillona bellerophon spec. nov. (Ga- 
leommatacea). The Veliger, 27(1): 72-70. 

OLDFIELD, E., 1961 , The functional morphology of 
Kellia suborbicularis (Montagu), Montacuta fer- 
ruginosa (Montagu) and M. substriata (Montagu), 
(Mollusca, Lamellibranchiata). Proceedings of 
the Malacological Society of London, 34: 255- 
295. 

POPHAM, M. L., 1939, On Phlyctaenachlamys 
lysiosquillina gen. and sp. nov., a lamellibranch 
commensal in the burrows of Lysiosquilla macu- 
lata. British Museum (Natural History), Great Bar- 
rier Reef Expedition 1928-29, Scientific Reports, 
6(2): 61-84. 

POPHAM, M. L., 1940, The mantle cavity of some 
of the Erycinidae, Montacutidae and Galeomma- 
tidae with special reference to the ciliary mecha- 
nisms. Journal of the Marine Biological Associa- 
tion of the United Kingdom, 24(2): 549-587. 

POWELL, A. W. В., 1932, On some New Zealand 
pelecypods. Proceedings of the Malacological 
Society of London. 20(pt. 1): 65-72, pi. 6. 

PURCHON, R. D., 1958, The stomach in the Eu- 



BIOLOGY OF DIVARISCINTILLA 



195 



lamellibranchia: stomach Type IV. Journal of Zo- 
ology (London). 131 : 487-525. 

RICHARDSON, K. C, L. JARRETT, & E. H. FINKE, 
1960, Embedding in epoxy resins for ultra-thin 
sectioning in electron microscopy. Stain Technol- 
ogy, 35: 3^3-323. 

STASEK, С R., 1964, Synopsis and discussion of 
the association of ctenidia and labia! palps in the 
bivalved Mollusca. The Veliger, 6: 91-97. 

STIMPSON, W., 1855, On some remarkable ma- 
rine Invertebrata inhabiting the shores of South 
Carolina. Proceedings of the Boston Society of 
Natural History. 5: 110-11 7. 



THIELE, J., 1934, Handbuch der Systematischen 
Weichtierkunde, Teil 3. G. Fischer, Jena, 779- 
1022 pp. 

TRYON, G. W., JR., 1872, Catalogue and synon- 
ymy of the family Galeommidae. Proceedings of 
the Academy of Natural Sciences of Philadel- 
phia, 1872: 222-226. 

VOKES, H. E., 1980, Genera of the Bivalvia: a sys- 
tematic and bibliographic catalogue (revised and 
updated). Paleontological Research Institution, 
Ithaca, New York, xxvii + 307 pp. 

Revised f\/ls. accepted 18 April 1989 



MALACOLOGIA, 1989, 31(1): 197-203 

KARYOTYPIC EVOLUTION IN PLEUROCERID SNAILS. I. GENOMIC DNA 
ESTIMATED BY FLOW CYTOMETRY 

Robert T. Dillon, Jr. 

Department of Biology 

College of Charleston 

Charleston, South Carolina, U.S.A. 29424 

ABSTRACT 

Total genomic DNA was measured in 16 species of North American pleurocerids, representing 
all six living genera. A constant value of 2.1 pg DNA/fiaploid genome was obtained, consistent 
witfi values from othier mesogastropoods and other mollusks with similar chromosome number. 
The relationship between DNA content and evolutionary radiation is called into question. 

Key words: Snails, freshwater, Pleuroceridae, cytogenetics, flow cytometry, DNA. 



INTRODUCTION 

Few groups of North American mollusks 
are as common, diverse, important, or poorly 
understood as the pleurocerid snails. They 
are the most conspicuous element of the 
macrobenthos in many rivers and streams, 
and as such have figured prominently in a 
number of ecological investigations (e.g. El- 
wood & Nelson, 1972; Sumner & Mclntire, 
1982; Hawkins & Furnish, 1987; Dillon & 
Davis, in review). Their occurrence in numer- 
ous, isolated populations, easily sampled 
year round, has made them ideal models for 
evolutionary research (Chambers, 1980, 
1982; Dillon, 1984, 1988a, 1988b). Yet their 
systematics are so confused that the specific 
identity of the populations inhabiting much of 
the United States is problematic. 

The first monographic treatment of the fam- 
ily was given by Tryon (1873). He catalogued 
about 500 nominal species, which he placed 
in nine genera. The currently accepted sys- 
tem of classification is due to a series of pa- 
pers by Goodrich (e.g. 1940, 1942). Goodrich 
recognized somewhat over 100 species and 
subspecies, and his revision formed the basis 
of the classification by Burch & Tottenham 
(1980) and Burch (1982) that I use here. 
Burch recognized six living genera: lo, Juga, 
Leptoxis (including Goodhch's Anculosa, Ni- 
tocris, and Eurycaelon), Lithasia, Pleurocera, 
and Elimia. Burch resurrected "Elimia" on the 
strength of Pilsbry & Rhoads' (1896) type 
designation, failing to note that Pilsbry subse- 
quently reversed himself (Walker, 1918:149). 



Elimia is a composite group, and thus I use 
the much more familiar name favored by 
Tryon and Goodrich, Goniobasis. 

Very little distinguishes many of these gen- 
era. Pleurocera is distinguished from Gon- 
iobasis by a short "canal" on the anterior lip 
of the shell, a feature that is inconspicuous or 
absent in many species. Detailed compari- 
sons of morphology, anatomy, life history, 
and ecology failed to find any other distinc- 
tions (Dazo, 1965). Juga is distinguished from 
Goniobasis by being from western North 
America, not eastern or central. It seems 
clear that the systematic relationships of pleu- 
rocerid snails need to be re-examined. 

Analysis of karyotype has proven to be a 
powerful tool in evolutionary studies (White, 
1973). But very little is known about the cyto- 
genetics of any pleurocerid species. In a re- 
view of molluscan karyotypes, Patterson 
(1969) found two reports for North American 
pleurocerids: n = 18 for Goniobasis laquéala 
and n = 20 for G. livescens. Dillon (1982) 
reported n = 1 7 in G. próxima. The most thor- 
ough study to date has been made by Cham- 
bers (1982), who found n = 18 in Florida Go- 
niobasis and described striking variation in 
arm length ratios. So the evidence available 
suggests that karyotypic variation does occur 
in the Pleuroceridae. In this series of papers, 
I will survey the six genera to see whether 
karyotype may be used to elucidate the sys- 
tematics and evolution of this important family 
of freshwater snails. 

Karyotypes are traditionally compared by 
constructing idiograms for each species, re- 



197 



198 



DILLON 



producing the relative sizes and centromeric 
positions of the chromosomes. Idiograms are 
generally standardized to unit length, so that 
each species is assumed to have the same 
amount of genetic material. An increase in 
chromosome number is viewed as a centric 
fission. This is a reasonable assumption — 
many convincing examples of such "Robert- 
sonian" events are known. But an increase in 
chromosome number could also represent 
additional genetic material. Additional chro- 
mosomes may be incorporated into a genome 
by coincident nondisjunction in the parents, or 
by large-scale gene duplication (Ohno, 1970). 
Thus it is of great value to determine the total 
genomic DNA content of each species to be 
karyotyped. In this first paper of the series, I 
estimate the genomic DNA content of a vari- 
ety of pleurocerid snails using flow cytometry. 

Flow cytometry is one of the most sensitive 
techniques available for quantifying cellular 
DNA. The technique has been used to dis- 
criminate between human chromosomes 
(Gray et al., 1975) and identify structural ab- 
normalities in the chromosome compliments 
of cell lines (review by Arkesteijn et al., 1987). 
The technique has been thoroughly reviewed 
by Melamed et al. (1979), Van Dilla et al. 
(1985), and Shapiro (1988). 

Briefly, tissues in an aqueous suspension 
are stained with a dye that intercalates into 
double-stranded nucleic acid. I used propid- 
ium iodide, after treatment with ribonuclease 
to eliminate double-stranded RNA. Then the 
suspension is channeled at high speed 
through a narrow aperture, using mechanics 
similar to those of the familiar Coulter counter. 
Each individual particle passes through a la- 
ser, which excites a red-fluorescent emission 
proportional to its DNA content. The degree to 
which the laser beam is scattered by each 
particle provides an estimate of the particle's 
size. The emissions of the individual particles 
are captured by photosensors and displayed 
in a scatter plot, which enables the operator to 
distinguish individual, whole cells from debris 
and clumped cells. Then the red fluorescence 
of the whole-cell fraction is plotted in a histo- 
gram, with fluorescence measured in arbi- 
trary units called "channel numbers." Since 
the relationship between channel number and 
DNA content is effectively linear, a flow cy- 
tometer calibrated with known samples can 
be used to estimate the DNA content of an 
unknown. 

The contribution of mitochondrial DNA to 
total red fluorescence has generally been 



found to be negligible (Melamed et al., 1979). 
Correction for any background mtDNA levels 
can be made by using a single tissue type for 
both the unknowns and the calibration stan- 
dards. 



METHODS 

The following populations were sampled: 

Goniobasis acutocarinata (Lea) — Small 
creek flowing into the Powell River at Virginia 
Highway 662 bridge, 0.5 km E of Stickeys, 
Lee County, Virginia. Goodrich (1940) synon- 
ymized this species under G. clavaeformis 
(Lea). 

Goniobasis alabamensis (Lea) — Coosa 
River at tailwater of Mitchell Dam, 20 km E of 
Clanton, Chilton County, Alabama. 

Goniobasis catenaria dislocata (Reeve) — 
"Intermittent" tributary of Big Poplar Creek at 
South Carolina Highway 6 bridge, 3 km SE of 
Elloree, Orangeburg County, South Carolina. 

Goniobasis floridensis (Reeve) — Blue 
Spring at Florida Highway 6, Madison County, 
Florida. Site 8 of Chambers (1980). 

Goniobasis livescens (Menke) — Portage 
Creek at Toma Road bridge, 5 km S of Pinck- 
ney, Washtenaw/Livingston County line, Mi- 
chigan. Station 2 of Dazo (1965). 

Goniobasis próxima (Say) — Mitchell River 
at North Carolina Highway 1330 bridge, 2.8 
km N of Mountain Park, Surrey County, North 
Carolina. Site MTCH of Dillon (1982, 1984). 

Goniobasis simplex (Say) — same site as G. 
acutocarinata. 

lo fluvialis (Say) — Powell River by small 
road just S of Virginia line, Hancock County, 
Tennessee. 

Juga hemphilli (Henderson) — Oak Creek 
11 km W of Corvallis, Benton County, Ore- 
gon. 

Leptoxis (Mudalia) carinata (Brug.) — Pratts 
Run at U.S. Highway 340 bridge, Waynes- 
boro, Augusta County, Virginia. 

Leptoxis praerosa (Say — same site as lo 
fluvialis. 

Lithasia duttoniana (Lea) — Duck River at 
Tennessee Highway 11 bridge, 10 km N of 
Farmington, Marshall County, Tennessee. 

Lithasia verrucosa (Raf.) — French Broad 
River at Cement Shoals, 1 km downstream 
from Kimberlin Heights, Knox County, Ten- 
nessee. 

Pleurocera acuta Raf. — Dazo's (1965) sta- 
tion 2, same as G. livescens. 

Pleurocera canaliculatum (Say) — Elk River 



DNA IN PLEUROCERIDS 



199 







1 cm 



FIG. 1. From left, Lithasia duttoniana, Goniobasis catenaria dislócala, G. alabamensis, G. acutocarinata, 
Juga tiemptiilli. 



at bridge 8 km W of Fayetteville, Lincoln 
County, Tennessee. 

Pleurocera uncíale (Reeve) — same site as 
lo fluvialis. 

The shell morphology of many of these 
species is quite variable, as are the species 
concepts of many prior workers in pleurocerid 
taxonomy. Typical shells from several of the 
less common taxa are shown in Fig. 1. 
Voucher specimens for all populations are de- 
posited in the Academy of Natural Sciences 
of Philadelphia. 

Techniques for sample preparation were 
based on Allen (1983), Buzzi (1989), and 
standard clinical methods. Foot muscle was 
excised from living snails and ground, with 
powdered glass, in a clear polystyrene tube 
with 600 |jlI of phosphate buffered saline. This 
buffer was modified from Allen (1983): NaCI 
8.0 g/l, KCl 0.20 g/l, MgCl2 0.10 g/l, Na2HP04 
1.15 g/l, KH2PO4 0.20 g/l. Drops of cold ab- 
solute ethanol were added, while vortexing, to 
bring the final ethanol concentration up to 
70%. Samples fixed in this manner were held 
at least overnight at 3°C, and sometimes as 
long as two weeks. 

An RNAse solution was prepared by dis- 
solving 50 mg ribonuclease A (Sigma type III- 
A) in 50 ml 1.12% trisodium citrate and heat- 
ing at 80°C for 10 minutes. The solution was 
frozen in 2 ml aliquots. On the morning of flow 
cytometric analysis, fixed tissue samples 
were centrifuged, aspirated, and resuspend- 
ed in a several ml of a solution containing 47 
ml phosphate buffered saline, 1 ml 1 .2% (v/v) 
Nonidet P-40, and one aliquot of RNAse so- 
lution. Since the lot of ribonuclease lll-A that I 
received had 105 units of activity per ml, the 



final RNAse activity in phosphate buffered sa- 
line was approximately 4 units/ml. Tubes 
were incubated at room temperature for 30- 
60 minutes. 

Each sample was vortexed, drawn into a 1 
ml tuberculin syringe, and forced through a 52 
jxm Nytex screen. Then 20 ^JLl of a 0.10% (w/v) 
propidium iodide solution was added per ml of 
tissue suspension, and incubated at room 
temperature for 30-60 minutes prior to anal- 
ysis on an Ortho Spectrum III flow cytometer. 

In collaboration with W. Buzzi, a calibration 
curve was constructed using human leuko- 
cytes and tissue samples from four mollusk 
species of known genomic content. We anal- 
ysed four Crassostrea virginica (Gmelin), five 
Mercenaria mercenaria (L.), and two llyan- 
assa obsoleta (Say), all collected from the 
Charleston, South Carolina, area. We ob- 
tained four Mytilus edulis L. from Milford, Con- 
necticut. The total genomic DNA of С virgin- 
ica is given by Swanson et al. (1981 :134), and 
values for the remaining mollusks are from 
Hinegardner (1974). Three individuals of G. 
catenaria dislocata were included as un- 
knowns. 

Goniobasis catenaria dislocata served as 
the standard in all subsequent analyses. Sev- 
eral fresh G. catenaria were analysed first, 
followed by four to six individuals of a second 
pleurocerid species. The peak red fluores- 
cence for each sample was noted, as well as 
the concentration of countable cells. Aliquots 
from samples of the two species were com- 
bined into a third tube such that cell concen- 
trations were equalized. The combined sam- 
ple was then re-analysed and the resulting 
histogram of red fluorescence inspected for 



200 



DILLON 





JO 












30 


• 






/ 1 


0) 

и 

с 
ф 

«л 
0) 

О 


25 
20 


• 




i/ 


/w 


U- 

> 


15 


- 


У 


л- 

С 






10 
5 


/ 


/ЬА 










Z 









о 



0.5 1.0 1.5 2.0 2.5 3.0 
рд DNA / haploid genome 



3.5 



FIG. 2. Calibration curve. О — Crassostrea, M — Mytilus, С — Mercenaria, G — Goniobasis catenaria dislocata, 
H — Ьиглап, I — liyanassa. 



evidence that the two peaks were non-over- 
lapping. 



RESULTS 

The calibration curve is shown in Fig. 2. An 
excellent fit to the linear hypothesis у = 10. 8x 
- 3.76 was obtained, with r^ = 0.98. So 
given a mean relative fluorescence of 17.3, I 
estimated that G. catenaria dislocata has 2.1 
pg DNA/haploid genome. 

A typical comparison between G. catenaria 
dislocata and an unknown (G. próxima in this 
case) is shown in Fig. 3. This particular sam- 
ple of G. catenaria tissue came from a snail 



collected the previous day, and shows two 
peaks — a strong gap 1 peak and a lower gap 
2/mitosis peak with twice the fluorescence. 
Only snails freshly collected in warm weather 
generally showed a gap 2 peak. Even if tem- 
perature and photoperiod were controlled and 
the snails fed commercial fish food ad lib, gap 
2 peaks generally disappeared after only a 
day or so in captivity, as shown in the G. próx- 
ima sample. In fact, it is evident that DNA 
synthesis has already been discontinued in 
the G. catenaria individual analysed, since no 
S-phase cells, with DNA contents intermedi- 
ate between Gl and G2, are apparent in Fig- 
ure 3. So although the snails in my aquaria 
always appeared healthy, cell division in foot 



DNA IN PLEUROCERIDS 



201 



muscle tissue was apparently disrupted al- 
most immediately. 

The rapid loss of cells at gap 2 and mitosis 
in captive snails did not affect the accuracy of 
sample comparisons. The much stronger, 
sharper gap 1 peaks were used as the basis 
for comparison in all cases. 

No difference was detected between the 
peak red fluorescence of G. catenaria dislo- 
cata and that observed in any other species of 
pleurocerid examined. Figure 3 shows that an 
equal mixture of G. catenaria cells and G. 
próxima cells shows no evidence of two gap 1 
peaks. This result was obtained in all compar- 
isons. 



DISCUSSION 

It would appear that all 16 pleurocerid spe- 
cies in my sample, representing six genera, 
have a uniform genomic DNA content of 2.1 
pg DNA/haploid genome. Hinegardner (1974) 
found that seven species of mesogastropods 
range from 0.67-2.4 pg DNA/haploid ge- 
nome. A vermetid was the only cerithiacean 
examined, with 1.5 pg DNA/haploid genome. 
So the value I have obtained for pleurocehds 
is consistent. 

From a broad comparison of gastropod or- 
ders, Hinegardner suggested that high 
amounts of DNA appear to be associated with 
evolutionary radiation. But in spite of their 
rather average-sized genome, the Pleurocer- 
idae have radiated extensively. Hinegardner's 
generalization may not hold for freshwater 
groups, where dispersal is generally much 
more restricted and the potential for differen- 
tiation greater. 

Across the five kingdoms, there is a general 
relationship between genome size and de- 
gree of organismal complexity or "evolu- 
tionary advancement" (Hinegardner, 1976). 
The "C-value paradox" arose when it was 
noted that some organisms, such as some 
flowering plants and amphibians, have 
amounts of DNA ("C-values") much greater 
than more advanced eukaryotes. But the 
pleurocerid genome size is rather typical for 
mollusks, and for invertebrates in general. 

Hinegardner (1974) reported a correlation 
between chromosome number and DNA con- 
tent in gastropods significant at the 0.01 level. 
Extrapolating from his graph, a chromosome 
number of n = 13 to 18 would be predicted 
from the DNA content of North American 
pleurocehds. This is consistent with the lim- 



c 

о 
и 

и 



1000 


• 




500 


- 


\ 






Vyv_ 


1000 


■ 




500 


■ 






• 










1000 






500 


- 


1 




1 ^-^^-"^--^^ 



о 100 200 

Fluorescence 

FIG. 3. Example comparison of unknown and stan- 
dard. Top — fresh G. catenaria dislocata standard, 
showing gap 1 and gap 2 peaks. Middle — the un- 
known (G. próxima), showing gap 1 peak only. Bot- 
tom — equal mixture of standard and unknown, 
demonstrating complete overlap of gap 1 peaks. 



ited information available on pleurocerid kary- 
otypes. Ongoing studies will more thoroughly 
address the degree to which uniformity in ge- 
nomic DNA content reflects karyotypic con- 
servation in this family. Any variation in chro- 
mosome number among pleurocehds can be 



202 



DILLON 



viewed with some confidence as originating in 
Robertsonian fusion or fission. 



ACKNOWLEDGMENTS 

I thank Dr. Fred Thompson, Dr. Gary Lam- 
berti, and Randy Wildman for providing spec- 
imens; Steve Ahlstedt for locality data; and 
John Wise and Robert T. Dillon, Sr., for help 
with the collecting. Dr. Mariano LaVia was a 
gracious host at the Medical University of 
South Carolina's flow cytometry unit, and Jo 
Ann Koffskey provided expert technical assis- 
tance. Special appreciation is due to Bill 
Buzzi, who worked out most of these tech- 
niques and shared them with me. 



LITERATURE CITED 

ALLEN, S. K., 1983, Flow cytometry: assaying ex- 
perimental polyploid fish and shellfish. Aquacul- 
ture, 33:317-328. 

ARKESTEIJN, G. J. A., A. С M. MARTENS, R. R. 
JONKER, A. HAGEMEIJER, & A. HAGENBEEK, 
1987, Bivariate flow karyotyping of acute myelo- 
cytic leukemia in the BNML rat model. Cytometry, 
8:618-624. 

BURCH, J. В., 1982, North American freshwater 
snails: identification keys, generic synonymy, 
supplemental notes, glossary, references, index. 
Walkerana, 4:1-365. 

BURCH, J. В., & J. L. TOTTENHAM, 1980, North 
American freshwater snails: species list, ranges, 
and illustrations. Walkerana, 3:1-215. 

BUZZI, W., 1 989, Growth and survival of larval and 
juvenile polyploid clams, Mercenaria mercenaria. 
MS thesis. College of Charleston, SC. 

CHAMBERS, S. M., 1980, Genetic divergence be- 
tween populations of Goniobasis occupying dif- 
ferent drainage systems. Malacologia, 20: 63- 
81. 

CHAMBERS, S. M., 1982, Chromosonal evidence 
for parallel evolution of shell sculpture pattern in 
Goniobasis. Evolution, 36:113-120. 

DAZO, B. C, 1965, The morphology and natural 
history of Pleurocera acuta and Goniobasis 
livescens (Gastropoda: Cerithiacea: Pleurocer- 
idae). Malacologia, 3:1-80. 

DILLON, R. T, Jr., 1982, The correlates of diver- 
gence in isolated populations of the freshwater 
snail, Goniobasis próxima. Ph.D. dissertation. 
University of Pennsylvania, Philadelphia. 183 pp. 

DILLON, R. T, Jr., 1984, Geographic distance, en- 
vironmental difference, and divergence between 
isolated populations. Systematic Zoology, 33: 
69-82. 

DILLON, R. T, Jr., 1988a, The influence of minor 
human disturbance on biochemical variation in a 



population of freshwater snails. Biological Con- 
servation, 43:137-144. 

DILLON, R. T, Jr., 1988b, Evolution from trans- 
plants between genetically distinct populations of 
freshwater snails. Genética, 76:111-119. 

DILLON, R. T, Jr., & K. B. DAVIS, In review. The 
diatoms ingested by freshwater snails: temporal, 
spatial, and interspecific variation. 

ELWOOD, J. W., & D. J. NELSON, 1972, Periphy- 
ton production and grazing rates in a stream 
measured with a ^^P material balance method. 
Oikos, 23:295-303. 

GOODRICH, C, 1940, The Pleuroceridae of the 
Ohio River system. Occasional Papers of the Mu- 
seum of Zoology, University of Michigan, no. 417: 
21 pp. 

GOODRICH, C, 1942, The Pleuroceridae of the 
Atlantic coastal plain. Occasional Papers of the 
Museum of Zoology, University of Michigan, no. 
456:6 pp. 

GRAY, J. W., A. V. CARRANO, L. L. STEINMETZ, 
M. A. VANDILLA, D. H. MOORE, B. H. MAYALL, 
& M. L. MENDELSOHN, 1975, Chromosome 
measurement and sorting by flow systems. Pro- 
ceedings of the National Academy of Science, 
72:1231-1234. 

HAWKINS, С P., & J. K. FURNISH, 1987, Are 
snails important competitors in stream ecosys- 
tems? Oikos, 49:209-220. 

HINEGARDNER, R., 1 974, Cellular DNA content of 
the Mollusca. Comparative Biochemistry and 
Physiology, 47A:447-460. 

HINEGARDNER, R., 1976, Evolution of genome 
size. In Molecular evolution, AYALA, F. J., ed. 
Sinauer Associates, Sunderland, Mass, pp. 179- 
199. 

MELAMED, M. R., P. F. MULLANEY, & M. L. MEN- 
DELSOHN, eds., 1979, Flow cytometry and sort- 
ing. John Wiley & Sons, N.Y. 716 pp. 

OHNO, S., 1970, Evolution by gene duplication. 
Springer- Verlag, NY. 160 pp. 

PATTERSON, С M., 1969, Chromosomes of mol- 
luscs. In Proceedings of the Symposium on Mol- 
lusca, Part II. Marine Biological Association of 
India. Bangalore Press, Bangalore, India, pp. 
635-686. 

PILSBRY, H. A., & S. N. RHOADS, 1896, Contri- 
butions to the Zoology of Tennessee, No. 4, Mol- 
lusca. Proceedings of the Academy of Natural 
Sciences of Philadelphia. 1896:487-506. 

SHAPIRO, H. M., 1988, Practical flow cytometry, 
2nd ed. A. R. Liss, NY. 353 pp. 

SUMNER, W. T, & С D. MCINTIRE, 1982, Grazer- 
periphyton interactions in laboratory streams. Ar- 
chiv fur Hydrobiologie, 93:135-157. 

SWANSON, С P., T MEVE, & W. J. YOUNG, 
1 981 , Cytogenetics: the chromosome in division, 
inheritance, and evolution, 2nd ed. Prentice Hall, 
Englewood Cliffs, NJ. 

TRYON, G. W., Jr., 1873, Land and freshwater 
shells of North America. Part IV. Strepomatidae. 
Smithsonian Miscellaneous Collections, no. 253: 
435 pp. 



DNA IN PLEUROCERIDS 



203 



VAN DILLA, M. A., P. N. DEAN, O. D. LAERUM, & 
M. R. MELAMED, eds., 1985, Flow cytometry: 
instrumentation and data analysis. Academic 
Press, Orlando, PL. 288 pp. 

WALKER, В., 1 91 8, A synopsis of the classification 
of the freshwater Mollusca of North America, 
north of Mexico. University of Michigan Museum 



of Zoology Miscellaneous Publications No 6213 
PP- 
WHITE, M. J. D., 1973, Animal cytology and evo- 
lution, 3rd ed. Cambridge University Press, Lon- 
don. 

Revised Ms. accepted 12 March 1989 



MALACOLOGIA, 1989, 31(1): 205-210 



HABITAT SELECTION BY A FRESHWATER MUSSEL: 
AN EXPERIMENTAL TEST 

Robert C. Bailey 

Ecology and Evolution Group 

Department of Zoology 

University of Western Ontario 

London, Ontario, Canada 

N6A 5B7 



ABSTRACT 

Two groups of the freshwater mussel Lampsilis radiata siliquoidea (Barnes, 1823) were col- 
lected in Inner Long Point Bay, Lake Erie. The first group of nnussels was collected from sandy, 
turbulent areas of the bay, while the second group was collected from soft-bottomed, muddy 
areas. The sand-collected mussels were larger and thicker-shelled than the mud-collected 
group, which is consistent with previously observed correlations between shell form and habitat 
in this and other Unionidae species. I placed 50 individuals from each of these two groups into 
each of two artificial ponds. Each pond contained equal areas of sand and mud in "checker- 
board" fashion, and each mussel was placed at random coordinates on the bottom of the ponds. 
After four months, two-thirds of the mussels were found in the mud sediment. About 80% of the 
mussels initially placed in mud stayed there, while about half of the mussels initially placed in 
sand moved to mud. Sand-collected mussels had a stronger tendency, relative to the mud- 
collected group, to either stay in mud if they started there or move to mud from sand. The results 
support the hypothesis that habitat selection has evolved in this unionid species, but are not 
consistent with the hypothesis that the two groups of mussels represent specialists for the 
habitats from which they were collected. 

Key words: habitat selection; Unionidae; shell morphology; specialists; sediment preference. 



INTRODUCTION 

As discussed by both Kat (1 982) and Hueh- 
ner (1987), there has been more observa- 
tional and anecdotal evidence than detailed, 
experimental study of habitat selection in 
freshwater mussels. This has lead to a re- 
markable lack of understanding of their ability 
(or lack of ability) to select habitat. Clearly this 
knowledge is important in understanding the 
relative niche breadths of each species, as 
well as the degree of niche overlap among 
species. 

Short-term (three-hour) experiments in the 
laboratory by Huehner (1987) indicated that 
most populations of Anodonta grandis Say, 
1829, and Lampsilis radiata (Gmelin, 1791) 
show a preference for sand over gravel. The 
other species tested, Elliptio dilatata (Ra- 
finesque, 1820), showed no substrate prefer- 
ence. Huehner (1987) commented on the be- 
havioural and morphological plasticity of 
Lampsilis radiata. In his laboratory experi- 
ments, one population of L. radiata sili- 



quoidea showed a preference for sand, while 
another had no preference. In this study, I 
tested the ability of two Lampsilis radiata sil- 
iquoidea forms to select substrate over a rel- 
atively long (four-month) experimental period. 
Lampsilis radiata siliquoidea (Barnes, 
1823) is one of many freshwater mussels 
whose shell morphology and growth rate vary 
with habitat. Bailey & Green (1988) and Hinch 
et al. (1986) found thicker-shelled, faster- 
growing L. r. siliquoidea in the turbulent, 
sandy sediment areas of Inner Long Point 
Bay, Lake Erie, when compared to conspe- 
cific mussels from the more quiescent, muddy 
areas of the bay. Many authors have claimed 
that correlations between the habitat of fresh- 
water mussels and their shell form are due to 
differential adaptation resulting in specialist 
phenotypes. Wilson & Clark (1914) sug- 
gested that larger, flatter shell forms are bet- 
ter adapted to burrowing in the coarse sub- 
strates of fast current areas in streams, while 
smaller, more obese (large width-to-length ra- 
tio) shells maintain a mussel's buoyancy In 



205 



206 



BAILEY 



100 



ТНЯВГ9Ч 




О l-l-l 



0.00 



0.01 0.1 1 

Grain size (mm) 

FIG. 1 . Particle size distribution of substrates used in the pond experiment ( SAND, MUD). 



soft substrates. Eagar (1978) clainned that 
more obese shells allow for a greater volume 
of soft tissue, thereby improving the "meta- 
bolic and functional activity" of mussels in qui- 
eter waters. Stanley (1970) considered the 
functional morphology of the entire Class Bi- 
valvia and drew conclusions similar to those 
of Wilson & Clark (1914). Although these hy- 
potheses seem reasonable, clear tests of 
their predictions have not been made. 

In the present study, L r. siliquoidea from 
both turbulent and quiescent areas of Inner 
Long Point Bay, Lake Erie, were used in a 
substrate selection experiment carried out in 
artificial ponds. Mussels from the two areas 
differed morphologically in a manner more or 
less consistent with the adaptive hypotheses 
outlined above. I predicted that if these two 
phenotypes represented specialists for differ- 
ent habitats, the availability of both fine- and 
coarse-grained sediments in artificial ponds 
would lead to differential habitat selection by 
the two groups of mussels. 



MATERIALS AND METHODS 

On 11 June 1985, 100 similarly aged (8- 
1 2-year-old) L. r. siliquoidea were collected 
using SCUBA from each of low and high ex- 
posure areas in Inner Long Point Bay, Lake 



Erie. Inner Long Point Bay is a large (75 km^), 
shallow (z = 2.5 m) bay with a heteroge- 
neous distribution of high, medium, and low 
exposure areas grading into one another (see 
Bailey, 1988, for a map of exposure areas). 
The mussels were collected from well within 
one high and one low exposure area in the 
bay, in each case within 50 m of the boat. 
These mussels were transplanted into two ar- 
tificial ponds on the campus of The University 
of Western Ontario. Each pond measured 5 
X 9 m and had a depth of about one meter. 
One week before collecting the mussels, 
equal areas of "sand" and "mud" sediment, 
obtained from Southwinds Sand and Gravel 
(London, Ontario), were spread to a depth of 
about 1 5 cm prior to filling the ponds with city 
water using taps located at the side of each 
pond. The sediment was added in checker- 
board fashion such that there were two rect- 
angular areas of each sediment in each of the 
ponds. The sand used was "golf course 
sand"; the mud was from silt deposits created 
by the wastewater from washing crushed 
gravel. Percent loss on ignition, determined 
as described in Bailey (1988), was nil for both 
substrates. Particle size analysis, using wet 
sieving and hygrometer analysis (Bowles, 
1978), indicate that the mud was finer and 
more heterogeneous than the sand (Fig. 1). 
A number was etched onto each mussel's 



MUSSEL HABITAT SELECTION 207 

TABLE 1. Loglinear model analysis showing the significance of "SOURCE" (area where mussel was 
collected in Inner Long Point Bay), "INITIAL SUBSTRATE" (substrate in which the mussel was initially 
planted in the experimental ponds), and the interaction of the two effects in predicting the final substrate 
of the mussels. 



Source 



df 



SOURCE 

INITIAL SUBSTRATE 

INTERACTION 



2.9 
17.4 

0.13 



0.08 

<0.001 

0.72 



shell and 50 individuals from each exposure 
area were placed at randonnly generated co- 
ordinates on the bottom of each of the ponds 
on 12 June 1985. By 18 June, there was am- 
ple evidence of mussel movement within the 
ponds (i.e. tracks). Because of concern about 
the use of city water, various physico-chem- 
ical and biological parameters were moni- 
tored. Chlorine (Hach Model CN-70) and dis- 
solved oxygen (Hach Model AI-33) analyses 
were carried out throughout the experimental 
period and showed total chlorine concentra- 
tions of 0.2-0.3 mg • L"^ (tap water in London 
was about 0.6-0.8 mg • L^) and dissolved 
oxygen concentrations of 90-100% satura- 
tion. Temperature over the experimental pe- 
riod ranged from 16-29°C. Flow rate from the 
taps into the ponds was checked daily and 
kept at 125 mLsec \ Qualitative sampling 
with a plankton net on 2 July 1 985 revealed 
abundant insect, crustacean Zooplankton, 
Hydracarine, and algal populations in both of 
the ponds. Four months after planting the 
mussels in the ponds (9 October 1985), 151 
individuals (142 alive) were recovered over a 
two-day period (using SCUBA) and the ponds 
were drained. Eight additional (dead) mussels 
were found the following day, and 23 dead 
mussels were recovered from the dry ponds 
the following spring. Because the live and 
dead mussels did not differ in their distribution 
patterns, data on all recovered individuals 
were used in the statistical analyses. Shells of 
the recovered mussels were cleaned, dried, 
weighed, and measured (length, height, and 
width as defined in Bailey & Green 1 988), and 
the morphological differences between those 
collected in the sand and the mud were con- 
sistent with differences observed by Bailey & 
Green (1988). 

The habitat from which each mussel was 
originally collected in the field ("SOURCE"), 
the substrate in which the mussel was initially 
placed in one of the ponds ("INITIAL 
SUBSTRATE"), and the interaction of these 
two factors were tested as predictors of the 



mussel's final substrate "choice" using a log- 
linear model (Fienberg, 1980). SAS Proc Cat- 
mod (SAS Institute Inc., 1982) was used for 
the analysis. 



RESULTS 

Two-thirds of the mussels recovered (124/ 
182) at the end of the experiment were found 
in the mud substrate. The loglinear model 
analysis (Table 1) showed that this prefer- 
ence for mud was somewhat influenced (p = 
0.08) by the "SOURCE" of the mussels and 
more strongly affected by their "INITIAL 
SUBSTRATE." There was no interaction be- 
tween these two effects. Compared to mus- 
sels collected in the muddy areas of Inner 
Long Point Bay, more of the mussels col- 
lected in exposed, sandier areas of the bay 
tended to either stay in mud if they started 
there, or move to mud from sand (Fig. 2). In 
both groups, there was a tendency for those 
initially placed in mud to stay there, but those 
initially placed in sand had about a 50/50 
chance of switching to mud (Fig. 2). 



DISCUSSION 

Although both the sand- and mud-collected 
mussels appeared to select habitat, the two 
forms differed only in the magnitude (rather 
than the nature) of their habitat selection. A 
greater proportion of the sand-collected mus- 
sels starting in mud stayed in mud, and a 
greater proportion of sand-collected mussels 
moved to mud from sand, but both groups 
showed similar basic patterns of habitat 
choice (Fig. 2). There are at least two expla- 
nations for this: (i) the two phenotypes do not 
represent specialists for different habitats, 
and (ii) the two substrate types used in the 
experiment did not adequately recreate the 
habitat choices available to these mussels in 
their natural environment. 



208 



BAILEY 



■ Sand-collected 
DMud-collected 




Group 

FIG. 2. Initial placement and final substrate of recovered mussels. The percentage of the total for a given 
SOURCE group (i.e. sand- or mud-collected) is given (MM: initial = mud, final = mud; MS: initial = mud, 
final = sand; SS: initial = sand, final = sand; SM: initial = sand, final = mud). 



It has previously been proposed that the 
smaller, more obese shells of freshwater 
mussels living In soft mud habitats simply re- 
flect a non-adaptive response to poorer grow- 
ing conditions. Food supply may be reduced 
in these areas (Ball, 1922; Stansbery, 1970; 
Kat, 1982), but this hypothesis has never 
been tested. Feeding behavior may also differ 
in soft sediment habitats. Ellis (1936) ob- 
served that mussels in muddy water had their 
valves closed 75-90% of the time, while 
those in silt-free water were closed less than 
50% of the time. He also found that heavy 
silting killed most of the mussels kept in ex- 
perimental tanks. Kat (1982) argued that the 
net intake of energy would be reduced on 
muddy substrates because the mussels 
would require more energy to maintain proper 
filtering position. As in the case of the 
"adaptive hypotheses" (see Introduction), lit- 
tle direct evidence has been collected to re- 
ject either the "adaptive" or "environmental" 
hypotheses of variation in shell morphology. 

Thus, the difference in shell morphology 
between sand- and mud-collected mussels 
may indicate different growing conditions 
rather than differential adaptation to their re- 
spective habitats. If this were true, the two 
phenotypes would not represent specialists 
for the two habitat areas in Inner Long Point 



Bay, and no difference in habitat choice would 
be expected. Sand-collected mussels may 
have exhibited a greater degree of pickiness 
because of size-dependent controls on the 
proximal mechanism of habitat selection in 
these mussels. Perhaps differences in short- 
term fitness of mussels in the two sediments 
(e.g. filtering efficiency, maintenance of shell 
position), which would provide the necessary 
cues for stimulating habitat selection, were 
not as great for the mud-collected mussels. 

The substrate choices available in the pond 
experiment may not have adequately repre- 
sented habitat variation in the natural environ- 
ment. The most obvious evidence supporting 
this contention is the clear choice of the 
"mud" sediment in the ponds by mussels col- 
lected from the sandy area of Inner Long 
Point Bay. The particle size distribution of 
"typical" sediment samples from muddy and 
sandy areas in Inner Long Point Bay (Fig. 3) 
are clearly more similar to the "mud" than the 
"sand" sediment in the ponds (Fig. 1), al- 
though the "mud" sediment in the ponds was 
somewhat more heterogeneous than the nat- 
ural sediments. Also, there were many differ- 
ences between the sandy and muddy areas in 
the bay that were not recreated in the exper- 
iment, such as macrophyte and fingernail 
clam communities (Bailey, 1988), organic 



MUSSEL HABITAT SELECTION 



209 



100 




0.00 



0.01 0. 

Grain Size (mm) 



FIG. 3. Particle size distribution of "typical" substrate collected from sandy (- 
of inner Long Point Bay, Lake Erie. 



-) and muddy ( ) areas 



content (Bailey, 1988) and penetrability (Bai- 
ley, personal observation) of the sediment, 
and the actual turbulence that created and 
maintains the sediment variation in the bay. 
None of these correlated environmental dif- 
ferences were present in the pond experi- 
ment, and thus weakened its relevance to the 
natural environment. On the other hand, the 
grain size stimulus for habitat selection in L. r. 
siliquoidea must have been quite strong to 
have generated the observed results. 

Even though both groups of mussels ap- 
peared to select the mud sediment in the 
ponds, the nature of this and similarly de- 
signed substrate selection experiments (e.g. 
Meier-Brook, 1969; Gale, 1971; Huehner, 
1987) allows for another interpretation. The 
relatively slow-moving bivalves must move 
through the two substrates available. If one 
of the substrates is considerably harder to 
move through than the other, the mussels will 
accumulate in that substrate and appear to 
have "chosen" it at the end of the experiment. 
This possibility, which may be likened to a 
food choice experiment in which one of the 
diet alternatives makes it physically impossi- 
ble for the animal to eat anything else, might 
be called the "stuck in the mud" hypothesis 
(R. H. Green & S. G. Hinch, personal com- 
munication). Although regular observations of 



the ponds revealed numerous tracks through 
both sediment types, detecting any difficulty 
in movement was beyond the scope of this 
study. 

If one does accept that habitat selection 
was demonstrated tDy L. r. siliquoidea, how 
relevant Is this to the behavior of the mussel 
in its natural habitat? Many authors have 
found that juvenile mussels, after finishing a 
life stage during which they are parasitic on 
fish, occupy a habitat somewhat different 
from adults of the same species (e.g. Lefevre 
& Curtis, 1912; Isely, 1911; Coker et al., 
1 921 ). Perhaps at some time between the ju- 
venile dropping from the fish host and the rel- 
atively sedentary adult stage (Strayer, 1981; 
but cf. Salmon & Green, 1983), selection of 
an appropriate adult habitat should occur. 
Whether or not habitat selection would evolve 
would depend on how much would be gained 
by selecting habitat (i.e. benefits of habitat se- 
lection) relative to the time and energy spent 
searching for the habitat (i.e. costs of hab- 
itat selection). The hypothesis that the ability 
to select habitat has evolved in L. r. sili- 
quoidea seems credible. This experiment has 
shown (with the aforementioned reservations) 
that the ability to select habitat exists In these 
mussels. This evidence strengthens conclu- 
sions from observational, frequency of occur- 



210 



BAILEY 



rence data and short-term laboratory experi- 
ments (e.g. Huehner, 1987). There is no 
evidence, however, that the sand- and mud- 
collected mussels from Inner Long Point Bay 
specialize on different habitat types. Either 
the difference in shell phenotype between the 
groups is a non-adaptive, environmentally in- 
duced effect or the habitats available in the 
pond experiment were not suitable for detect- 
ing a difference in preference. 



ACKNOWLEDGEMENTS 

To those who toiled in the ponds (Cindy 
Walker, Scott Hinch, Helene Dupuis, Karen 
Watkinson), I give thanks. The Canada Cen- 
tre for Inland Waters (Burlington) loaned me a 
boat for field work in Inner Long Point Bay. 
The Ontario Ministry of Natural Resources 
loaned some docking space. Miles Keenley- 
side facilitated use of the ponds. R. H. Green, 
P. Handford, T. M. Laverty, and D. Strayer 
read the manuscript and improved it with their 
comments. This project was funded by 
NSERC Operating and Ontario Ministry of the 
Environment grants to R. H. Green, and an 
NSERC Postgraduate Scholarship to RGB. 



LITERATURE CITED 

BAILEY, R. C, 1988, Correlations between species 
richness and exposure: freshwater molluscs and 
macrophytes. Hydrobiologia, 162: 183-191. 

BAILEY, R. С & R. H. GREEN 1988, Within-basin 
variation in the shell morphology and growth rate 
of a freshwater mussel. Canadian Journal of Zo- 
ology, 66: 1704-1708. 

BALL, G. H., 1922, Variation in freshwater mussels. 
Ecology, 3: 93-121. 

BOWLES, J. E., 1978, Engineering properties of 
soils and their measurement, 2nd edition. Mc- 
Graw-Hill. 

COKER, R. E., A. F. SHIRA, H. W. CLARK & A. D. 
HOWARD, 1921, Natural history and propaga- 
tion of freshwater mussels. Bulletin of the U.S. 
Bureau of Fisheries, 37: 77-181. 

EAGAR, R. M. C, 1978, Shape and function of the 
shell: a comparison of some living and fossil bi- 



valve molluscs. Biological Reviews, 53: 169- 
210. 

ELLIS, M. M., 1936, Erosion silt as a factor in 
aquatic environments. Ecology, 17: 29-42. 

FIENBERG, S. E., 1980, The analysis of cross- 
classified categorical data. MIT Press. Cam- 
bridge, Mass. 

GALE, W. F., 1971, An experiment to determine 
substrate preference of the fingernail clam, 
Sphaehum transversum (Say). Ecology, 52: 
367-370. 

HINCH, S. G., R. С BAILEY & R. H. GREEN, 1986, 
Growth of Lampsilis radiata (Bivalvia, Unionidae) 
in sand and mud: a reciprocal transplant experi- 
ment. Canadian Journal of Fisheries and Aquatic 
Science, 43: 548-552. 

HUEHNER, M. K., 1987, Field and laboratory de- 
termination of substrate preferences of unionid 
mussels. Ohio Journal of Science, 87: 29-32. 

ISELY, F. В., 191 1 , Preliminary note on the ecology 
of the early juvenile life of the Unionidae. Biolog- 
ical Bulletin, 20: 77-80. 

KAT, P. W., 1982, Effects of population density and 
substratum type on growth and migration of El- 
liptic complánate (Bivalvia, Unionidae). ¡\Aalaco- 
logical Reviews, 15: 119-127. 

LEFEVRE, G. & W. С CURTIS, 1912, Studies on 
the reproduction and artificial propagation of 
freshwater mussels. Bulletin of the U.S. Bureau 
of Fisheries, 30: 105-201. 

MEIER-BROOK, C, 1969, Substrate relations in 
some Pisidium species (Eulamellibranchiata: 
Sphaeriidae). Malacologia, 9: 121-125. 

SALMON, A. & R. H. GREEN, 1983, Environmental 
determinants of unionid clam distribution in the 
Middle Thames River, Ontario. Canadian Journal 
of Zoology, 61: 832-838. 

SAS INSTITUTE INC, 1982, SAS user's guide: sta- 
tistics. SAS Institute Inc. Gary, N.C. 

STANLEY, S. M., 1970, Relation of shell form to life 
habits in the Bivalvia. Geological Society of 
America Memoirs, 125: 1-296. 

STANSBERY, D. H., 1970, A study of the growth 
rate and longevity of the naiad Amblema plicata 
(Say 1817) in Lake Erie. Bulletin of the American 
t^alacological Union, 37: 78-79. 

STRAYER, D. L., 1981, Notes on the microhabitats 
of unionid mussels in some Michigan streams. 
American l\/lidland Naturalist, 106: 411-415. 

WILSON, С В. & H. W. CLARK, 1914, The mussels 
of the Cumberland River and its tributaries. U.S. 
Bureau of Fisheries Report No. 781 . 

Revised Ms. accepted 27 March 1989 



MALACOLOGIA, 1989, 31(1): 211-216 

SPERMATOCYTE CHROMOSOMES AND NUCLEOLUS ORGANIZER REGIONS 

(NORs) IN TRICOLIA SPECIOSA (MÜHLFELD, 1824) (PROSOBRANCHIA, 

ARCHAEOGASTROPODA) 

R. Vitturi & E. Catalane 
Institute of Zoology, University of Palermo, Via Arch i raf i 18-90123 Palermo, Italy 

ABSTRACT 

The chromosome complement, n = 8 and 2n = 16, of Tricolia speciosa is at present the 
lowest chromosome number found within the Archaeogastropoda (Mollusca: Prosobranchia). 
The karyotype consists entirely of bi-armed chromosomes. No heterotypic elements were ob- 
served in analyses of meiotic and mitotic chromosomes. An analysis of the nucleolar organizer 
region (NOR) by silver staining is reported. Tricolia speciosa presents an intraspecific variability 
in Ag-NOR pattern as revealed by differences in the number of Ag-NORs per cell within a cell 
population. 

Key words: Tricolia; karyology; nucleous organizer regions. 



INTRODUCTION 

Three thousand living species distributed in 
22 families are currently recognized by Franc 
(1968) within the prosobranch order Archae- 
ogastropoda. Because karyological informa- 
tion is only available for 76 species from nine 
families (Vitturi et al., 1982; Nakamura, 1982, 
1 983, 1 986), it is clear that many archaeogas- 
tropod species and families remain com- 
pletely unexplored. 

Previous studies on mitotic chromosomes 
morphology in 46 of the 76 examined species 
(Nakamura, 1986) revealed that 10-20 per- 
cent of the chromosome complements of ar- 
chaeogastropods consist of sub-telocenthc 
(ST) and acrocentric (A) chromosomes, with 
higher values of metacentric (M) and sub-me- 
tacentric (SM) elements in the karyotypes of 
those species characterized by a low number 
of chromosomes, such as Patellidae and Ac- 
maeidae. Moreover, within this order, the 
haploid chromosome number varied from n = 
9 (Patellidae) to n = 21 (Trochidae), with in- 
termediate values as briefly summarized in 
Table 1. Nakamura (1986) noted, however, 
that chromosome numbers were quite con- 
stant within each family, except for the Hali- 
otidae and Fissurellidae, in which there was 
some variation. 

The location of nucleolus organizer regions 
has been reported mainly for mammalian spe- 
cies (Goodpasture & Bloom, 1975; Pardue & 
Hsu, 1975: Markovic et al., 1978; Traut et al.. 



1984), and for a relatively few species of fish 
(Kligerman & Bloom, 1 977; Foresti et al., 1 981 ; 
Thode et al., 1983, 1985; Thode, 1987). 

With regard to Mollusca, results with silver 
staining have been described for the genera 
Bulinus and Biomphalaria (Mollusca, Planor- 
bidae) (Goldman et al., 1983). 

In the present paper, we describe sperma- 
tocyte chromosome of the species Tricolia 
speciosa, which belongs to the family Pha- 
sianellidae (Archaeogastropoda) previously 
unexplored at a karyological level. Addition- 
ally, we report here our findings concerning 
the distribution and behaviour of nucleolar or- 
ganizer regions (NORs) in this species. 



MATERIALS AND METHODS 

Thirty sexually mature male specimens of 
Tricolia speciosa collected in February 1987 
in the Gulf of Palermo were employed. Taxo- 
nomic identification of the specimens was 
made according to the guidelines of Parenzan 
(1970), and voucher shells of ten specimens 
were deposited at the Museum of the Institute 
of Zoology of the University of Palermo. 

Meiotic chromosomes were obtained by 
treating testes according to the squashing 
technique described for other molluscan spe- 
cies (Vitturi et al., 1982). In order to obtain 
mitotic chromosomes, testes of ten speci- 
mens were treated before squashing with 



211 



212 VITTURI & CATALANO 

TABLE 1. Haploid chromosome numbers in Archaeogastropoda; (1) including one species reported to 
have various chromosome numbers 

Total 

number of species examined 

Family with n = 9 10 11 12 13 14 15 16 17 18 19 20 21 species 

Acmaeidae — 14 — — — — — — — — — — — 14 

Patellidae 5____________ 5 

Neritidae*'> 1 1122— 1— — — — — — — 23 

Haliotidae _____2 — 2— 4 — — — 8 

Fissurellidae ____i 1 1 з_____ ß 

Trochidae*^' _________ 13 _1 1 14 

Turbinidae — — — — — — — — — 3 — — — 3 

Stomatellidae — — — — — — — — — 1 — — — 1 

Helicinidae — — — — — — — — — 3 — — — 3 

Total 6 15 1 22 1 4 1 5 24 1 1 76 



0.025% colchicine in double distilled water for 
20 minutes. 

The same slides, after removal of the 
cover-glass, were then stained with silver ni- 
trate following the procedure of Howell & 
Black (1980). 

Acetic-orcein slides were photographed 
with a Wild phase contrast microscope, and 
NOR-banded slides with a Wild light micro- 
scope. 

Mitotic chromosomes were interpreted on 
the basis of the arm ratio, following the no- 
menclature proposed by Levan et al. (1964) 



OBSERVATIONS 
Acetic-orcein slides 

At the pachytene stage, all bivalents were 
tightly paired and their outlines were irregular 
(Fig. 1). 

The analysis of 64 diakinetic plates gave 
the haploid number of 8 chromosomes (Fig. 
2a). When disparate chromosome counts oc- 
curred (one plate with 6 chromosomes, three 
plates with 7, and four plates with 9 chromo- 
somes), the discrepancy was usually attrib- 
uted to either loss or breakage of bivalents. 
Almost all bivalents homogeneously stained 
appeared chiasmatic (Fig. 2b), and their 
lengths ranged from 2 |xm to 3 |xm. 

At the spermatogonial metaphase stage 
(Fig. 3a), all 16 elements showed no achro- 
matic area, with the exception of a pale me- 
dial zone corresponding to the centromere re- 
gion, and thus appeared randomly distributed 
on the squashing plane. From an analysis of 
the idiogram (Fig. 3b, one plate is repre- 
sented) combined from the cfiromosomes of 



five metaphase plates and arranged on the 
basis of their decreasing size and centromere 
position (Fig. 4, Table 2), it appears that all 
pairs were metacentric except for one (Figs. 
3b, 4, arrows) that was sub-metacentric. 

NOR-banding slides 

Analysis of nuclei stained by the silver 
method revealed a variability in the number of 
nucleoli/nucleus, and the frequencies appear 
in Table 3. 

In Figure 5, the two areas showing an in- 
tense silver deposit were of larger dimensions 
than the six areas observed in Figure 7. 

In Figure 6, a nucleus with three nucleoli is 
visible. 

A summary of the state at diakinesis is as 
follows: 60% of analysed spreads show a 
completely NOR negative appearance (Fig. 
8), 30% have 2-3 elements with NORs rep- 
resented by minute dots (Fig. 9, arrows), and 
10% have almost all elements with Ag gran- 
ules (Fig. 10, arrows). 

Mitotic chromosomes at the prophase 
stage often show NOR positive areas not as- 
sociated with the chromosomes (Fig. 11, 
arrow). In the same figure, the element indi- 
cated by two arrows has a large telomeric 
NOR-band. Variability in the number of NOR 
positive elements and the size of Ag-NORs 
was also observed. 

At the metaphase stage, spreads with ei- 
ther three Ag-NOR chromosomes (Fig. 1 2, ar- 
rows) or with five or six NOR-elements were 
present (Fig. 13, arrows). 

DISCUSSION 

From our observations, it seems that the 
course of spermatogenesis in Thcolia spe- 



CHROMOSOMES IN TRICOLIA 



213 



'^щщ.т-шшщ 



5 ■* Vè^^^JÊ^ 



¿^ 



«^ О 

. 10 д|т 



2й ^^> 



10 >im 



» 3b 



-f Ш*- ^ ВС :> st II It i% il •« 



X 






4P •* 









t I t 



У 



5 -um 



ш Щ 






10 Д1т 



,_JOjjm_r|| 
I И| 



10 iJm 



1*^' 












10 Aim 



t 



m * ччт 

>*• 10 Aim , 



.'^ 



jum 



10 -"f" I J 



12 



lí*" 



Л I 



• • •. 



ЯУ. 



i* * 



7 ^ 



10 üm 



I0jjm 



A m 



FIG. 1. Pachytene chromosomes in male gonads of Tricolia speciosa. 

FIG. 2. (a) and (b) diakinetic bivalents of T. speciosa. 

FIG. 3. (a) spermatogonial metaphase chromosomes and (b) karyotype of T. speciosa (arrow indicates 

sub-metacentric pair). 

FIG. 4. Idiogram constructed from five metaphase plates of T. speciosa (arrow indicates sub-metacentric 

pair). 

FIG. 5. Ag-nucleus with two nucleoli. 

FIG. 6. Ag-nucleus with three nucleoli. 

FIG. 7. Ag-nucleus with six nucleoli. 

FIG. 8. NOR-negative diakinetic plate of T. speciosa. 

FIG. 9. NOR-positive diakinetic plate of T. speciosa (arrows indicate Ag-positive elements). 

FIG. 10. NOR-positive diakinetic plate of T. speciosa (arrows indicate Ag-positive elements). 

FIG. 11. Mitotic chromosomes at prophase stage of T. speciosa (arrows indicate Ag-NORs). 

FIG. 12. f\/litotic metaphase chromosomes of T. speciosa (arrows indicate Ag-NORs). 

FIG. 13. Mitotic metaphase chromosomes of T. speciosa (arrows indicate AG-NORs). 



214 



VITTURI & CATALANO 



TABLE 2. Mean length and arm ratio of the 
chromosomes of five metaphase plates of Tricolia 
speciosa. 

Chromosome Mean length, Arm ratio Centromere 
pairs ^JL ± SD mean position 



1 


2.6 ± 0.58 




M 


2 


2.5 ± 0.50 




M 


3 


2.4 ± 0.54 




M 


4 


2 ± 0.45 




M 


5 


2 ± 0.44 




M 


6 


1.9 ±0.36 




M 


7 


1.9 ± 0.36 




M 


8 


1.8 ± 0.34 


.4 


SM 



TABLE 3. Frequency of nucleoli/nucleus 
Tricolia speciosa. 


in 


No. of nucleoli/nucleus 


1 


2 3 4 5 


6 


Nuclei 32 
Frequencies 12 

% 


40 107 51 28 
15 41 19 11 


5 
2 



dosa does not differ from that of other mol- 
luscan species (Patterson, 1969). Cytological 
characteristics such as pachytene chromo- 
somes with irregular outlines and chiasmatic 
bivalents, constantly reported within the Mol- 
lusca (Vitturi et al., 1982; Vitturi & Catalane, 
1984; Vitturi et al., 1985b; Vitturi et al., 1986), 
were observed. 

Distant somatic pairing between homolo- 
gous chromosomes at the metaphase stage 
has been described for Haliotis tubercolata 
(Prosobranchia, Archaeogastropoda) (Co- 
lombera & Tagliaferh, 1983a) and Acantho- 
chiton crinitus (Polyplacophora) (Colombera 
& Tagliaferh, 1983b) but was not seen in our 
preparations. In fact, a random distribution of 
the mitotic metaphase chromosomes on the 
squashing plane was observed. 

The absence of heterotypic elements 
among spermatocyte bivalents, and of heter- 
omorphic pairs among male mitotic chromo- 
somes, allowed us to exclude a XY sex-de- 
termining mechanisms in Tricolia speciosa. At 
present, within the Archaeogastropoda only 
species included in the family Neritidae show 
a male XO sex-chromosome system (Naka- 
mura, 1983; Vitturi & Catalane, 1988). How- 
ever, a chromosome value of 8 bivalents ob- 
served for Tricolia speciosa suggests that this 
species has the lowest chromosome number 
within the Archaeogastropoda (Table 1). 



If we accept the idea that evolution in gen- 
eral (Mayr, 1970; Colombera & Lazzaretto- 
Colombera, 1978), and within the phylum 
Mollusca in particular (Vitturi et al., 1982; Ra- 
sotto & Cardellini, 1983; Vitturi et al., 1985a), 
proceeds via a decrease of chromosome 
number, although exceptions are certainly 
known (Vitturi et al., 1983), then the special- 
ization of Tricolia speciosa is apparent. More- 
over, it is held that evolved karyotypes are 
more symmetrical than those observed in the 
generalized species (Ohno, 1970; Colombera 
& Vitturi, 1978; Vitturi et al, 1987). If so, the 
specialization of Tricolia speciosa, which is 
remarkable in having all bi-armed chromo- 
somes, would be further supported. 

Data obtained from this study suggest that 
the Ag-staining pattern was, in this species, 
variable, as shown by the differences in the 
number of nucleoli/nucleus and in the number 
of chromosomes involved in the nucleolar or- 
ganization. This variability, previously re- 
ported in fish (Howell & Black, 1979; Foresti 
et al., 1981; Thode et al., 1983) and in mam- 
mals (Goodpasture & Bloom, 1975; Hender- 
son et al., 1976;Devetal., 1977; Mikelsaaret 
al., 1 977a,b; Winking et al., 1 980), is currently 
interpreted as a differential transcriptional ac- 
tivity of the ribosomal DNA (Miller et al., 
1976). Our results showing a correlation be- 
tween the number of nucleoli and their dimen- 
sions seem to be consistent with the idea that 
nucleoli in interphase tend to fuse (Goldman 
etal., 1983). 

Because chromosomes stained with acetic- 
orcein showed no achromatic zones, NORs 
are in our opinion unrelated to any satellite 
region in the species under study. In Gobius 
fallax (Pisces, Gobiidae) (Thode et al., 1983) 
and in other fish species (Almeida Toledo et 
al., 1981), the same conclusion was reached. 

Comparatively small Ag dots in the chromo- 
somes of diakinesis involving from zero to al- 
most all bivalents, were observed. This fact 
leads us to speculate that in this species, as 
in human cells (Schwarzacher et al., 1978), a 
decrease in the NORs activity at meiotic 
metaphase-l occurs. However, in Tricolia 
speciosa it seems that a higher number of 
elements are involved in this activity at mei- 
otic metaphase-l rather than at mitotic stages. 



ACKNOWLEDGMENTS 

This research was supported by grant: 
Ricerca Scientifica 60%, 1 986-87. 



CHROMOSOMES IN TRICOLIA 



215 



LITERATURE CITED 

ALMEIDA TOLEDO, L.F., F. FOREST! & F.S.A. 
TOLEDO, 1981, Constitutive heterochromatin 
and nucleolus organizer region in the knife fish 
Apteronotus albifrons (Pisces, Apteronotidae). 
Expehentia, 37: 953-954. 

COLOMBERA, D. & I. LAZZARETTO-COLOM- 
BERA, 1 978, Chromosome evolution in some ma- 
rine invertebrates. In B. BATTAGLIA & J. BEARD- 
MORE, eds., "Marine organisms," pp. 487-525. 

COLOMBERA, D. & F., TAGLIAFERRI, 1983a, 
Chromosomes from male gonads of Haliotis tu- 
bercolata and Haliotis lamellosa (Haliotidae, Ar- 
chaeogastropoda, Mollusca). Caryologia, 36: 
231-234. 

COLOMBERA, D. & F. TAGLIAFERRI, 1983b, 
Chromosome from male gonads of Acanthochi- 
ton crinitus (subclass Polyplacophora, Mollusca). 
Caryologia, 36: 293-297. 

COLOMBERA, D. & R. VITTURI, 1978, I chromo- 
somi nel sottordine Phlebobranchiata (Ascidia- 
cea). Caryologia, 31: 479-485. 

DEV, V.G., R. TANTRAVAHI, D.A. MILLER & O.J. 
MILLER, 1977, Nucleolus organizers in Mus 
musculus subspecies and in the RAG mouse cell 
line. Genetics, 86: 389-398. 

FORESTI, F., L.F. ALMEIDA TOLEDO & F.S.A. 
TOLEDO, 1981, Polymorphic nature of nucleolus 
organizer regions in fishes. Cytogenetics and 
Cell Genetics, 31 : 137-144. 

FRANC, A., 1968, in P.P. GRASSE, ed., "Traité de 
Zoologie," Tome V, Mason, Paris: pp. 40-324. 

GOLDMAN, M.A., P.T LOVERDE, CL. CHRIST- 
MAN, D.A. FRANKLIN, F. MATTHEWS, R.J. 
PICHFORD & CS. RICHARDS, 1983, Nucleolar 
organizer regions in Biomphalaria and Bulinus 
snails. Expehentia, 39: 911-913. 

GOODPASTURE, С & S.E. BLOOM, 1975, Visu- 
alization of nucleolar organizer regions in mam- 
malian chromosomes using silver staining. Chro- 
mosoma, 53: 37-50. 

HENDERSON, A.S., E.M. EICHER, T YU & K.C 
ATWOOD, 1976, Variation in hbosomal RNA 
gene number in mouse chromosomes. Cytoge- 
netics and Cell Genetics, 17: 9-17. 

HOWELL, W.M. & D.A. BLACK, 1979, Location of 
the nucleolus organizer regions in the sex chro- 
mosomes of the banded killifish, Fundulus di- 
aphanus. Copela, 3: 544-546. 

HOWELL, W.M. & D.A. BLACK, 1980, Controlled 
silver-staining of nucleolus organizer regions with 
a protective colloidal developer: a 1 -step method. 
Expehentia, 36: 1014-1015. 

KLIGERMAN, A.D. & S.E. BLOOM, 1977, Distribu- 
tion of F-bodies, heterochromatin, and nucleolar 
organizers in the genome of the central mud- 
minnow, Umbra limi. Cytogenetics and Cell Ge- 
netics, 18: 182-196. 

LEVAN, A., K. FREDGA & A.A. SANDBERG, 1964, 
Nomenclature for centromeric position of chro- 
mosomes. Hereditas, 52: 201-220. 

MARKOVIC, V.D., R.G. WORTON & J.M. BERG, 



1978, Evidence for the inheritance of silver- 
stained nucleolus organizer regions. Human Ge- 
netics, 41: 181-187. 

MAYR, E., 1970, L'evoluzione delle specie anima- 
lio. Giulio Einaudi editore s.p.a., Torino, Vol. II, 
pp. 485. 

MIKELSAAR, A.V., M. SCHMID, W. KRONE, H.G. 
SCHWARZACHER & W. SCHNEDL, 1977a, Fre- 
quency of Ag-stained nucleolus organizer re- 
gions in the acrocentric chromosomes of man. 
Human Genetics, 37: 73-77. 

MIKELSAAR, A.V., H.G. SCHWARZACHER, W. 
SCHNEDL & P. WAGENBICHLER, 1977b, Inher- 
itance of Ag-stainability of nucleolus organizer re- 
gions investigation in 7 families with trisomy 21. 
Human Genetics, 38: 183-188. 

MILLER, D.A., U.G. DEV, R. TANTRAVAHI & O.J. 
MILLER, 1976a, Suppression of human nucleo- 
lus organizer activity in mouse-human somatic 
hybrid cells. Experimental Cell Research, 101: 
235-243. 

NAKAMURA, H.K., 1982, Karyotypes of three spe- 
cies of Notoacmea (Gastropoda, Acmaeidae). 
Seto Marine Biology Laboratory, XXVII (1/5): 1 7- 
23. 

NAKAMURA, H.K., 1983, Karyological studies of 
Neritidae (Streptoneura: Archaeogastropoda). I. 
Chromosomes of five species from Hong Kong, 
with special reference to the sex chromosomes. 
Proceedings of the Second International Work- 
shop on the Malacofauna of Hong Kong and 
Southern China (Eds. B. Morton & D. Dugeon): 
257-273. 

NAKAMURA, H.K., 1986, Chromosomes of Ar- 
chaeogastropoda (Mollusca: Prosobranchia), 
with some remarks on their cytotaxonomy and 
phylogeny. Seto Mahne Biology Laboratory, 
XXXI (3/6): 191-267. 

OHNO, S., 1970, The enormous diversity in ge- 
nome sizes of fish as a reflection of nature's ex- 
tensive experiments with gene duplication. 
Transactions of the American Fishery Society, 99 
(1): 120-130. 

PARDUE, M.L. & TC HSU, 1975, Locations of the 
18S and 28S hbosomal genes on the chromo- 
somes of the Indian muntjak. Journal of Cell Bi- 
ology, 64: 251-254. 

PARENZAN, P., 1970, Carta d'identità delle con- 
chiglie del Mediterráneo (Gasteropodi), vol. I. 
BIOS Taras Eds. 

PATTERSON, СМ., 1969, Chromosomes of Mol- 
luscs. Proceedings of Symposium on molluscan 
Mahne Biology Association, India 2: 635-686. 

RASOTTO, M. & P. CARDELLINI, 1983, The chro- 
mosomes of Dendrodoris limbata and Dendro- 
doris grandiflora (Gastropoda, Mollusca). Caryo- 
logia, 36: 175-181. 

SCHWARZACHER, H.G., A.V. MIKELSAAR & W. 
SCHNEDL, 1978, The nature of the Ag-staining 
of nucleolus organizer regions. Electron and light 
microscopic studies on human cells in inter- 
phase, mitosis and meiosis. Cytogenetics and 
Cell Genetics, 20: 24-39. 



216 



VITTURI & CATALANO 



THODE, G., 1987, Karyotype analysis of the ding- 
fish, Lepadogaster condollei Risso ( Gobiesoci- 
formes). Cytobios, 51: 163-169. 

THODE, G., M.C. ALVAREZ, V. GILES & E. GAR- 
CIA, 1985, Chromosome complement, C- 
banding and Ag-NOR location in Ophysurus ser- 
pens (Ophichtidae, Anguilliformes). Cytobios, 43: 
73-77. 

THODE, G., J. CANO & M.C. ALVAREZ, 1983, As- 
sociation of nucleolus organizer chromosomes 
shown by silver staining in Gobius fallax. Journal 
of Heredity, 74: 480-482. 

TRAUT, W., H. WINKING & S. ADOLPH, 1984, An 
extra segment in chromosome 1 of wild Mus 
musculus a C-band positive homogeneously 
staining region. Cytogenetics and Cell Genetics, 
38: 290-297. 

VITTURI, R., P. CARBONE & E. CATALANO, 
1985a, The chromosomes of Pycnodonta co- 
chlear (Poli) (r\/lollusca, Pelecypoda). Biologis- 
ches Zentralblatt, 104: 177-182. 

VITTURI, R. & E. CATALANO, 1984, Spermatocyte 
chromosomes in 7 species of the sub-class 
Prosobranchia (Mollusca, Gastropoda). Biologis- 
ches Zentralblatt, 103: 69-76. 

VITTURI, R. & E. CATALANO, 1988, A male XO 
sex-determining mechanism in Theodoxus me- 



ridionalis (Neritidae) (Prosobranchia, Archaeo- 
gastropoda). Cytologia, 53: 131-138. 

VITTURI, R., E. CATALANO & M. MACALUSO, 
1986, Chromosome studies in three species of 
the Gastropod superfamily Littorinoidea. Malaco- 
logical Review, 19: 53-60. 

VITTURI, R., E. CATALANO, M. MACALUSO & A. 
MAIORCA, 1987, Spermatocyte chromosomes 
in six species of Neogastropoda (Mollusca, 
Prosobranchia). Biologisches Zentralblatt, 106: 
81-88. 

VITTURI, R., E. CATALANO, M. MACALUSO & N. 
PARRINELLO, 1985b, The chromosomes of cer- 
tain species of the sub-class Opisthobranchia 
(Mollusca, Gastropoda). Biologisches Zentral- 
blatt, 104: 701-710. 

VITTURI, R., A. MAIORCA & E. CATALANO, 1983, 
The karyology of Teredo utriculus (Gmelin) (Mol- 
lusca, Pelecypoda). Biological Bulletin, 165: 
450-457. 

VITTURI, R., M.B. RASOTTO & N. FARINELLA- 
FERRUZZA, 1982, The chromosomes of 16 mol- 
luscan species. Bollettino di Zoología, 49: 61-71 . 

WINKING, H., K. NIELSEN & A. GROPP, 1980, 
Variable positions of NORs in Mus musculus. Cy- 
togenetics and Cell Genetics, 26: 158-164. 

Revised Ms. accepted 4 October 1 988 



MALACOLOGIA, 1989,31(1): 217-227 

FEEDING EXPERIMENTS ON AND ENERGY FLUX IN A NATURAL POPULATION 

OF THE EDIBLE SNAIL HELIX LUCORUM L. (GASTROPODA: PULMONATA: 

STYLOMMATOPHORA) IN GREECE 

A. Staikou & M. Lazahdou-Dimithadou 

Laboratory of Zoology 

Department of Biology, 

University of Ttiessaloniki, 

54006 Thessaloniki, Greece 

ABSTRACT 

Energy flux in Helix lucorum was studied using as food Lactuca sativa, Urtica dioica and 
Petasites albus. The highest daily consumption and assimilation rates were observed in newly 
hatched snails and the lowest rates in adult snails. Assimilation efficiency, mean monthly pro- 
duction, as well as the growth (Pg/1%) and ecological (Pg/A%) efficiencies, fluctuated with 
season, the generation and with the physiological state of the snails. Snails fed on L. sativa 
showed higher assimilation efficiency than those fed on U. dioica or P. albus. Ingestion rate was 
found equal to 19.7% if snails were fed on U. dioica and 14.6% if they were fed on P. albus. 
Energy flow through H. lucorum population was 51 .7 Kcal/m^/year if snails were fed on U. dioica 
and 29.2 Kcal/m^/year if they were fed on P. albus. 

Key words: feeding experiments; energy flux; consumption; nutritional budget; I4elix lucorum. 



INTRODUCTION 

Ingestion and assimilation are two essential 
phases of energy transport from one trophic 
level to another, and thus they compose an 
important part of ecosystem functioning. Ter- 
restrial gastropods, as primary consumers, 
play an Important role in matter and energy 
transport from producer level to upper trophic 
levels. 

Studying the role of terrestrial molluscs In 
dynamics of woodland ecosystems, first 
Lindquist (1941) and then Mason (1970b) 
stressed the need for quantitative studies on 
food consumption and assimilation. Many 
studies have been published since on terres- 
trial pulmonates, such as those of Stern 
(1968, 1975) on Arion ruf us and Agriolimax 
reticulatus, of Jennings & Barkham (1976) on 
Arion ater, of Zelfert & Shutov (1978, 1981) 
on Bradybaena fruticum and Eobania vermic- 
ulata, of Lazarldou-Dlmitrladou & Daguzan 
(1978) on Euparypha pisana, of Charrier & 
Daguzan (1980) on Helix aspersa, and of 
Lazahdou-Dimitrladou & Kattoulas (sub- 
mitted) on Eobania vermiculata. Similarly, 
many studies on population bioenergetics of 
freshwater pulmonates and prosobranchs ex- 
ist, such as those of Aldridge et al. (1986) on 
Viviparus georgianus, of Russell-Hunter et al. 



(1983, 1984) on Helisoma trivolvis and Lym- 
naea palustris, and of Aldridge (1 982) on Lep- 
toxis carinata, etc. 

The present study forms part of a wider In- 
vestigation Into ecophysiology of the edible 
snail Helix lucorum In Greece. Reported here 
are the results concerning the experiments on 
food consumption and assimilation in the lab- 
oratory, and estimates of population metabo- 
lism in the field. 



METHODS AND MATERIALS 

The experiment lasted from April 1984 to 
March 1985. The snails used in the experi- 
ment were collected from a natural habitat of 
Helix lucorum in the Logos region of Edessa, 
northern Greece, where its ecology and biol- 
ogy have been studied. Every month, nine 
snails from each generation present In the 
field at that time were collected and trans- 
ferred to the laboratory. The number of gener- 
ations was known from the demographic 
analysis of the populations of Helix lucorum. 
Adult snails have a wide overlap of genera- 
tions that could not be distinguished; there- 
fore, the assumption that adults of different 
ages (snails with an age of three years or 
more) belong to the same cohort was taken 



217 



218 



STAIKOU & LAZARIDOU-DIMITRIADOU 



into account (Staikou et al., 1988). During 
winter months, that is December, January 
and February, no experiments were done be- 
cause the snails hibernated. 

The experiments were carried out in the 
laboratory under semi-natural conditions. 
Lighting followed the natural cycles, and tem- 
perature coincided with natural temperature 
at the given month. Snails were kept in indi- 
vidual glass chambers (40 x 20 x 15 cm), 
and high humidity (~ 85%) was supplied by a 
piece of sponge soaked with water and a 
small pot of water. Three different kinds of 
food were used: (a) Lactuca sativa, which in 
general is considered as "good" food for 
snails, and (b) Urtica dioica and Petasite sal- 
bus, which were the most abundant food re- 
sources in the study site of Helix lucorum. 
There were three replicates for each kind of 
food, that is nine replicates for each genera- 
tion. Each month, 36 to 54 chambers were 
used depending on the number of genera- 
tions that existed in the field. 

The methodology followed was that of Laz- 
aridou-Dimitriadou & Daguzan (1978). The 
amount of excrement produced daily was de- 
termined by means of a marker technique 
(Phillipson, 1960). Before the experiment, 
snails were fed carrot slices, which resulted in 
colour-marked faeces. Then the snails were 
weighed, measured and exposed to the ex- 
perimental food, which was presented in 
pieces of specific surface area (4 cm x 4 
cm). Food was replaced every 24 hours. The 
amount of food consumed was estimated as 
the area grazed by the individual snails, mea- 
sured by the surface area of the remnant with 
a planimeter. This method was used as it 
gave the best results after preliminary exper- 
imentation with other methods, such as the 
difference of live weight between the food 
given to the snails and control food material 
kept under the same conditions (Bogucki & 
Helczyk-Kazecka, 1977). The method of dry- 
ing food material, weighing, then rehydrating, 
giving it to the snails, and then drying and 
weighing again, which has been shown useful 
in feeding experiments with aufwuchs as food 
for fresh- water snails (Tashiro et al., 1980), 
could not be used, as our snails would not 
feed on food treated this way. The dry weight 
of food consumed was calculated for each 
kind of food by using the regression equations 
of leaf surface on leaf dry weight (16 cm^ of L. 
sativa is equal to 0.043 ± 0.0064 g dry 
weight, 16 cm^ of U. dioica is equal to 0.0518 
± 0.0095 g dry weight, and 1 6 cm^ of P. albus 



is equal to 0.0379 ± 0.0037 g dry weight). 
The above equations were obtained by using 
60 pieces of 16 cm^ from each kind of food; 
30 pieces were collected during spring (April) 
and 30 more during autumn (October); their 
dry weight was obtained after drying in vacuo 
in the presence of СаСОз. The faeces of 
each snail were collected, dried and weighed. 

After the experimental period, which lasted 
seven days, the snails were again measured, 
weighed and given carrot food. The faeces 
were collected until the coloured marker 
faeces appeared. At the end of each experi- 
ment, the snails were killed, the shell was 
separated from the body and both were dried 
in vacuo at room temperature in the presence 
of СаСОз. Dry weights of shell and body were 
taken seven days later. 

To quantify the daily consumption and as- 
similation rates as well as the growth and eco- 
logical efficiencies, the same formulae as 
Lazaridou-Dimitriadou & Kattoulas (sub- 
mitted) and the I.E. P. global productivity sym- 
bols listed by Petrusewicz & f^acfayden 
(1970) were used. 



C(mg) 



Daily consumption rate = 



Daily faecal production rate 



Daily assimilation rate = 



Assimilation efficiency = 

C(mg) 

Production (Pg) or GP = the amount of dry tis- 
sue elaborated in 
the snail body and 
shell per unit of time 
(mg/month) 



lie — 

LW.(g) 


FU(mg) 
'''' = L.W.(g) 


C(mg)-FU(mg)* 


L.W.(g) 


C(mg)-FU(mg) 



Growth efficiency 



Pg(mg) 



X 100 



(or gross growth efficiency) Mmg) 

Pg(mg) 

Ecological efficiency = x 100 

(or net growth efficiency) A(mg) 



*C (mg)— FU (mg) stands for ТА (total assimilated) accord- 
ing to conventional component labels (Russell-Hunter & 
Buckley, 1983) 



ENERGY FLUX IN HELIX LUCORUM 



219 



[where С** = dry weight of food consumed 
daily, FU*** = dry weight of faeces produced 
dally, L.W. = mean snail live weight 
(body + shell), I = dry weight of food ingested 
per month (C(mg) x 30), A = dry weight of 
food assimilated per month (C(mg) x 30 — 
F(mg) X 30)]. 

Monthly production, that is dry-weight gain 
of each snail could not be directly measured. 
It was extrapolated by the regressions of the 
dry body and shell weight in relation to the 
largest shell diameter(D) and the calculated 
organic content of the shell. Different regres- 
sions were used for juvenile and adult H. lu- 
corum, because it was known from the study 
of the relative growth that their growth rate 
differs (Staikou et al., 1988): 

For D < 22 mm the following regressions, 
where Wb = dry body weight and Ws = dry 
shell weight, were used: 

Log Wb = 2.592 Log D -3.884 (N = 123, r^ 

= 0.884) 
Log Ws = 3.16 Log D -4.7 (N = 123, r^ 

= 0.835) 

for 21 mm < D < 36 mm there were used: 

Log Wb = 2.801 Log D -4.11 (N = 163, r^ 

= 0.754) 
Log Ws = 3.865 Log D -5.527 (N = 163, r^ 

= 0.802) 

and for D > 36 mm there were used the fol- 
lowing: 

Log Wb = 3.338 Log D -4.945 (N = 118, r^ 

= 0.319) 
Log Ws = 3.114 Log D -4.408 (N = 118, r^ 

= 0.398) 

For the determination of the shell organic 
matter, a known quantity of homogenated 
shell material was treated with 5 N HCl solu- 
tion, the remainder was treated with distilled 
water six to seven times to wash away the 
calcium chloride (CaCl2) left and then dried at 
65°C. The shell organic matter was deter- 
mined as the residual weight of dry shell 
weight left after the above-described treat- 
ments. The replicability of these measures 
was checked by burning a known quantity of 



**C{mg) stands for Tl (total ingested) according to conven- 
tional component labels 
(Russell-Hunter & Buckley, 1983) 
***FU stands for NA (not assiглilated) according to conven- 
tional component labels 
(Russell-Hunter & Buckley, 1983) 



homogenated shell material, after drying it to 
constant weight, in a muffle-furnace at 560°C 
to obtain by difference an ash-free dry weight. 

The best method of computing organic 
growth is microbomb calorimetry, that is as- 
sessment of energetic equivalents of organic 
biomass, or analyses of fat, protein and car- 
bohydrates at all stages (Russell-Hunter et 
al., 1968). Another widely used method is es- 
timating organic carbon by wet oxidation 
(Russell-Hunter et al., 1968), and the C/N ra- 
tio at all stages. In this study, bomb calorim- 
etry was used mainly to produce comparable 
results with most of the existing studies on 
terrestrial snails. Thus, all rates of consump- 
tion, egestion and assimilation, as well as 
production and growth and ecological effi- 
ciencies, were computed in terms of both dry 
weight and energetic values. The energy con- 
tent of H. lucorum body, shell organic matter, 
and faeces, as well as the energy content of 
the three food materials, was determined on a 
Phillipson microbomb calorimeter. For each 
sample, two subsaniples were burnt and 
whenever a difference greater than 0.05 ap- 
peared a third and sometimes a fourth sub- 
sample was used. 

Appendices with detailed calculations of all 
the rates and efficiencies used can be ob- 
tained by the Department of Zoology, School 
of Biology, Aristotelion University of Thessal- 
oniki, 54006 Thessaloniki, Greece. 



RESULTS 

The percentage of organic matter in the 
shell of H. lucorum was found to equal 1 .7%. 
The caloric value of the organic material of 
the shell was 4.797 ± 0.24 cal/mg ash free 
dry weight. 

A comparison between dry weight of food 
eaten and dry weight of faeces produced re- 
vealed a positive correlation between the 
above two parameters for all food materials 
used. Coefficient correlation was very high for 
animals fed on Lactuca (r = 0.951 , N = 43) 
and Petasites (r = 0.907, N = 43) and some- 
what lower for animals fed on Urtica (r = 
0.751, N = 43). 

The highest values for daily consumption 
rate were observed in newly hatched snails, 
aged one month for animals fed on Lactuca 
(89.55 mg/g) and Urtica (61.73 mg/g), and in 
juveniles aged three months for animals fed 
on Petasites (28.26 mg/g). Values of this pa- 
rameter declined with age and became very 



220 



STAIKOU & LAZARIDOU-DIMITRIADOU 



low in mature animals with a largest shell di- 
ameter greater than 35 mm. For animals fed 
on Lactuca, the lowest value observed was 
2.63 mg/g. For those fed on Urtica, 1.11 mg/g, 
and for those fed on Petasites, 0.07 mg/g. 
Values of daily faecal production rate and 
daily assimilation rate followed the general 
pattern of daily consumption rate for all kinds 
of food used. 

The values of the above parameters of the 
individual nutritional budget of the snails were 
also influenced by the time of the year or by 
the physiological state of the animals. Thus, 
high values appeared during spring, espe- 
cially in May, and autumn (September, 
October). Also, high values in adult snails 
were shown in June before the reproductive 
period. 

Overall assimilation efficiency was higher 
and more constant in animals fed on Lactuca 
(82%) than in animals fed on Urtica (73%) 
and Petasites (59%) (Table I). Values of this 
parameter calculated for the different gener- 
ations showed that young snails prefer Lac- 
tuca and Urtica and show a smaller prefer- 
ence for Petasites. Mature snails show a 
marked preference for Lactuca while their as- 
similation efficiency was almost the same, 
ranging from 30%-80% when fed on Urtica or 
Petasites (Figs. 1-3). 

Values of mean monthly production (Pg), 
growth (Pg/I) and ecological (Pg/A) efficien- 
cies varied with the season and/or the phys- 
iological state of the snails, becoming highest 
in June irrespective of the kind of food. Also, 
high values were observed in September or 
November and sometimes in March and April 
(Figs. 1-3). 

In general, values of growth and ecological 
efficiencies were higher in snails fed on Pet- 
asites and lower in snails fed on Urtica or Lac- 
tuca (Table I). It has to be stressed, though, 
that these values were underestimated be- 
cause mucus production was not taken into 
consideration (Lamotte & Stern, 1987). 

Using the values of the calorific content of 
the body and the excrement of the snails at 
the end of each experiment, it was possible to 
convert the parameters of the individual nutri- 
tional budget of H. lucorum in caloric values 
(Table II). 

Values of monthly ingestion (El) and 
monthly assimilation (EA) fluctuated accord- 
ing to the season or/and the physiological 
state of the animals. Highest values were al- 
ways observed in late spnng (May) and in au- 
tumn (September, October) (Figs. 1-3). 



Snails fed on L. sativa also showed high val- 
ues in June and July (Fig. 1). 

Fluctuations in production (EPg), gross 
growth (EPg/EI) and net growth (EPg/EA) ef- 
ficiencies follow the same pattern as when 
these parameters are calculated in terms of 
dry weight (Figs. 1-3). 

Ingestion rate, which shows the popula- 
tions' impact on the environment, was esti- 
mated from the values of annual turnover ra- 
tio (P/B = 1.24) (Staikou et al., 1988) as well 
as the values of growth efficiency EPg/EI 
(which were 0.65 and 0.85 when snails were 
fed on Urtica or Petasites respectively). In- 
gestion rate was found equal to 19.7% if 
snails were fed on Urtica and 14.6% if they 
were fed on Petasites. 

Energy spent for egg production was cal- 
culated by: (a) the caloric content of eggs 
(Table II), (b) the mean number of eggs laid 
(50.5 ± 21 .3), and (c) their mean weight (0.43 
± 0.12 g.). By multiplying the eggs laid per 
snail per year by their mean weight, and then 
this number by the caloric content of eggs, the 
mean reproductive output in terms of energy 
values for any adult snail was assessed. 
Knowing the duration of life of H. lucorum (14 
years) and the number of eggs laid by an in- 
dividual the first and following years (Staikou 
et al., 1988), the reproductive output in ener- 
getic values was calculated for the life span of 
H. lucorum. 

It was known by the feeding experiments 
the energy ingested and assimilated by an 
individual till its maturity, as well as the energy 
assimilated during a year of an adult's life. 
Multiplying the last value, which corresponds 
to an adult's life, by the number of years an 
adult snail may live after the attainment of its 
maturity, it was possible to compute the indi- 
vidual energy budget of H. lucorum during its 
life time (Table I). It was found that of the total 
assimilated energy a snail spends 14.8% for 
growth, 2.3% for egg production, and 82.9% 
in metabolic energy when fed on Urtica, and 
22.5%, 3.7% and 73.8%, respectively, when 
fed on Petasites. It was also found that the 
reproductive output was equal to 13.6% and 
14.3% of the non-metabolic assimilated en- 
ergy when snails were fed on Urtica and Pet- 
asites respectively. 



DISCUSSION 

As stated by Russell-Hunter et al. (1968), 
the organic material in the shell represents 



ENERGY FLUX IN HELIX LUCORUM 



Ю CD Ю 
■^' C\i ОЗ 



h- О 00 

ю ih c\i 

о CJ т- 

со см 



221 



CÛ 



со о о 
■^ см ю 



со ■* см 



со см 1- 
■^ -г-' c\i 



CJ) -^ in 
о О) о 

■* 00 со 

Ю ■* 00 

ti ■^ "^ 
со см 



^ со со 

т- со со 

о Г^ тз- 

00 1^ т- 

LO ■* Ю 

ю со 



см см о 
со о со 
1^ см ю 

о см Tí 

■^ со со 

Ю 1- 



г: •^ о см 

со со см со 
со ^í- ю ■^ 



■* -^ со 

со -^ 'í- 

■^ со см 

■^ t^ 1- 

t^ см ■.- 



;? '" о 

t^ со СП 
О) см ю 

со 00 о 

t^ о т- 

00 со 



о 

g ^ 

« Е 

OJ (Я 

^ < 



ее с: 

о О) о 1_ 



t^ см см 
ю ■.- см 



оз 


1— 


см О) 'í 


00 


со 


со 1- 


о 






со 


ш 





со о 

CD ССЗ 1Г> 





ОС) Tf 


t 


СМ CM со 


LO 


00 


ю 




in 




(J) 




CM 





Q. _ i= 



о ш ш :s Ш Ш Ш I 



222 



STAIKOU & LAZARIDOU-DIMITRIADOU 



25000 т 

El 20000 ■■ 
(cal) 

15000 + 

EA 
(cal) 10000 



5000 Я 



Lactuca 






Ir' 

ó 
I I I I I I I I I I I I I I I 






и V \д ft »^' \\ 
' VI ° '^ 



v4 
о 



-1—1 — I I I — I I I I I I — (— I 



9 11 4 6 

84 185 



10 3 5 7 9 11 4 6 8 10 3 
I 86 I 87 188 



EA/EI 
% 



100 

90 

80 

70.. 

бО" 

50 ■• 

EPg/EA 40- 

EPg/EI 
% 



S-o'\ 



0-0-0 



\ ,0-0-0-0.0' 



л 'ж\ 



0-0. 

,.0-0-0, / °-° 



L.4 Л.,.ч /\,/\ ,.._ Да .,. 

о j -ж I I ж«* lili x-^-t-x t I I т-х-жч-Ж|».ж-«-ж->-х-ж-»:1н 

9 11 468 10 3579 11 468 10 3 
84 I 85 I 86 I 87 I 88 



EPg 
(cal) 





. .. . /\ I 1/ \ / 

Q 4-O-t— i-O 'T I I I I oP-t-O I I I O-O-O-t-P'O-O-t-O-t -О-О-Он 

9 11 468 10 3579 11 468 10 3 
84 185 186 I 87 I 88 

Months 

FIG. 1. Mean monthly ingestion (cal) (El), mean monthly assimilation (cal) (EA), mean assimilation effi- 
ciences (%) (EA/EI), ecological efficiencies (%) (EPg/EA), growth efficiencies (%) (EPg/EI), and mean 
monthly production EPg (cal) during the life cycle of Helix lucorum fed on Lactuca sativa. 
[For the construction of Figure 1, the feeding experimental results of each generation (G1,G2,G3 or G4) 
known to be present in the field each month from April 1984 to March 1985 (according to the already 
published life-cycle data of Staikou et al., 1988) were combined in computations assuming that the first 
generation (Gl) is followed by G2 at the end of the first year and G2 is followed by G3 at the end of the 
second year and G3 by G4 at the end of the third year. So feeding and growth parameters could be followed 
monthly from hatching till the maturity of the snails (except during winter time when the snails hibernate) that 
is for 3.5 years from September 1984 to March 1988, although feeding experiments lasted one year.] 



EA/EI 
% 



EPg/EA 
% 

EPg/EI 
% 



ENERGY FLUX IN HELIX LUCORUM 
Utrica 



223 



El 
(cal) 

EA 
(cal) 



14000 т 

12000 

10000 

вооо ■■ 
бооо •■ 



2000 
О 




H/V t' 4 s/^^ \í°' 




H — I I I I ( — I I I — t—t- 



I I I I I I I I t I t I I 



9 11 4 6 8 10 ,3 5 7 9 11 4 6 8 10 
84 185 I 86 I 87 I 



100 ■• 
90 ■• 



70 

60 ■• 
50 •■ 

40 •■ 
30 ■• 
20 



o°'\ о 




ЧлР\ 1/ \Ä mí 

о i-X-t-^X'. till ^-l—t-X I I I I хжн-жж-ж-нж-нж-ж-жн 



9 11 4 
84 185 



68 10 3579 11 46 
I 86 I 87 



10 



EPg 
(cal) 



3000 
2500 
2000 

1500 
1000 



500 ■• 
О 

-I-' 




/\ 




/ 



O-t-Q.Q '? I I I I ol(— t-O till О O-t-OO O-f-O-bOOOH 

1 11 4 6 8 10 3 5 7 9 11 4 6 8 10 3 
84 185 I 86 I 87 I 88 

Months 

FIG. 2. Mean monthly ingestion (cal) (El), mean monthly assimilation (cal) (EA), mean assimilation efficien- 
cies (%) (EA/EI), ecological efficiencies (%) (EPg/EA), growth efficiencies (%) (EPg/EI) and mean monthly 
production EPg (cal) during the life cycle of Helix lucorum fed on Urtica dioica 

[For the construction of Figure 2, the feeding experimental results of each generation (G1,G2,G3 or G4) 
known to be present in the field each month from April 1984 to March 1985 (according to the already 
published life-cycle data of Staikou et al., 1988) were combined in computations assuming that the first 
generation (Gl) is followed by G2 at the end of the first year and G2 is followed by G3 at the end of the 
second year and G3 by G4 at the end of the third year. So feeding and growth parameters could be followed 
monthly from hatching till the maturity of the snails (except during winter time when the snails hibernate) that 
is for 3.5 years from September 1984 to March 1988, although feeding experiments lasted one year.] 



224 



STAIKOU & LAZARIDOU-DIMITRIADOU 



Petasites 



9000 •■ 
8000 •• 
7000 •■ 
^1 6000 

(cal) 5000 

4000 t 

EA 3000 
(cal) . 



2000 
1000 



Ô до \ Д 

IJ ^ V/ 




0V!^O-H 

9 11 

84 185 



I I I о I I I I I I I I O-t— I- 

468 10 3579 



1 — I I I I I P=o I I I — I 



11,4 6 8 10 3 
187 I 88 




EPg/EI 



2500 T 



EPg 
(cal) 



1000 



500 




911468 10 3579114 68 10 3 
84 I 85 I 86 I 87 I 88 

Months 

FIG. 3. Mean monthly ingestion (cal) (El), mean monthly assimilation (cal) (EA), mean assimilation efficien- 
cies (%) (EA/EI), ecological efficiencies (%) (EPg/EA), growth efficiencies (%) (EPg/EI) and mean monthly 
production EPg (cal) during the life cycle of Helix lucorum fed on Petasites albus. 
[For the construction of Figure 3, the feeding experimental results of each generation (G1,G2,G3 or G4) 
known to be present in the field each month from April 1984 to March 1985 (according to the already 
published life-cycle data of Staikou et al; 1988) were combined in computations assuming that the first 
generation (Gl) is followed by G2 at the end of the first year and G2 is followed by G3 at the end of the 
second year and G3 by G4 at the end of the third year. So feeding and growth parameters could be followed 
monthly from hatching till the maturity of the snails (except during winter time when the snails hibernate) that 
is for 3.5 years from September 1984 to March 1988, although feeding experiments lasted one year.] 



ENERGY FLUX IN HELIX LUCORUM 225 

TABLE 2. Calorific content, of Lactuca sativa, Urtica dioica and Petasites albus leaves, as well as of 
shell and egg matter of Helix lucorum (where N = number of trials, s = standard deviation) 



Data 



Lactuca sativa 
(N = 60) 
Urtica dioica 
(N = 60) 
Petasites albus 
(N = 60) 
Mean shell 
organic matter 
(N = 9) 

Mean egg matter 
(N = 4) 



Mean 

cal/mg ± s 

with ash 



Mean 
cal/mg ± s 
without ash 



3.7689 ± 0.2116 
2.4093 ± 0.2584 
3.9119 ± 0.0933 

4.2143 ± 0.2830 
3.1325 ± 0.0482 



4.0040 ± 0.2813 
3.2705 ± 0.2383 
4.2741 ± 0.0882 

4.7973 ± 0.2452 
3.8750 ± 0.0799 



Mean ash 
weight 

% ± s 



7.55 ± 0.51 

26.50 ±1.11 

8.14 ± 0.19 

18.40 ± 0.12 
20.52 ± 0.42 



Mean water 
weight 
% ± s 



92.00 ± 0.006 
72.20 ± 0.009 
89.00 ± 0.008 

75.20 ± 5.30 
82.40 ± 3.20 



Stored energy that is never turned over until 
death, except v\/here external erosion or inter- 
nal shell resorption takes place. The percent- 
age of the organic matter in the shell of H. 
lucorum was found lower than that reported 
for H. aspersa by Charrier & Daguzan (1980) 
and for Eobania vermiculata by Lazaridou- 
Dlmltriadou & Kattoulas (submitted). It was 
somewhat similar to that reported by Lazari- 
dou-Dlmitriadou & Daguzan (1978) for Eu- 
parypha pisana. The calorific content of the 
shell was similar to that reported by Hughes 
(1970) for the bivalve Scrobicularia plana, by 
Lazaridou-Dimitriadou & Daguzan (1978) for 
E. pisana, and by Charrier & Daguzan (1980) 
for H. aspersa. It was slightly lower than that 
reported by Lazaridou-Dimitriadou & Kattou- 
las (submitted) for E. vermiculata. 

High values of daily consumption rate, daily 
faecal production rate and daily assimilation 
rate In newly hatched snails may be due to 
their higher metabolic rate in relation to older 
ones. The same phenomenon has been ob- 
served in Arion ater (Jennings & Barkham, 
1976), in Agriolimax laevis (Stern, 1979), in 
Eobania vermiculata (Zeifert & Shutov, 1978; 
Lazaridou-Dimitriadou & Kattoulas, submit- 
ted), in Euparypha pisana (Lazaridou-Dimitri- 
adou & Daguzan, 1978), and in many non- 
marine prosobranch gastropods (Aldridge et 
al., 1986). The season of the year seemed to 
influence the values of the above parameters. 
The peaks observed in spring (mainly in May) 
were probably related to the fact that this is 
the period of maximum activity for H. lucorum 
in the field (Staikou et al., 1988). Minor peaks 
observed in autumn (e.g. September or/and 
October) were probably due to the fact that 
snails are less active than in May but accu- 
mulate food reserves prior to hibernation. 



Seasonal fluctuations in values of these pa- 
rameters have been also reported by Lazari- 
dou-Dimitriadou & Kattoulas (submitted) for 
E. vermiculata. Seasonal degrowth has been 
shown in freshwater pulmonate gastropods 
(Russell-Hunter, 1983, 1984). 

High assimilation efficiencies in animals fed 
on Lactuca have been reported by Bogucki & 
Helczyk-Kazecka (1977) for adult H. pomatia 
and by Charrier & Daguzan (1980) for H. as- 
persa. Mason (1970a) and Richardson 
(1975a) also found that snails show higher 
assimilation rates when fed on Lactuca and 
much lower when fed on Urtica. Assimilation 
efficiency drops in October or November just 
before hibernation and in May or June when 
snails are fed on Urtica and in July-August 
when fed on Petacites when higher tempera- 
tures occurred in 1984 (Staikou et al., 1988, 
fig. 2). The less constant assimilation effi- 
ciency when snails are fed on Urtica and Pet- 
asites might be attributed to their different 
quality each month, because they were col- 
lected from the field, whereas Lactuca came 
from cultivations throughout the year. The ef- 
fects of food quality on assimilation and dif- 
ferential catabolism have been shown in non- 
marine gastropods (Aldridge et al., 1986). 
Assimilation efficiency in animals fed on Ur- 
tica is somewhat similar to that reported by 
Jennings & Barkham (1976) for Arion ater 
(69%) and by Lazaridou-Dimitriadou & Kat- 
toulas (submitted) for E. vermiculata (81%). It 
is higher than that reported by Mason (1 970a) 
for Hygromia striolata (52.40-8.78%) and 
Discus rotundatus (47.70-8.89%) feeding on 
Urtica. The low efficiencies in the latter case 
may be due to the fact that Mason ran his 
experiments at 10°C. 

The peaks observed in the values of mean 



226 



STAIKOU & LAZARIDOU-DIMITRIADOU 



monthly production (Pg), growth (Pg/I) and 
ecological (Pg/A) efficiencies correspond to 
the months after the most rapid growth 
(March, April and mainly in June), or prior to 
hibernation (September to November) when 
food reserves are accumulated. The above 
differences were also assessed as changes 
in overall efficiencies or in physiological rates 
in non-marine prosobranch gastropods (Al- 
dridge et al., 1986). 

Ingestion rate when snails are fed only on 
Urtica accords with the value mentioned by 
Lazaridou-Dimitriadou & Kattoulas (sub- 
mitted) for E. vermiculata and by Richardson 
(1975b) and Williamson (1975) for Cepaea 
nemoralis. 

The calorific content of the snail's body is 
comparatively low in relation to that of other 
animals (Slobodkin & Richman, 1961; Slo- 
bodkin, 1962; Golley, 1961), and this is prob- 
ably due to the low quantity of lipids in the 
snail's body (Hughes, 1970). Knowing the dif- 
ference in the calorific content of the bodies of 
the mature snails before and after the repro- 
ductive period (June: 12896.2 cal-August: 
11647.3 cal = 1248.9 cal.) and the total re- 
productive output (1674 cal) (Table I), it was 
possible to calculate that 25.4% of the energy 
spent for egg production comes from concur- 
rent trophic input. 

Estimates of reproductive output as propor- 
tion of total assimilated energy (when snails 
were fed on Urtica or Petasites) accord well 
with the values given by Calow (1978) for 
some iteroparous fresh-water gastropods, 
whereas they were much lower than the val- 
ues given for the semelparous species. Esti- 
mates of reproductive output as a proportion 
of non-metabolic assimilated energy were 
lower than all the values reported by the same 
author for the iteroparous species, such as H. 
lucorum, and this may be related to the longer 
life span of H. lucorum. 

Knowing the mean number of snails in ev- 
ery size class/m^ (Staikou et al., 1988), as 
well as the quantity of food consumed by 
them per month, it was possible to estimate 
the annual consumption and the annual fae- 
cal production of the snails in the field. If 
snails would feed only on Urtica, annual con- 
sumption would equal 15.81 g/m^, and annual 
faecal production, 3.54 g/m^ (equivalent val- 
ues in calories were 51.7 Kcal/m^/year and 
17.3 Kcal/m^/year); mean assimilation effi- 
ciency for all size classes was 77.6%. If snails 
would feed only on Petasites, the annual con- 
sumption would equal 6.8 g/m^, and annual 



faecal production, 2.6 g/m^ (equivalent values 
in calories were 29.2 Kcal/m^/year and 13.1 
Kcal/m^/year); mean assimilation efficiency 
for all size classes was 61 .8%. The annual 
consumption values found in this study, are 
higher than those found by Mason (1970b) for 
different snail species fed on beech litter, by 
Jennings & Barkham (1976) for Arion ater, 
and by Zeifert & Shutov (1978) for Brady- 
baena fruticum. These differences may be 
due to the different density of the snail spe- 
cies in the field or to the different food used for 
the above studies. The difference in field den- 
sity may also be another reason for the higher 
values of annual consumption and energy 
flow through the population of E. vermiculata 
fed only on Urtica (Lazaridou-Dimitriadou & 
Kattoulas, submitted). 



ACKNOWLEDGMENTS 

Thanks are extended to K. Asmi for her 
technical assistance. Financial support was 
provided by the Minister of Agriculture. 



LITERATURE CITED 

ALDRIDGE, D., W. D. RUSSELL-HUNTER & D. 
BUCKLEY, 1986, Age-related differential catab- 
olism in the snail, Vivíparas georgianas, and its 
significance in the bioenergetics of sexual dimor- 
phism. Canadian Journal of Zoology, 64: 340- 
346. 

ALDRIDGE, D., 1982, Reproductive tactics in rela- 
tion to life-cycle bioenergetics in three natural 
populations of the freshwater snail Leptoxis car- 
inata. Ecology, 63(1): 196-208. 

BOGUCKI, A. & B. HELCZYK-KAZECKA, 1977, Ef- 
ficiency of food assimilation in the Roman snail 
Helix pomatia L. Bulletin de la Société des Amis 
des Sciences et des Lettres de Poznan, Série D, 
17 Livraison: 159-167. 

CALOW, P., 1978, The cost of reproduction— A 
physiological approach. Biological Review, 54: 
23-40. 

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

GOLLEY, F. В., 1961, Energy values of ecological 
materials, Ecology, 42: 581-584. 

HUGHES, R. N., 1970, An energy budget for a 
tidal-flat population of the bivalve Scrobicularia 
plana (Da Costa). Journal of Animal Ecology, 39 
(2): 357-381. 

JENNINGS, J. J. & J. P. BARKHAM, 1976, Quan- 



ENERGY FLUX IN HELIX LUCORUM 



227 



titative study of feeding in woodland by the slug 
Arion ater. Oikos, Copenhangen, 27: 168-173. 

LAMOTTE, M. & G. STERN, 1987, Les bilans én- 
ergétiques chez les mollusques pulmones. Hali- 
otis. 16: 103-128. 

LAZARIDOU-DIMITRIADOU, M. & J. DAGUZAN, 
1978, Consommation alimentaire, production et 
bilan énergétique chez Euparypha pisana 
(Müller) (Gastéropode Pulmoné). Annales de la 
Nutrition et de l' Alimentation. 32: 1317-1350. 

LAZARIDOU-DIMITRIADOU, M. & M. KATTOU- 
LAS (submitted). Energy flux in a natural popu- 
lation of the land snail Eobania vermiculata 
(Müller) (Gastropoda Pulmonata Stylommato- 
phora) in Greece. 

LINDQUIST, В., 1941, Experimentelle Untersu- 
chungen über die Bedeutung einiger Landmol- 
lusken für die Zersetzung der Waldstern. K. 
Fysiogr. Lunds Läkaresällskaps Förhandlingar, 
II: 1-13. 

MASON, С. F., 1970a, Food feeding rates and as- 
similation in woodland snails. Oecologia (Berlin), 
4: 358-373. 

MASON, С. F., 1970b, Snail populations, beech lit- 
ter production, and the role of snails in litter de- 
composition. Oecologia (Berlin), 5: 215-239. 

PETRUSEWICZ, K. & A. MACFADYEN, 1970, Pro- 
ductivity of terrestrial animals. Principles and 
methods. IBP Hand. 13 Blackwell, Oxford. 

PHILLIPSON, S., 1960, A contribution to the feed- 
ing biology of Mitopus morio F. (Phalangida). 
Journal of Animal Ecology, 29: 35-43. 

RICHARDSON, A. M. M., 1975a, Food, feeding 
rates and assimilation in the snail Cepaea nemo- 
ralis (L). Oecologia (Berlin), 19: 59-70. 

RICHARDSON, A. M. M., 1975b, Energy flux in a 
natural population of the land snail Cepaea ne- 
moralis. Oecologia (Berlin), 19(2): 141-164. 

RUSSELL-HUNTER, W. D., R. MEADOWS, M. AP- 
LEY & A. BURKY, 1968, On the use of a "wet- 
oxidation" method for estimates of total organic 
carbon in mollusc growth studies. Proceedings of 
the Malacological Society of London, 38: 1-11. 

RUSSELL-HUNTER, W. D. & D. E. BUCKLEY, 
1983, Actuarial Bioenergetics of nonmarine mol- 
luscan productivity. In: The l\/lollusca 6: 463-503, 
ed. by W. D. Russell-Hunter, Academic Press, 
Inc., N.Y. 



RUSSELL-HUNTER, W. D., D. ALDRIDGE, J 
TASHIRO & B. PAYNE, 1983, Oxygen uptake 
and nitrogenous excretion rates during ovenwin- 
ter degrowth conditions in the pulmonate snail, 
Helisoma trivolvis. Comparative Biochemistry 
and Physiology, 74A (3): 491-497. 

RUSSELL-HUNTER, W. D., A. BROWNE & D. AL- 
DRIDGE, 1984, Ovenwinter tissue degrowth in 
natural populations of freshwater pulmonate 
snails (Helisoma trivolvis and Lymnaea palus- 
tris). Ecology, 65(1): 223-229. 

SLOBODKIN, L.B., 1962, Energy in animal ecol- 
ogy. Advances in Ecological Research. Aca- 
demic Press. (Gragg, J. В.), I: 69-101. 

SLOBODKIN, LB. & S. RICHMAN, 1961, Calories/ 
gm. in species of animals. Nature, 191:299. 

STAIKOU, A., M. LAZARIDOU-DIMITRIADOU & N. 
FARMAKIS, 1988, Aspects of the life cycle, pop- 
ulation dynamics, growth and secondary produc- 
tion of the edible snail Helix lucorum Linnaeus 
1758 (Gastropoda Pulmonata) in Greece. The 
Journal of Molluscan Studies, 54: 139-155. 

STERN, G., 1968, Recherches sur le bilan éner- 
gétique de la limace Arion rufus L. en période de 
croissance. Doctorat 3e cycle, Paris. 

STERN, G., 1975, Effet de la temperature sur la 
production et al consommation chez Agriolimax 
reticulatus (M.) en période de croissance. Bulle- 
tin Ecologique, VI:501-509. 

STERN, G., 1979, Bilans pondéral et énergétique 
de croissance et de reproduction chez Agriolimax 
laevis (Müller), Pulmonata, Limacidae. Bulletin 
de la Société Zoologique de France, 104(2): 
147-161. 

TASHIRO, J., W. ALDRIDGE, & W. D. RUSSELL- 
HUNTER, 1980, Quantifiable artificial rations for 
molluscan grazing experiments. Malacological 
Review, 13: 87-89. 

WILLIAMSON, P., 1975, Use of ^^Zn to determine 
the field metabolism of the snail Cepaea nemo- 
ralis L. Ecology, 56: 1185-1192. 

ZEIFERT, D. V. & S. V. SHUTOV, 1978, Role of 
certain terrestrial molluscs in the transformation 
of leaf litter. Ekologiya, 5: 58-61. 

ZEIFERT, D. V. & S. V. SHUTOV, 1981, The con- 
sumption of leaf litter by land molluscs. Pedobio- 
logia, 21: 159-165. 

Revised Ms. accepted 16 July 1989 



Publication dates 
Vol. 28, No. 1-2 19 January 1! 
Vol. 29, No. 1 28 June 1988 
Vol. 29, No, 2 16 Dec. 1988 
Vol. 30, No. 1-2 1 Aug. 1989 



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, 19th and the 



Parkway, Philadelphia, Pennsylvania 19103, U.S.A. Publication date 
Vol. 27, No. 2-17 December, 1986 



WHY NOT SUBSCRIBE TO MALACOLOGIA? 

ORDER FORM 
Your name and address 



Send U.S. $17.00 for a personal subscription (one volume) or U.S. $27.00 for an 
institutional subscription. Mal<e checks payable to "MALACOLOGIA." 

Address: Malacologia, Academy of Natural Sciences 
Nineteenth and the Parkway, Philadelphia 
PA 19103, U.S.A. 



31(1) 



MALACOLOGIA 



1989 



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 translations 
(complete with main titles). Include all foreign 
accents. Both American and British spellings 
are allowed. 

4. Unless indicated otherwise below, con- 
tributors should follow the recommendations 
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 consistent. 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. 

10. Use the metric system throughout. Mi- 
cron should be abbreviated \xm. 

11. Illustrations are printed either in one 
column or the full width of a page of the 
journal, so plan accordingly. The maximum 
size of a printed figure is 13.5 x 20.0 cm 
(preferably not as tall as this so that the cap- 
tion does not have to be on the opposite 
page). 

12. Drawings and lettering must be dark 
black on white, blue tracing, or blue-lined 
paper. Lines, stippling, letters and numbers 
should be thick enough to allow reduction by 
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 jim," not "163 ixm"). 
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 dupli- 
cate 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 Litera- 
ture 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 reproduc- 
tions. 

22. Authors who want illustrations returned 
should request this at the time of ordering re- 
prints. Otherwise, illustrations will be main- 
tained for six months only after publication. 



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. 



SUBSCRIPTION COSTS 

25. For Vol. 31 , personal subscriptions are 
U.S. $17.00 and institutional subscriptions 
are U.S. $27.00. Address inquiries to the 
Subscription Office. 



VOL. 31, NO. 1 MALACOLOGIA 1989 

CONTENTS 

W. F. PONDER, R. HERSHLER, & B. J. JENKINS 

An Endemic Radiation of Hydrobiid Snails from Artesian Springs in Northern 
South Australia: Their Taxonomy, Physiology, Distribution and Anatomy 1 

R. TRIEBSKORN 

Ultrastructural Changes in the Digestive Tract of Dereceras reticulatum (Müller) 
Induced by a Carbamate Molluscicide and by Metaldehyde 141 

KENNETH С EMBERTON 

Retraction/Extension and Measurement Error in a Land Snail: Effects on Sys- 
tematic Characters 1 57 

PAULA M. MIKKELSEN & RÜDIGER BIELER 

Biology and Comparative Anatomy of Divariscintilla yoyo and D. troglodytes, Two 
New Species of Galeommatidae (Bivalvia) from Stomatopod Burrows in Eastern 
Florida 175 

ROBERT T. DILLON, JR. 

Karyotypic Evolution in Pleurocerid Snails. I. Genomic DNA Estimated by Flow 
Cytometry 1 97 

ROBERT С BAILEY 

Habitat Selection by a Freshwater Mussel: An Experimental Test 205 

R. VITTURI & E. CATALANO 

Spermatocyte Chromosomes and Nucleolus Organizer Regions (NORs) in Tri- 

colia Speciosa (Mühlfeld, 1824) (Prosobranchia, Archaeogastropoda) 21 1 

A. STAIKOU & M. LAZARIDOU-DIMITRIADOU 

Feeding Experiments on and Energy Flux in a Natural Population of the Ed- 
ible Snail Helix Lucorum L. (Gastropoda: Pulmonata: Stylommatophora) in 
Greece 217 



MCZ 
VOL 31, NO. 2 LIBRARY 1990 

JUN 6 1990 

HARVARD 
UNIVERSITY 



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 

Nineteenth Street and the Parkway 

Philadelphia, Pennsylvania 19103, U.S.A. 



EUGENE COAN 

California Academy of Sciences 

San Francisco, CA 



Co-Editors: 



Assistant Managing Editor: 

CARYL HESTERMAN 

Associate Editors: 



CAROL JONES 
Vasser College 
Poughkeepsie, NY 



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, President 
Museum of Comparative Zoology 
Cambridge, Massachusetts 

JOHN BURCH, Vice-President 

MELBOURNE R. CARRIKER 
University of Delaware, Lewes 

GEORGE M. DAVIS 
Secretary and Treasurer 

CAROLE S. HICKMAN 
University of California, Berkeley 



JAMES NYBAKKEN, President-Elect 
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 MALACOLOGICA Treasurer, UNITAS MALACOLOGICA 

Rijksmuseum van Natuurlijke Koninklijk Belgisch Instituut 

Historie voor Natuurwetenschappen 

Leiden, Netherlands Brüssel, Belgium 



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

ELMER G. BERRY, 
Germantown, Maryland 



Emeritus Members 

ROBERT ROBERTSON 

The Academy of Natural Sciences 

Philadelphia, Pennsylvania 

NORMAN F. SOHL 
U.S. Geological Survey 
Reston, Virginia 



Copyright & 1990 by the Institute of Malacology 



1990 
EDITORIAL BOARD 



J. A. ALLEN 

Marine Biological Station 

Millport. United Kingdom 

E. E. BINDER 

Muséum d'Histoire Naturelle 

Genève. Switzerland 

A. J. CAIN 

University of Liverpool 
United Kingdom 

P. CALOW 

University of Sheffield 
United Kingdom 

A. H. CLARKE, Jr. 
Portland, Texas. U.S.A. 

B. С CLARKE 
University of Nottingham 
United Kingdom 

R. DILLON 

College of Charleston 

SC. U.S.A. 

С J. DUNCAN 
University of Liverpool 
United Kingdom 

E. FISCHER-PIETTE 

Museum National d'Histoire Naturelle 

Pahs, France 

V. FRETTER 
University of Reading 
United Kingdom 

E. GITTENBERGER 
Rijksmuseum van Natuurlijke Histohe 
Leiden, Nethehands 

F. GIUSTI 

Université di Siena, Italy 

A. N. GOLIKOV 
Zoological Institute 
Leningrad, U.S.S.R. 

S. J. GOULD 
Harvard University 
Cambridge, Mass., U.S.A. 



A. V. GROSSU 
Universitatea Bucuresti 
Romania 

T. HABE 
Tokai University 
Shimizu, Japan 

R. HANLON 

Manne Biomedical Institute 

Galveston, Texas, U.S.A. 

A. D. HARRISON 
University of Waterloo 
Ontario, Canada 

J. A. HENDRICKSON, Jr. 
Academy of Natural Sciences 
Philadelphia, PA, U.S.A. 

K. E. HOAGLAND 

Association of Systematics Collections 

Washington, DC, U.S.A. 

B. HUBENDICK 
Naturhistonska Museet 
Göteborg, Sweden 

S. HUNT 

University of Lancaster 

United Kingdom 

R. JANSSEN 

Forschungsinstitut Senckenberg, 
Frankfurt am Main, Germany 
(Federal Republic) 

R. N. KILBURN 
Natal Museum 
Pietermaritzburg, South Africa 

M. A. KLAPPENBACH 

Museo Nacional de Histoha Natural 

Montevideo, Uruguay 

J. KNUDSEN 

Zoologisk Institut & Museum 

Kobenhavn. Denmark 

A. J. KOHN 

University of Washington 

Seattle. U.S.A. 

Y. KONDO 

Bernice P. Bishop Museum 

Honolulu, Hawaii, U.S.A. 



A. LUCAS 

Faculté des Sciences 

Brest. France 



S. G. SEGERSTRÁLE 
Institute of Marine Research 
Helsinki. Finland 



C. MEIER-BROOK 
Tropenmedizinisches Institut 
Tübingen. Germany (Federal Republic) 

H. K. MIENIS 

Hebrew University of Jerusalem 

Israel 

J. E. MORTON 
The University 
Auckland. New Zealand 

J. J. MURRAY, Jr. 
University of Virginia 
Charlottesville. U.S.A. 

R. NATARAJAN 
!\Aarine Biological Station 
Porto Novo, India 

J. OKLAND 
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. 

W. F. PONDER 
Australian Museum 
Sydney 

R. D. PURCHON 

Chelsea College of Science & Technology 

London. United Kingdom 

Ql Z. Y. 

Academia Sínica 

Qingdao, People's Republic of China 

N. W. RUNHAM 

University College of North Wales 

Bangor, United Kingdom 



G. ALAN SOLEM 



F. STARMUHLNER 

Zoologisches Institut der Universität 

Wien. Au st na 

Y. I. STAROBOGATOV 
Zoological Institute 
Leningrad. U.S.S.R. 

W. STREIFE 
Université de Caen 
France 

J. STUARDO 
Universidad de Chile 
Valparaiso 



T. E. THOMPSON 



S. TILLIER 

Muséum National d'Histoire Naturelle 

Pans. France 

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

W. S. S. VAN BENTHEM JUTTING 
Domburg. Netherlands 

J. A. VAN EEDEN 
Potchefstroom University 
.South Africa 

N. H. VERDONK 
Rijksuniversiteit 
Utrecht. Netherlands 

B. R. WILSON 

Dept. Conservation and Land Management 

Netherlands. Western Australia 

H. ZEISSLER 

Leipzig. Germany (Democratic Republic) 

A. ZILCH 

Forschungsinstitut Senckenberg 
Frankfurt am Main. Germany (Federal 
Republic) 



MALACOLOGIA, 1990, 31(2); 229-236 

TIDAL MICROGROWTH BANDS IN SIPHONARIA GIGAS (GASTROPODA 
PULMONATA) FROM THE COAST OF COSTA RICA 

D. J. Crisp; J. G. Wieghell^ & С A. Richardson^ 

ABSTRACT 

Siphonaria gigas growing on the coast of Costa Rica under a semi-diurnal tidal regirne lays 
down one microgrowth band per tide. This relationship was used to measure the rate of incre- 
mental growth at the anterior and posterior margins of the shell. The growth rate was somewhat 
irregular, and the anomalies at each margin were shown probably to compensate each other. 
Barnacle cover probably reduced growth rate. An approximate curve of diameter increase 
against time, assuming the Bertalanffy equation, is given. 

Key words: Siphonaria, growth rates, microgrowth bands. 



INTRODUCTION 

Within the calcareous skeletons of many 
living marine invertebrates occupying the in- 
tertidal zone or shallow sublittoral are minute 
banding patterns, known as microgrowth 
bands. These may best be seen by making 
acetate peel replicas of a polished and etched 
section of the shell cut along the direction in 
which additional shell is laid down during its 
growth. These microgrowth bands appear as 
a series of light and dark bands when viewed 
under the microscope by transmitted light. 
The darker bands are usually narrower and 
have been termed "growth bands" while the 
lighter intervening areas were termed "growth 
Increments" (Richardson et al., 1979) al- 
though both bands and increments represent 
additions to growth of the shell. 

Such tidal banding patterns have been 
demonstrated in a variety of animal groups, 
quite independently of their phylogenetic ori- 
gins. They were first demonstrated by Evans 
(1972, 1975) in the Pacific cockle Clinocar- 
dium nuttallii, and are perhaps most clearly 
expressed in other members of the Cardiá- 
cea. Richardson and his co-workers' studies 
that underlie our present understanding of the 
endogenous and exogenous nature of tidal 
bands, their relation to other environmental 
factors, and their interaction with spring and 
neap tidal changes were carried out with the 
European cockle, Cerastoderma edule, under 



the semi-diurnal tidal regime of northwestern 
Europe (Richardson et al., 1979, 1980a, b, 
1 981 ). The few gastropods studied so far con- 
tain tidal increments in the coiled whorls of the 
shell in typical forms or, in the case of limpets, 
along the corresponding region, viz. outer 
sides of the shell (Ekaratne & Crisp, 1982, 
1984). The evidence for tidal bands in the 
primitive polyplacophoran molluscs is less 
certain, but a regular 28-day periodic series of 
patterns were observed in New Zealand chi- 
tons by Jones & Crisp (1985) suggesting a 
tidal periodicity over the 14-day lunar cycle. 
Barnacle growth, analysed by Bourget & 
Crisp (1975a, 1975b, 1985) in Balanus bal- 
anoides, also was found to show periodic 
growth with tidal banding in the shell, and sim- 
ilar banding patterns were demonstrated also 
in Elminius modestus (Crisp & Richardson, 
1975). 

Of particular interest are the marine pulmo- 
nales. Pulmonates are believed to have 
evolved air breathing from the main stock of 
marine gastropods to fit them to life on land. 
Siphonaria browses on rocks in the littoral 
zone and has evolved by convergent evolu- 
tion a shell morphology like that of archaeo- 
gastropod limpets and a similar behaviour 
pattern (Morton, 1968; Barnes, 1982). The 
question arises whether shell growth occurs 
In increments separated by tidal bands or 
whether it grows more or less continuously 
without reference to tides. 



' School of Animal Biology [now Biological Sciences], University College of North Wales, Bangor, Gwynedd, United Kingdom 

LL57 2UW 

267, Etwall Rd., Hall Green, Birmingham, United Kingdom 828 OLF 

^School of Ocean Sciences, Manne Science Laboratories, Menai Bridge, Anglesey, United Kingdom LL59 5EH. 

All correspondence and reprint requests to Dr. С A. Richardson, 



229 



230 CRISP ET AL. 

TABLE 1 . Positions of sites and conditions of growth in three groups of Siphonaria gigas. 













No. 


of days 








Band-dating 


Date of 


Date of 


between marking 


Chthamalus 


Group no. 


Coast 


technique 


marking 


collection 


and collection 


presence 


1 


P.M.E. 


file-marked 
identity no. 


5.VII.85 


17.VIII.85 




44 


absent 


2 


P.M.W. 


identity no. 
only 


5.VII.85 


17. VIII. 85 




44 


absent 


3 


P.M.W. 


identity no. 
only 


5.VII.85 


17. VIII. 85 




44 


present 



P.M.E.: Punta Mala East; P.M.W.: Punta Mala West 



MATERIALS AND METHODS 

Siphonaria gigas Sowerby, 1825, was col- 
lected from two shores on the west coast of 
Costa Rica, Punta Mala West and Punta Mala 
East, Guanacaste Province (Orlega 1985, 
1 986; Sutherland & Otega, 1 986), from mid to 
high level of the intertidal zone. The condi- 
tions of growth and site details of three groups 
of animals used in this investigation are sum- 
marised in Table 1 . The specimens of group 1 
only were "file-marked" at the growing edge 
adjacent to the rock surface at the time of 
low water on 5 July 1985, without removal 
from the rock. Simultaneously, a small plastic 
tag was fixed to the side of the shell with ar- 
aldite for individual identification. Each indi- 
vidual of groups 1, 2 and 3 were so labelled, 
but only those of group 1 were also file 
marked. A file mark in the European arche- 
gastropod limpet Patella vulgata causes a 
cleft to be formed that can be related to a 
particular growth band, giving it the relevant 
date of the edge of the shell at that point. 
Ekaratne & Crisp (1984) described alternative 
methods of "band dating" shells and found 
file marking to be one of the more reliable 
techniques. However, they noted that it usu- 
ally reduced shell growth rate for a number of 
days aftenwards so that one or two bands im- 
mediately after file marking might be lost com- 
pletely. Similarly, the file marking procedure 
was found to result in a slight growth check in 
S. gigas, which could be seen as a weak ring 
running around the surface from the original 
file mark, and as a cleft seen in section (Fig. 
1 ). Similarly, some of the shells that had been 
simply given an identity tag also appeared to 
have been affected by the disturbance, pro- 
ducing a small cleft. Since these individuals 
had neither been removed from the rock nor 
filed, the disturbance was minimal and in 



some individuals it was not possible to identify 
such a cleft so that the bands could not be 
dated. 

After having been collected on 17 August 
1 985, the animals were immediately killed and 
the tissues removed from the shell. On arrival 
in the United Kingdom, any adherent barna- 
cles or debris were removed from the outside 
of the shell, the tag was removed, the shells 
scrubbed, dried and labelled. The identity 
number was written in indelible ink on the in- 
side of the shell, and any external ridge as- 
sociated with the file mark or attachment of the 
tag was also outlined with an arrow pointing to 
it (Fig. 2). Thus, the disturbance mark or cleft 
in the acetate peel could be related to the 
appropriate band seen in section. Each shell 
was embeddded in "metaset" resin, left for at 
least 1 5 h to harden, and cut by hacksaw along 
its maximum diameter. It was smoothed and 
polished as recommended by Richardson et 
al. (1979) using a series of increasingly fine 
abrasives (340, 1 20 wet and dry trimite paper), 
and polished for 30 seconds on cloth soaked 
in household metal polish "Brasso." It was 
washed in mild detergent and finally etched for 
20 minutes in a 1% "Decal," a formic-acid- 
based histological decalcifying fluid. After a 
further rinse in distilled water, it was air dried 
for 2-3 hours and the section was ready for 
replication. The appropriate size of acetate 
sheet (replicating material) was cut out, wetted 
briefly with ethyl acetate and laid on the sec- 
tion with air bubbles eliminated as far as pos- 
sible. The section and replica were placed un- 
der a plastic box to reduce the rate of 
evaporation of ethyl acetate. After at least 15 
minutes the replica was peeled oft and kept flat 
by holding between coverslip and microscope 
slide. Peels are best viewed in a low power 
phase contrast microscope in air, not in mount- 
ing medium. 



TIDAL BANDS IN SIPHONARIA 



231 




FIG. 1 . An enlarged photograph of the aceiate peel replica of Siphonaria gigas in the region of growth. GC: 
cleft indicating growth check. Parallel dark lines indicate tidal bands seperated by increments. The growth 
bands are superimposed on light and dark patches caused by vaned orientations of crystallites which are 
generally orthogonal to the direction of the growth bands and increments. 



Counting Bands 

Where possible the cleft or growth check 
clearly associated with a file mark or thought 
to be caused by disturbance through tagging 
was identified. The first band at this mark was 
taken as the datum for counting the number of 
increments between the check and shell 
edge. As can be seen in Figure 2, growth is 
not symmetrical around the shell, but the an- 
terior end becomes steeper than the poste- 
rior, as in most limpets. Thus, any section in 
the anterior-posterior plane exposes two 
growth regions, the anterior being shorter 
than the posterior. Assuming that the shell in- 
creases all round by concurrent increments, 
the band width should be shorter along the 
anterior half but the number should be the 
same. Band counts were made from the 
growth checks to the anterior (A) and poste- 
rior (P) margins of the shell, and each count 
was repeated. From the band counts and as- 
sociated statistical tests we sought the an- 
swer to the following questions. 

1 . Did the anterior and posterior profiles 
manifest the same number of bands? 

2. Were the bands laid down at tidal 
intervals? 



3. Were growth rates influenced by 
spring or neap tidal periods? 

4. Did barnacle cover influence growth 
rate? 

5. Did locality influence growth rate? 

Measurements 

Before embedding the shells in resin, each 
shell was scrubbed clean, dried and weighed 
to the nearest 10 mg (W) and its longer diam- 
eter (D) and height (H) measured using ver- 
nier calipers within 0.1 mm. The total growth 
between the disturbance mark and the ante- 
rior and posterior margins was measured us- 
ing a calibrated eyepiece graticule to an ac- 
curacy of ±1%. 

Tidal data 

The Tidal Institute at Bidston kindly sup- 
plied tidal data for the Standard Port of Pan- 
ama (Balboa) for July 1985. A normal semi- 
diurnal pattern of tides operates at Punta 
Mala, with a maximum range of 5 m at springs 
and a minimum of 2 m at neaps. There were 
83 tidal emersions between 5 July and 1 7 Au- 
gust 1985. 



232 



CRISP ETAL. 





FIG. 2. (a) Side view of Siphonaria gigas. Arrow 
points to disturbance nnark. NG: new growth after 
disturbance, t: marking tag. (b) View of the interior 
of the shell showing Identity (no. 3) and direction of 
saw cut along the anterior (A) and posterior (P) 
direction passing through the apex (black dot). 

RESULTS 

1 . Shell Measurements 

A comparison of the relationships between 
shell weight, shell diameter and shell height 
revealed no significant difference among the 
three groups of shells described in Table 1 . In 
all cases, volume calculated as V = 7tD^H/12 
rose with the weight (W), not significantly de- 
viating from isometry (Table 2). The mean val- 
ues of weight/volume was 0.9248 ± 0.0152, 
but shell shape changed significantly with in- 
crease in size, as shown in Table 2. The av- 
erage angle subtended by the shell to the hor- 
izontal (H) was obtained as = Tan ^ (2H/D), 
and Tan в increased with size as expected, 
the shell becoming taller. The mean value of в 
over all individuals was 36.6° with 95% confi- 
dence limits 35.7°-37.5°. 

2. Microgrowth Bands 

In many shells it was far from easy to iden- 
tify the growth check with certainty. Only in 
those shells where the distance measured 



from the external ridge to the edge of the 
growing tip corresponded with the distance of 
the cleft in the replica to the tip of the replica 
were the observations on numbers of bands 
included. The number of specimens giving 
countable bands and fulfilling the above crite- 
ria were six with file marks from Punta Mala 
East, four from Punta Mala West in the non 
barnacled area, and four from Punta Mala 
West in the barnacled area. The total number 
was thus 14, which was sufficient to deter- 
mine the periodicity of the banding, but insuf- 
ficient to validate such other questions as dif- 
ferences between sites, influence of barnacle 
settlement, and effect of spring and neap tidal 
periods. 

Counts of Microgrowth Bands 

The bands of the 14 selected shells were 
each counted along the anterior and posterior 
margins twice, except for two shells where the 
posterior margin was not accurately count- 
able (Table 3). 

If from all values for first readings (Count 1 ) 
are subtracted those of Count 2 an estimate 
of the reliability of counting can be made. 
There are 26 values, and the matched pairs t 
test for 25 degrees of freedom gives t = 0.64 
(N.S.). There is therefore no significant differ- 
ence between the two counts (p = 0.53). 

The standard error of the difference be- 
tween count 1 and Count 2 is 1 .223 so that 
any count based on the mean of Count 1 and 
2 can be relied upon to have a standard error 
of only ± 0.864 and confidence limits ± 3.7 of 
the average count observed, when p = 0.05. 

Similar "matched pairs" t test was applied 
to the 12 average count differences between 
the counts at the anterior and posterior mar- 
gins. The probability that counts at both mar- 
gins could have come from the same popula- 
tion of values was 0.74, p = 0.45, showing 
that the results were not significantly different 
and that each margin could be regarded as a 
replicate count. When all 52 observations 
were assembled they gave a mean band 
number of 81.04 with 5% confidence limits 
lying between 80.31 and 81.77. This result 
can be compared with the theoretical value 
for one band per tide of 83.00, or of one band 
per day of 44.00. Though significantly less 
than 83.00 (t = 5.35, p = 0.0001) the defi- 
ciency from an exact tidal periodicity is only 
2.4%. This is of the magnitude to be expected 
as a result of the disturbance caused by the 



TIDAL BANDS IN SIPHONARIA 



233 



TABLE 2. Changes in shell characteristics with size (33 individuals). 





Derived formula 




Students t value 






for 33 


Expected index 


for deviation 




Regression 


individuals 


for isometry 


from expectation 


Significance 


Log V on Log W 


1 .09W° ^^ 


1.000 


-1.29 


Not Sig. 


Log H on Log V 


7.81 V° ^2 


0.333 


+ 5.49 


Sig. 0.0001 


Log H on Log W 


8.11W°^' 


0.333 


+ 4.60 


Sig. 0.0001 


Log D on Log V 


22.10V°23 


0.333 


5.48 


Sig. ■ 0.0001 


Log D on Log W 


22.75W°2e 


0.333 


-4.10 


Sig. 0.0001 


Log H on Log D 


01-I7D13603 


1.000 


+ 4.09 


Sig. -0.0001 


Tan H on W 


0.660 + 0.0537W 


0.000 


t 5.62 


Sig. < 0.0001 



TABLE 3. Band counts in the anterior and posterior margins of shells of Siphonaria gigas specimens 
grown in two natural environments. 







Number 


of bands counted 


Number 


of bands counted 


Site and 


Specimen 
number 


in 


A' 


margin 


in 


■P' 


margin 


conditions 


count 1 




count II 


count 1 




count II 


P.M.E. no 


G 


82 




80 


80 




81 


Chthamalus 


E 


81 




82 


82 




84 




В 


83 




82 


82 




79 







80 




80 


79 




80 




С 


82 




81 


78 




79 




R 


83 




82 


83 




83 


P.M.E. no 


4 


78 




78 


— 




— 


Chthamalus 


5 


80 




80 


77 




79 




6 


82 




83 


80 




80 




8 


81 




81 


80 




81 


P.M.W. 


T 


79 




81 


83 




82 


heavy 


S 


81 




82 


82 




82 


Chthamalus 


N 


82 




82 


84 




83 




F 


81 




82 


— 




— 


('A') Anterior and (P) 


Posterior margins 















date marking and is clearly not daily but tidal 
banding as with many other marine molluscs. 

Growth Rates 

The longer profile at the posterior end of the 
shell implies a greater rate of growth than at 
the shorter anterior profile. The average total 
increment over 83 tidal cycles was indeed 
slightly higher at the postenor end 1.46 ± 
0.13 mm than at the anterior (1.41 ± 0.12 
mm) but not significantly so. 

The two sites without barnacles present 
gave growth rates that did not differ signifi- 
cantly, showing increments in length of each 
margin at Punta Mala East of 0.0339 and at 
Punta Mala West of 0.0401 mm day '. At 
Punta Mala West in the presence of Chtha- 
malus the increment in length averaged at 
0.0305 mm day \ When the growth rates at 



N 


S 


N 


S 
1 


N 




□ DO° 


дДД^ 


,.лд--'''''' 


□ o° 


дД '' 






aa° i/. 


лд лД '' 








a^'' 











20 40 60 80 

NUMBER OF TIDES 

FIG. 3. Details of growth of a single individual Si- 
phonaha gigas with very clear microgrowth lines. At 
anterior border, (A) and at postenor border (n)- N: 
neap tides, S: spring tides. 

the two Chthatvalus-hee shores were com- 
bined and averaged (0.036 mm day ^) they 
were not significantly higher than that at 



234 



CRISP ET AL 



s N s 



20 40 60 

NUMBER OF TIDES 



FIG. 4. Sum of growth at anterior and posterior bor- 
ders of the same individual (Fig. 3) showing decel- 
eration of growth rate with age. N: neap tide, S; 
spring tide. 




12 

AGE IN MONTHS 



FIG. 5. Growth curve of a sample of four Siphonaria 
gigas in barnacled area based on tidal bands. L = 
L X (1 - e '^'), L y- = 27.2 mm, К = 0.00206 
day V Rate at 17 mm = 0.047 mm day ~^ 



Punta Mala West where Chthamalus was 
present (t - 1.146, p = 0.25). Similarly, a 
comparison of Punta Mala West shores with 
and without Chthamalus gave t = 1 .37, p = 
0.20, again a figure not usually regarded as 
significant. However, as measured, the 
growth rate in the presence of Chthamalus 
appears to have been reduced by 24%. 

It should be noted that the growth rates of 
each side of the shell when added together 
and adjusted for shell slope (Ekaratne & 
Crisp, 1984) give the growth rate of the shell 
in height or diameter, which are the measure- 
ments usually quoted. However, a detailed 
measurement of both borders, anterior (A) 
and posterior (P) of shell N, giving the incre- 
ment over each of two tidal cycles from the 
shell edge to 83 bands behind, as reproduced 
in Figure 3, shows that growth is far from uni- 
form. It will be seen that both borders give 
sharp increases in length, and then slow 
down. Although A and P borders are coarsely 
correlated positively, since both are growing 
their random fluctuations appear to take place 
independently and without any common ref- 
erence to the tidal cycle. Furthermore, if the 
increments of the A and P borders are 
summed and A + P is plotted against the 
number of the tidal event (Fig. 4), the resulting 
plot appears more regular. In order to test the 
possibility of compensatory growth, the incre- 
ments at each border were listed and the 
mean increment subtracted to give the esti- 
mated acceleration or deceleration of growth 
for that tidal cycle at each border. These 



anomalies were then regressed against each 
other. They gave a negative correlation coef- 
ficient of 0.281 for 39 degrees of freedom and 
a probability of random variation of only 0.076 
in support of compensatory growth. Thus, it 
seems likely that the apparently random fluc- 
tuations at the A and P borders are not en- 
tirely random, but negatively correlated. 
When one border grows rapidly, growth at the 
other border is suppressed so that the total 
growth is more regular at either border. After 
such an episode, the roles reverse and the 
other border catches up. A similar compensa- 
tion mechanism was noted by Chsp & Patel 
(1967) in regard to the growth of the lateral 
plates of the barnacle Elminius modestus. 

The general form of the growth curve, if the 
irregularities are ignored, is asymptotic, prob- 
ably close to the Bertalanffy model. However, 
if an attempt is made to determine the con- 
straints of the Bertalanffy equation using the 
plot of dL/dt against size L (see Crisp, 1985), 
these irregularities make the differentiation of 
L by t almost impossible. By using two values 
of dL/dt from the sum A + P = L over the 
whole 83 increments for the largest and 
smallest shells, we obtained a not very ap- 
proximate equation for Siphonaria gigas 
growth in an area with barnacles present (Fig. 
5). By measuring the average angles of the 
anterior (A) and posterior (P) margins (d) the 
sum of the growth at each has been con- 
verted to diameter (D) increase through the 
relation: 

dD = (dA + dP) Cos Й 



TIDAL BANDS IN SIPHONARIA 



235 



where 6 = Tan ^ (2H/D) and H, its mean 
value, was 36.6°, Cos в = 0.803. 



DISCUSSION 

Microgrowth bands with a tidal periodicity 
have been established in certain barnacles 
(Bourget & Crisp, 1975a, b; Crisp & Richard- 
son, 1975), bivalves (Evans, 1972, 1975; 
Richardson et al., 1979, 1980a, b, 1981; 
Richardson, 1987), and probably in Poly- 
placophora (Jones & Crisp, 1985). All these 
are marine animals inhabiting the intertidal 
zone. Crisp (1989), reviewing the phenome- 
non, gave various lines of evidence to sug- 
gest that harder and more perfectly crystalline 
parts of the shell comprised the bands and 
that these formed when the body fluids were 
temporahly at a lower pH due to an accumu- 
lation of carbon dioxide and perhaps organic 
acids duhng emersion. All shell-secreting in- 
vertebrates exposed to the air and closed 
temporarily to avoid water loss, would be 
likely to experience acidosis and thus would 
slow down or prevent secretion of calcium 
carbonate. 

The siphonarian gastropods differ from all 
the above groups in being regarded as be- 
longing to a group, the subclass Pulmonata, 
superorder Basommatophora, that has be- 
come adapted to terrestrial life. Typically the 
mantle cavity has reduced external access by 
a narrow pore, its vascularised roof functions 
as a lung, the animal has lost the ctenidia and 
operculum, and it lays a shelled egg. How- 
ever, Siphonaria itself is only partially modi- 
fied. It retains or re-develops aquatic respira- 
tion through the siphon situated on the right 
side, it has secondary branchial lamellae on 
the roof or the mantle cavity and has retained 
a pelagic larval stage. The strong marine af- 
finity has led, in the past, to the Siphonariidae 
being regarded as evolved from marine 
opisthobranchs and classified as a family of 
the tectibranchs. 

Whatever the origin of Siphonaria, their pa- 
telloid form and adherent physiology (Morton, 
1968) are so closely similar to those of the 
patelloid archegastropods that the presence 
of microgrowth lines in the shell of Siphonaria 
are likely to have been produced by the same 
factors as in Patella. The need to retain water 
when closely adhering to the rock and the 
consequent absence of respiratory exchange 
at the time of low water would similarly give 
rise to a fall in pH since there is then no ef- 



fective air breathing mechanism at work. 
Thus, it is not surprising that they too should 
lay down shell bands in synchrony with tidal 
emersion. 



ACKNOWLEDGEMENTS 

We thank Dr. Sonia Ortega for supplying 
the material on which this study was based 
together with the environmental information 
given in Table 1 . One of us (D.J.C.) wishes to 
thank the Leverhulme Trust, which provided 
him with financial support duhng the course of 
this work. Dr. N. W. Runham read and im- 
proved the text. 



LITERATURE CITED 

BARNES, R. D. 1982. Invertebrate zoology. Holt- 
Saunders Japan, Tokyo, 1089 pp. 

BOURGET, E. & D. J. CRISP, 1975a. Factors af- 
fecting deposition of the shell in Balanus bal- 
anoldes (L). Journal oí the Manne Biological As- 
sociation of the United Kingdom, 55, 231-248. 

BOURGET. E. & D. J. CRISP, 1975b. An analysis 
of the growth bands and ndges of barnacle shell 
plates. Journal of the Marine Biological Associa- 
tion of the United Kingdom, 55, 439-461 . 

CRISP, D. J. 1985. Energy flow measurements. In: 
Methods for the Study of Marine Benthos (Eds. 
N. A. HOLME & A. D. MCINTYRE). Pp. 284-372 
Oxford, Blackwell. 

CRISP, D, J, 1989. Tidally deposited bands in 
shells of barnacles and molluscs. In: Biomineral- 
isation (Ed. R. CRICK). New York, Plenum Press. 

CRISP, D. J. & E, BOURGET, 1985. Growth in bar- 
nacles. Advances in Marine Biology, 22, 199- 
244. 

CRISP, D. J. & B. PATEL, 1967. The influence of 
surface contour on the shapes of barnacles. Pro- 
ceedings of the Symposium of Crustacea. Marine 
Biological Association of India. Part II, 612-629. 

CRISP, D. J. & С A. RICHARDSON, 1975. Tidally 
produced internal bands in the shell of Elminius 
modestas. Mahne Biology, 33, 155-160. 

EKARATNE, S. U. K. & D, J. CRISP, 1982. Tidal 
micro-growth bands in intertidal gastropods, with 
an evaluation of band-dating techniques. Pro- 
ceedings of the Royal Society of London, B, 21 4, 
305-323. 

EKARATNE, S. U. K. & D. J. CRISP, 1984. Sea- 
sonal growth studies of intertidal gastropods from 
shell microgrowth band measurements, including 
a comparison with alternative methods. Journal 
of the Manne Biological Association of the Untied 
Kingdom. 64, 183-210. 

EVANS, J. W. 1972. Tidal growth increments in the 
cockle Clinocardium nuttalli. Science, 176, 416- 
417. 



236 



CRISP ETAL. 



EVANS. J. W. 1975. Growth and micromorphology 
of two bivalves exhibiting non-daily growth lines. 
In: Growth rhythms and the History of the Earth s 
rotation (Eds. G. D. ROSENBERG & S. K. RUN- 
CORN). Pp. 119-134. London, John Wiley & 
Sons. 

JONES. P. & M. CRISP. 1985. Microgrowth bands 
in chitons: evidence of tidal periodicity. Journal of 
Moiluscan Studies. 51. 133-137. 

MORTON. J. E. 1968. Molluscs. Hutchinson Uni- 
versity Library London. 244 pp. 

ORTEGA. S. 1985. Competitive interactions among 
tropical intertidal limpets. Journal of Expenmen- 
tal Manne Biology and Ecology. 90, 1 1-25. 

ORTEGA, S. 1986. Fish prédation on gastropods 
on the Pacific coast of Costa Rica. Journal of 
Expenmental Manne Biology and Ecology. 97, 
181-191. 

RICHARDSON. С A. 1987. Tidal bands in the shell 
of the clam Tapes philippinarum. (Adams & 
Reeve, 1850). Proceedings of the Royal Society 
of London B. 230, 367-387. 

RICHARDSON. С A.. D. J. CRISP & N. W. RUN- 
HAM, 1979. Tidally deposited growth bands in 



the shell of the common cockle, Cerastoderma 
edule (L). Malacologia. 5, 277-290. 

RICHARDSON, С A., D. J. CRISP & N. W. RUN- 
HAM, 1980a. Factors influencing shell growth in 
Cerastoderma edule. Proceedings of the Royal 
Society of London B. 210. 513-531. 

RICHARDSON, С A., D. J. CRISP & N. W. RUN- 
HAM, 1980b. An endogenous rhythm in shell 
deposition in Cerastoderma edule. Journal of the 
Manne Biological Association of the United King- 
dom. 60. 991-1004. 

RICHARDSON, С A., D. J. CRISP & N. W. RUN- 
HAM. 1981. Factors influencing shell depostion 
during a tidal cycle in the intertidal bivalve Ceras- 
toderma edule. Journal of the Mahne Biological 
Association of the United Kingdom. 61 , 465-476. 

SUTHERLAND, J. P. & S. ORTEGA, 1986. Com- 
petition conditional on recruitment and temporary 
escape from predators on a tropical rocky shore. 
Journal of Expenmental Marine Biology and 
Ecology, 95, 155-166. 



Revised Ms. accepted 21 September 1989 



MALACOLOGIA, 1990, 31(2): 237-248 

THE NUMBERS OF FRESHWATER GASTROPODS ON PACIFIC ISLANDS AND 
THE THEORY OF ISLAND BIOGEOGRAPHY 

Alison Haynes 

School of Pure and Applied Sciences, University of tfie South Pacific. P.O. Box 1 168 

Suva, FIJI. 

ABSTRACT 

The freshwater gastropod fauna of the Pacific islands of Beqa, Vanuabalavu, Waya, Roturлa 
(Fiji), Upolu, Savai'i, Tutuila (Samoa), Tongatapu, Vava'u (Tonga), Rarotonga (Cook Islands), 
New Georgia (Solomon Islands), Guam, Truk and Ponepe (Micronesia) is descnbed. Thirty eight 
species were found; 26 species belonged to the Nentidae, 10 to Thiandae, and one each to 
Assimineidae and Planorbidae. Using multiple regression analysis, the numbers of species on 
these and 1 1 other Pacific islands were shown to be correlated with the water area on the island 
and the distance the island was from a source of freshwater gastropods (accounting for 92% of 
the variation). Distance by itself was not a significant contributor. Islands with a small area of 
water showed a steeper species-water area curve, and the number of species on these islands 
was more correlated with distance than to water area. This was probably due to a higher 
extinction rate brought about by the drying up of the limited number of habitats. 

Key words: freshwater, gastropods. Pacific islands, island biogeography. 



INTRODUCTION 



METHODS 



Fauna! studies of angiosperms, birds and 
land snails in the Pacific have documented 
the ranges and distributions of the species 
in these taxa and have revealed examples 
of endemism and of species radiation (Car- 
Iquist, 1974; Diamond, 1984; Solem, 1959). 
These studies have also been used in discus- 
sions of the theory of island biogeography de- 
veloped by MacArthur & Wilson (1967). This 
theory suggests that because the immigration 
rate to near islands is greater than that to 
more distant islands and because the extinc- 
tion rate is greater on smaller islands than on 
larger islands, the equilibrium number of spe- 
cies tends to increase with island area. In the 
past, freshwater snail diversity has been dis- 
cussed in relation to this theory, with lakes 
and ponds being considered as islands of wa- 
ter isolated by land barriers (Lassen, 1975; 
Aho, 1984). 

The aims of this work were to establish 
what species of freshwater gastropods are 
present on Pacific islands and to find if the 
island faunas, some of which had already 
been described (Haynes, 1985, 1988a; Star- 
mühlner, 1976), supported the theory of is- 
land biogeography. 



Freshwater Gastropod Survey 

From 1983 to 1987, freshwater gastropods 
were collected from the islands of Beqa, Ro- 
tuma, Vanuabalavu, Waya (Fiji); Guam; Truk 
(Federated States of Micronesia); Savai'i 
(Western Samoa); New Georgia (Solomon 
Islands); Rarotonga (Cook Islands) (Fig 1). 
The fauna of these islands is described for the 
first time. Collections were also made from 
Ponepe (Federated States of Micronesia); 
Upolu (Western Somoa); Tutuila (American 
Samoa); Tongatapu, Vava'u (Tonga) (Table 
1). All islands are within the tropics. Guam is 
the most northerly at 14 N and Rarotonga is 
the most southerly at 22 'S. 

Freshwater gastropods were collected by 
hand from rocks, boulders and vegetation or 
were sieved with a 1 mm mesh from gravel 
and mud from streams, rivers and pools. 
Sampling took place both near the coast and 
inland to ensure that the gastropods found 
were representative of the whole fauna. Each 
site was searched for at least an hour, and all 
collections were made when the volume of 
water flowing in each stream was low to nor- 
mal. 



237 



238 



HAYNES 



170 



180° 



170° 



KAUAI» , 
160' 



20 



»С 



V, 



GEORGIA^ Ч 

GUADALCANAL«^ 



Fl 


л ISLANDS 


в 


Beqa 


G 


Gau 


К 


Kadavu 





Ovalau 


т 


Taveuni 


V 


Vanuabalavu 


W 


Waya 



Ч 

\ 

EFATE. 


■ROTUMA 


«SAVAII 
UPOLU" -TUTUILA 


VANUA LEVU - -r 

W.- ^^ ' V 
VITI LEVU^-Og 


■ VAVAU 


Ï<^' 






-TONGATAPU 



TAHITI^ 

20° 



•RAROTONGA 



FIG. 1 . Pacific islands from which collections of freshwater gastropods have been made. 



Identification of the snails followed Riech 
(1937), Starmühlner (1970, 1976), and 
Haynes (1984). 

Water temperature was recorded, and wa- 
ter samples were collected on New Georgia, 
Upolu, Savai'i, Tutuila, Tongatapu and Vav'u. 
These were analysed for pH, total ions (|xs 
cm ^) and hardness mg CaCOg 1 ^) by the 
Institute of Natural Resources, University of 
the South Pacific. Some collections were 
made on islands that were visited not prima- 
rily for collecting gastropods; on these islands 
no water samples were taken. 

All gastropod collections are housed in the 
Biology Department, University of the South 
Pacific. 

Island Biogeography 

Data for the 14 islands investigated are pre- 
sented in Table 2, along with data already 
published for other Pacific islands. The is- 
lands previously investigated are Viti Levu 
(Fiji) (Haynes, 1985); Vanua Levu, Ovalau, 
Gau, Kadavu, Taveuni (Fiji) (Haynes, 1988a); 
Guadalcanal (Solomon Island), Efate (Van- 



uatu), Tahiti (Starmühlner, 1976); New Cale- 
donia (Starmühlner, 1970); and Kauai (Ha- 
waii) (Burch & Patterson, 1971; Macioiek, 
1978). 

Stream length was estimated by measuring 
the length of all streams and rivers on 1 : 
50,000 or 1 :25,000 government maps of the 
islands. The water area was estimated by 
multiplying the stream length by a mean river 
or stream width of 10-50 m (depending on 
the island size) and by adding the area of 
standing water to it. 

The large, geologically old islands of New 
Guinea, New Caledonia and Viti Levu were 
considered to be the most likely sources of 
freshwater immigrants to the islands, so that 
the distances in Table 2 were measured from 
the nearest of these three islands to the island 
in question. The three large islands together 
with nearby islands form three generally ac- 
cepted biogeographical subrogions of the Pa- 
cific islands (Thome, 1963). The source is- 
lands possessed all freshwater gastropod 
species found on the smaller islands in their 
regions, with the exception of endemic spe- 
cies. Apart from Kauai (Hawaii), the endemics 



GASTROPODS ON PACIFIC ISLANDS 



239 



TABLE 1 . Study Sites 



Micronesia 

1. GUAM. Largest island in Micronesia. Formed from the union of two volcanoes. Yling River, Cetti and 
Asafines streams were sampled. 

2. TRUK (Moen). Moen is one of the many islands in the Trul< Lagoon. Winchen River and several 
small streams near the Continental Hotel were sampled. 

3. PONEPE. A rugged island with high rainfall. Nanepil, Lehnmesi and Pilenkiepu rivers and Enipas 
Stream were sampled. The collections were made by John Macioiek and John Ford (Macioiek & 
Ford, 1987), who assisted the author with collections on Guam and Truk. 

Solomon Islands 

4. NEW GEORGIA. A high volcanic island. Sampled along the length of Puha and Borora rivers. 

Western Samoa 

5. SAVAI A. Streams are confined to the south coast because of extensive lava fields on the north 
coast. Latolo Plantation Stream, Sili Village Stream, Mata'avai Pool, Asago Spring, and Sapavai'i 
Water Hole were sampled. 

6. UPOLU. A high volcanic island. Sampled Fallefa Falls, Le Mafa Pass Stream, Mulivai Stream, and 
along the Vaisigano River. 

American Samoa 

7. TUTUILA. Volcanic with short streams. Sampled Alofau, Lemafa Saddle and Le'ele streams, and 
Pala Lagoon. 

Tonga 

8. TONGATAPU. A coral island with no running water. Sampled coastal and inland ponds. 

9. VAVA'U. An elevated limestone cluster with no running water. Sampled pools and Lake Tuanuku. 

Cook Islands 

10. RAROTONGA. The only true volcanic island in the Cook Islands. Sampled Avatiu, Vaimanga and 
Avana streams and taro patches. (Lower courses of all streams were dry in September 1983.) 

Fiji 

1 1 . BEQA. 14 km offshore from the main island of Viti Levu. Sampled the length of the stream at 
Naceva and in Naduruvesi Creek. 

12. WAYA. In the Yasawa group. Sampled the two streams in the Yolobe area. 

13. VANUABALAVU. Largest island in the northern Lau group. Northern part uplifted coral, southern part 
volcanic. Sampled the two streams near Lomolomo. 

14. ROTUMA. An isolated volcanic island 500 km north of Viti Levu. The rock is porous, and there are 
no permanent streams. Wells and taro patches were sampled. 



were Fluviopupa brevior on Efate and Mel- 
anoides paxa and Melanoides peregrina on 
Upolu. New Caledonia, the source island for 
Efate, has three species of Fluviopupa that 
could have given rise to Fluviopupa brevior. 
Melanoides is a genus that shows much vari- 
ation within species, and the isolation on Upolu 
of one or more of the four Melanoides species 
from the source island Viti Levu could have 
given rise to Upolu's two endemic species. 

The freshwater gastropods on Kauai, like 
most taxa in the Hawaiian group, show con- 
siderable speciation. It has eight endemic 



freshwater gastropod species. Four of these, 
Neritina granosa Sowerby, N. vespertina 
Sowerby, Clithon cariosus (Wood), С neglec- 
tus (Pease), probably arose from species ar- 
riving from Southeast Asia or New Guinea. 
The four Lymnaeidae endemics (Erinna new- 
combi, E. aulacospira, Pseudisidora rubella 
and P. producta) probably had their origins in 
America, Melanoides tuberculata, Tarebia 
granifera (found elsewhere on Pacific 
islands), and Ferrissia sharpi probably arrived 
accidentally in recent times whereas Galba 
viridis was introduced from Asia about 1890 



240 



HAYNES 



TABLE 2. The 25 Pacific islands arranged according to area with the data used in multiple regression 

analysis. 













Stream 


Water 




No. of 


Area 


Height 


Distance 


length 


area 




species 


(km^) 


(m) 


(km) 


(km) 


(km^) 


Island 


У 


Xi 


X2 


X3 


X4 


X5 


New Caledonia 


30 


16750 


1618 


source 


3320 


166 


Viti Levu 


31 


10429 


1323 


source 


2585 


136 


Vanua Levu 


26 


5556 


1032 


60 


1230 


62 


Guadalcanal 


24 


5302 


2330 


200 


1855 


93 


Savai'i 


11 


1709 


1856 


800 


300 


16 


New Georgia 


20 


1470 


843 


200 


1080 


54 


Kauai 


12 


1432 


1598 


(6200) 


604 


83 


Upolu 


15 


1114 


1113 


840 


325 


17 


Tahiti 


15 


1042 


2241 


2440 


735 


37 


Efate 


18 


887 


646 


500 


370 


22 


Guam 


11 


541 


406 


1800 


122 


2.00 


Taveuni 


15 


470 


864 


10 


483 


11 


Kadavu 


17 


411 


838 


85 


398 


6 


Ponepe 


11 


334 


772 


1400 


270 


4 


Tongatapu 


3 


259 


19 


740 





0.25 


Gau 


16 


140 


750 


62 


197 


2 


Tutuila 


13 


137 


652 


1000 


128 


1.8 


Vava'u 


1 


118 


179 


800 





0.02 


Ovalau 


20 


101 


626 


17 


105 


1.5 


Rarotonga 


3 


67 


653 


2400 


87 


0.9 


Rotuma 


1 


47 


256 


500 





0.01 


Vanuabalavu 


4 


38 


290 


106 


10 


0.05 


Beqa 


13 


35 


439 


14 


37 


1.9 


Truk (Moen) 


4 


19 


369 


1300 


6 


0.03 


Waya 


9 


17 


580 


46 


20 


0.20 



(Burch & Patterson, 1971; Macioiek, 1978). 
Therefore, in the case of Kauai, distance from 
a source island is irrelevant. 

It was thought that little bias was introduced 
by using Starmühlner's figures for New Cale- 
donia, Guadalcanal, Efate and Tahiti. He col- 
lected from Upolu and Tutuila (Samoa) in 
1985 (Starmühlner, 1986) and reported 23 
species (one doubtful), which compares fa- 
vorably with the 22 species I found in 1987. 

Bishop Museum shell collections of fresh- 
water gastropods from Pacific islands were 
studied in 1985, and I undertook a revision of 
their nomenclature. The Bishop Museum col- 
lections, which are not extensive, contain no 
species additional to those I found. 

The data in Table 2 were the basis of mul- 
tiple regression analysis using the method de- 
scribed by Bliss (1970). The number of gas- 
tropod species on an island was used as the 
dependent variable and the other factors — is- 
land area, island height, island distance from 
the presumed source of gastropods, stream 
length and water area — were the independent 
variables. The four first independent variables 



were converted to logs whereas, for conve- 
nience, water area was first multiplied by 100 
before being converted to logs. 

The quantity of calcium ions (hardness) 
and total ions (conductivity) in the water can 
determine whether gastropods will be pre- 
sent. However, as the figures for hardness 
and conductivity for all streams and rivers 
tested (Table 4) were above 4.3 mg Ca 1 ^ + 
1.2 mg Mg 1 \ the amount that limits the 
presence of gastropods in freshwater (Aho, 
1984), they were not used in the multiple re- 
gression analysis. 



RESULTS 
Freshwater Gastropod Survey 

Thirty eight species of freshwater gastro- 
pods were collected from the 14 islands. 
Twenty six were Neritidae, 10 Thiaridae, one 
Planorbidae and one Assimineidae (Table 3). 

The species found most frequently was the 
parthenogenetic thiarid Melanoides tubercu- 



GASTROPODS ON PACIFIC ISLANDS 



241 



lata (Table 3). It was present on 1 1 of the 14 
islands investigated. This species is also 
found in East Africa, the Middle East, Asia 
and the Caribbean (Starmühlner, 1976). 

The stream-dwelling neritids, Nerltlna var- 
iegata (on 9 islands) and Septana procellana 
(on 8 islands), were the next most wide- 
spread. These were followed by the brackish- 
water gastropod Neritina turrita on 7 islands. 

Twelve of the species were present on both 
North and South Pacific islands. These were 
Melanoides tuberculata. Neritina turrita, N. 
variegata, N. pulligera, N. macgillvrayi. N. 
squamipicta, Neritodryas subsulcata. Clittion 
corona, Septaria porcellana, S. lineata, S. 
sanguisuga and Tarebia granifera. 

Although in this study Tliiara cancellata, 
Neritodryas cornea, Neritina labiosa, N. aspe- 
rulata and Clithon nucleolus were found only 
on New Georgia, the first three have been 
recorded from Papua New Guinea (Riech, 
1937; Starmühlner, 1976) and the last two 
from New Caledonia (Starmühlner, 1970). 

The only endemic species recorded were 
two thiarid species, Melanoides laxa and Mel- 
anoides peregrina on Upolu. 

Water temperature, pH, hardness and con- 
ductivity for the islands studied and results 
already published from other Pacific islands 
are given in Table 4. Gastropods were absent 
from Lake Tagimaucia, Taveuni where total 
ions (conductivity) (14-18 |jls cm ^) and 
hardness (0.8-5.0 mg Ca + Mg 1 ^) were 
low (Southern et al., 1986) (Table 4). Hard- 
ness and conductivity of other freshwaters 
were sufficient to support gastropods (Table 
4). 

Island Biogeography 

Because island area (XJ was correlated 
with stream length (X4) (r = 0.9377) and with 
water area (X5) (r = 0.8737), and water area 
(X5) was correlated with stream length (X4) (r 
= 0.9522), each made a similar contribution 
to the variation in the number of species (y). 
However, the variable water area correlated 
best with species numbers (r = 0.8412) 
(Table 2). 

Using the stepdown method of reducing the 
number of variables until only those having 
significant influence were left (Bliss, 1970), 
the best correlation obtained was with the 
vahables water area, distance from the 
source of gastropods (X3) and island height 
(X2). These vahables accounted for 93% of 
the vahation in species numbers. When is- 



land height was omitted, 92% of the correla- 
tion was still accounted for by water area and 
distance. As the contnbution of island height 
was not significant, the residuals of species 
numbers (y Y) from the equation Y - 9.898 
+ 4.9445X5 - 3.7935X3 were plotted as de- 
viations from the partial regression of species 
numbers against water area in Fig. 2. They do 
not depart much from linearity or from uniform 
scatter about the line. When distance was 
omitted, the correlation fell to 84%, showing 
that water area is the major conthbutor to the 
correlation, but distance is a significant con- 
tributor (p •; 0.001) when taken in combina- 
tion with water area. However, distance by 
itself is not significant (r = 0.3745) for 23 is- 
lands. 

When the eight small islands with least wa- 
ter area (i.e. Ovalau, Rarotonga, Tongatapu, 
Waya, Vanuabalavu, Truk, Vava'u and Ro- 
tuma) were considered separately, distance 
from source of gastropods was the largest 
contributing factor to the number of species 
per island. When combined with water area, 
the two factors contributed 91% of the corre- 
lation, whereas distance alone contributed 
81%. Species numbers were plotted against 
water area for the 25 islands in Figure 3. It is 
seen that the slope is steeper for the eight 
islands with small areas of water. 



DISCUSSION 

The parthenogenetic thiarid Melanoides tu- 
berculata, which was found on 1 1 of the 1 4 
islands, can easily be spread on plant mate- 
rial as it gives birth to live young. One speci- 
men reaching an island can start a new pop- 
ulation, and as it inhabits ponds, ditches and 
taro patches as well as streams and rivers, it 
can survive on such islands as Tongatapu, 
Vava'u and Rotuma which have no running 
water. 

The buliniform planorbid Physastra nasuta, 
which inhabits ponds as well as running wa- 
ter, was present on Tongatapu, Rarotonga 
and Tutuila. It has been found on other Pacific 
islands, such as New Caledonia (Starmü- 
hlner, 1970), Viti Levu (Haynes, 1984) and 
Vanua Levu (Haynes, 1988a). Walker (1984) 
suggested that the genus Physastra evolved 
in the Australian region and spread into 
Southeast Asia and the Pacific through New 
Guinea. Physastra nasuta vja^s collected from 
Tonga in 1832 (Solem, 1959), and it may 



242 



HAYNES 



E r^ CD 

3 ■^ LO T- 

o ^ 
CL 



CO 








>< 


r^ 


О 


СУ) 


CO 


Ч — 


CXD 




5 




in 




03 
СГ 

CD 


Ю 


О 


CO 


CO 


CO 


1 — 




•* 




CO 








O) 








с 
о 


r-- 


CO 


CO 




CD 


1Л 




о 




CD 












со 








ÛC 








3 








'to 


00 


CT) 


•^ 


> 


T — 


r- 




(0 


T— 


T — 




> 








3 








Q. 
2 








CD 


CD 


CO 


CO 


ю 






en 


CvJ 







= r-- CM CO 
2 CO LD T- 
Z5 -^ CO 



^ -^ CO ю 

о ■■- -^ ■■- 
Э ^ ^ 



i; о CO о 
g Г- 'í C\J 



M- C\J -.- 
со h- ■.- 
со Г-- 



ê^co^ 



о 



со 

о •.- 
in -^ 



е£ 



"О "^^ JD 

i? 2Í ш § 
J« < X Z 



X 



X 



X X 



XXX X X 



X X X X 



X XXX 



X X 



X X X X X 



X X 



X X 



XXX XXX X 



XXX 



X X 



,-^ та 



X 



S' ^ ^ 



'о о 



с 5 о :^ 



Ш то 

< с 

Û та 

ее та 

< та 

I- |~ 



с та 
та о 

iií о ел 7Í 



-о 

3 


s(Hii 
Gouh 

OUSS( 


о 


V) 


та -рШ 
с 5 ' — ' 


0) ,-^ 


с — :5 


с ° : — 


11 


sp/ra 
'fosa 
'ха(1 


S^ ел 3 
оЗ о t 


^^ 


та ^ -5 


clS та 


Q) ■— 




. — ' 



-Q со 



|^|^5 5555 5 



ш t 
< 5 

9та 
b с 
ее ■£ 
ш <ь 

Z ^ 



со о ^ о >, ^ 
" JD о 

w ^J ^ 



E 
со 


(Л 


■0) ф 


(Л 

CD 


с ОС 


та 


та 
Я, 

С35 


о — Sá 


та 


■0) р о. 


о 


сь 


-с .?> та 


с 


С 


■*^ ^ Z3 


:з 


та 


а> 3 сз- 


та 


ь 


Cl Cl ел 



со й 
'S та 
с У 
та о 



^^^г:гггг 



GASTROPODS ON PACIFIC ISLANDS 



243 



XXX XXX 



X X 



XX X 



XXX 



XXX 



XX XXXX XX X XX 



X X 



^ 

^"cn 



FC 



=5 tT3 

CD ii 



Ф 7 

:§ 8 
5 </) 


ata 

У) 
idema 

(Linné) 
ensis 


гг^_ 


^ -e ^ N- СП с 

:з ф тз 3 с СП 


!iec 
tilia 
toa 
nné 


■о 5 с Ü Р 1п 

3 9 2 '^ о => 




со со :£ СС О О 


~ — ' :_^ - — 


гг 


г: О CJO 



С СО Cl O 



Ф 
^^ > 

N o Ф 

■ä^ E cn^ 






(D 



СП С '^ Ï S tö 

. . . . ^ "■ " 
(JOÜO CO 



СП 

ш "5 

< со 
Q g _ 



3 с 4J g f п 

со * со с: и 
СО СО СЛ) СО 



< ^ 

_1 -С 

О. Q, 



СП о 



с 
СП 
ш 

Ш со 

< р 

Û Ь 
Ш СП СО 

= с СО 

1§^ 

< <С 



244 



HAYNES 



TABLE 4. The comparison of the water chemistry and the number of gastropods present in the streams 
of Pacific islands 





Temperature 




Conducitivity 


Hardness 


Number of 


Island 


С 


pH 


|jls cm ' 


mg CaCO^I ' 


gastropods 


Savai'i 


25 


7.1-7.2 


57.8-103.9 


11.3-26.1 


11 


Upolu 


26 


6.6 


78.5 


20.3 


15 


Tutuila 


27-30 


6.4-7.3 


146-152 


15.0-33.7 


13 


New Georgia 


25-26 


6.9-7.1 


181-183 


21.5-22.0 


20 


Vava'u 












(L. Tu'anuku) 


31 


7.5 


14040 


407 



(1 in pool) 


Tongatapu 












coastal pools 


— 


7.4 


1435-8061 


42.9-177 


3 


Ponepe' 


— 


— 


21-104 


6-46 


11 


Vit! Levu^ 


23-32 


5.0-7.5 


42.6-231 


19.5-99 


31 


Vanua Levu^ 


22-30 


6.0-7.0 


111.1-915 


36-252 


26 


Ovalau^ 


25-26 


6.7-7.0 


147.1-152.3 


56-60 


20 


L. Tagimaacia'' 


— 


5.0-6.5 


14-18 


0.8-5.0(Ca + Mg) 





streams^ 


21-22 


5.0-5.5 


36.1-66.7 


9-19.7 


15 


Kadavu^ 


25-27 


6.5-7.5 


36.1-66.7 


20-22 


17 


Gau= 


26-27 


7.0-7.7 


122-134 


52-55 


16 



1. Macioiek & Ford (1987), 2, Haynes (1985). 3, Haynes {1988a). 4. Southern et al. (1986). 



TABLE 5. Freshwater gastropod habitats on Fiji Islands and the gastropods that may inhabit them 



Habitats 



Gastropods 



1. Ponds, dalo (taro) patches, 
ditches & lakes 

2. Brackish water (shaded or 
mangrove areas) 

3. Brackish water (open areas, 
mouths of streams & rivers) 

4. Freshwater (influenced by the 
tide, lower courses of streams 
hvers) 

5. Fast flowing streams & rivers 
(substrate pebbles, stones & 
boulders) 



6. Cascades (substrate boulders 
rocks) 



Melanoides tuberculata. Physastra nasuta, Ferrissia noumeensis, 
Gyraulus montrouzieri 

Nehtina turrita. N. turtoni. N. auriculata. Clithon oualaniensis 

Neritina turrita, N. turtoni, N. auriculata. Clithon oualaniensis, С 
diadema, С. pritchardi, С. rarispina. С. spinosa. Septana linéala, 
Assiminea crosseana, Melanoides arthurii 

С pritchardi, С. diadema, S. linéala. Septana porcellana, 
Neritina squamipicta, Thiara amarula, T. bellicosa, T. scabra, T. 
terpsichore, Melanoides plicana. M. arthuni. M. aspirans 

M. tuberculata, M. lutosa. T. scabra. P. nasuta. F. noumeensis. 
Fluviopupa pupoidea. Fijidoma maculata. Neritina pulligera. N. 
petiti. N. canalis. N. percata. N. variegata. N. macgillvrayi. 
Neritodryas subsulcata. Neritilia rubida. С pritchardi. С 
olivaceus. S. porcellana. S. sanguisuga. S. suffreni. S. 
macrocephala 

S. porcellana. S. sanguisuga. S. suffreni. S. macrocephala 



have been transported to Tongatapu and 
Rarotonga on taro plants in recent times by 
man. 

The majority (26 species out of 38) of the 
snails collected were nerites (Table 3). It has 
been suggested that the brackish and fresh- 
water neritid genera, Clithon, Neritina, Neri- 
tilia. Neritodryas and Septana, evolved at dif- 
ferent times from the marine genus Nerita 



probably in the Southeast Asia region 
(Govindan & Natarajan, 1972; Starmühlner, 
1982). A few species have spread westward 
into the Indian Ocean, whereas many have 
spread eastward across the Pacific Ocean. 

In this survey, many more species of 
nerites were found in the South Pacific (25 
species) than in the North Pacific (11 spe- 



GASTROPODS ON PACIFIC ISLANDS 



245 




FIG. 2. Residual species numbers (y - Y) of fresh- 
water gastropods plotted as deviations from the 
partial regression of species numbers against water 
area. 




area (km) 



FIG. 3. The freshwater gastropod species num 
bers-water area curve for Pacific islands. B: Beqa 
E: Efate, Ga: Gau, Gd; Guadalcanal, Gm: Guam 
Kd: Kadavu, K: Kauai, O: Ovalau, NO: New Gale 
donia, NG: New Georgia, P: Ponepe, Ra 
Rarotonga, Rt: Rotuma, S: Savai'i, T: Tahiti, Tv 
Taveuni, Tg: Tongatapu, Tr: Truk, Tt: Tutuiia, U 
Upolu, Vb: Vanuabalavu, Va: Vanua Levu, Vv 
Vava'u, VL: Viti Levu, W: Waya. 



cies). All species found in the North Pacific 
were also present in the South Pacific (Table 
3). It appears that more species have moved 
south through the New Guinea-Solomon Is- 
land region than have moved north into Mi- 
cronesia. Such species as Clithon nucleolus, 
Neritina asperulata and N. labiosa malanisica 



do not appear to have dispersed further east 
than Solomon Islands and New Caledonia, 
whereas Clithon phtchardi. Septaria macro- 
cephala and S. suffreni probably arose in the 
South Pacific as they are not found as far 
north as Vanuatu and Solomon Islands (Fiq. 
1). 

Unlike land snails, which show consider- 
able speciation on Pacific Islands, e.g. 
Partula on Samoa and zonitids in Fiji (Solem, 
1959), comparatively few species of endemic 
freshwater gastropods have been found. Be- 
sides the two endemic species of Thiaridae, 
Melanoides laxa and M. peregnna collected 
from Upolu, other endemic species recorded 
on islands discussed in this paper are Fiji- 
doma macúlala (Thiaridae), Fluviopupa 
pupoidea (Hydrobiidae) (Viti Levu); an 
opisthobranch, Acochlidium sp. (Vanua 
Levu); Melanopsis frustulum, M. mariei 
(Thiaridae), Fluviopupa minor and two other 
Fluviopupa spp., Hemistomia caledonica (Hy- 
drobiidae), Physastra petiti (Planorbidae) 
(New Caledonia); Fluviopupa brevior (Efate) 
and the eight endemic species on Kauai men- 
tioned above (Mornson, 1954; Starmühlner, 
1970, 1976; Haynes, 1988b; Burch & Patter- 
son, 1971; Macioiek, 1978). 

Although man has probably helped in the 
distribution of Melanoides tuberculata and 
Physastra nasuta. which live in taro patches, 
It is unlikely that man has been responsible 
for the spread of other species to Pacific is- 
lands. Most freshwater gastropods do not live 
on vegetation but are found on the mud or 
rocks of stream or river bottoms. They are not 
favored as food and therefore the chance of 
them being spread purposely by man is small. 
Some brackish-water neritid species may 
cling to wooden boats and be carried to 
nearby islands. Other neritid and thiarid spe- 
cies may be rafted out to sea on tree trunks 
during flooding and be washed ashore at a 
river or stream mouth. However, many spe- 
cies are probably distributed from island to 
island as veliger larvae. Most neritid and 
brackish-water thiarid gastropods hatch as 
veligers. These may be swept out to sea and 
settle in brackish water at the mouth of a 
stream on another island. Ford (1979) re- 
ported long lines of young Neritina granosa 
migrating upstream on Hawaiian Islands. He 
believed that the veligers, after being swept 
down to the sea, spent some weeks in salt 
water before settling at the mouth of a stream 
and starting their migration upstream. There 
is no evidence to suggest that this occurs in 



246 



HAYNES 



all neritid species, but Nehtina. Clithon and 
Septana veligers kept in the laboratory can be 
acclimatized to sea water, and they have re- 
mained alive for 22 days without settling. This 
allows them time to be carried by currents to 
quite distant islands. However, they are more 
likely to reach and become established on is- 
lands that are near the source of the gastro- 
pod veligers. 

Island Biogeography 

According to the equilibrium theory of is- 
land biogeography (MacArthur & Wilson, 
1967), the greater the distance of an island 
from a source of colinization, the smaller the 
probability of colonization. However, if islands 
are the same distance from the source, immi- 
gration will be greater to the larger island. Iso- 
lated small populations on small islands will 
have a higher rate of extinction due to com- 
petition and population fluctuations. If further 
immigration occurs after all potential niches 
are filled, interspecific interactions will in- 
crease, and the extinction rate will increase 
and keep the species number in equilibrium. 

On the 25 Pacific islands considered, the 
total area of water was the main factor influ- 
encing the number of freshwater gastropod 
species present (explaining 84% of the vari- 
ation). Because island area and stream 
length are strongly correlated with water area, 
their influence on the number of species is 
incorporated in water area. Distance from the 
source contributes 8% to the variation in the 
number of species and unknown factors 7%. 
The contribution of height is also largely in- 
corporated in water area (r = 0.7312) be- 
cause an island with an altitude less than 300 
m usually will be without streams, and in gen- 
eral the higher an island the greater its stream 
length, water area and habitat diversity. 

The importance that distance contributes to 
species variation on small islands may be due 
to the strong possibility of the small area of 
water drying up and the consequent likelihood 
of extinction of some or all gastropods. The 
nearer such islands are to a source of gastro- 
pods the more likely immigration is to occur 
and the number of species to be restored. 
Ovalau (20 species) and Beqa (13 species), 
which are close to Viti Levu, have a relatively 
large number of species, whereas the more 
distant islands, such as Rarotonga (3 spe- 
cies), Truk (4 species) and Vanuabalavu (4 
species), have few species (Tables 2, 3). 

Freshwaters on Pacific islands can be di- 



vided into six distinct habitats: (1) ponds, taro 
patches, ditches; (2) shaded brackish water; 
(3) open brackish water; (4) freshwater influ- 
enced by the tide; (5) fast flowing streams 
and rivers; and (6) cascades (Table 5). Some 
are inhabited by only a few gastropods, and 
others are suitable for colonization by a large 
number of gastropod species. Small islands 
and islands of low elevation do not have all 
these different habitats, but those they do 
have fall into one of these categories. The 
species inhabiting the habitats are not all the 
same for each island group. 

The number of gastropod species on an is- 
land will partially depend on the number of 
each kind of habitat and their size. These are 
factors which account for some or all of the 
unknown 7% in variation of the number of 
gastropod species on islands. 

The steeper slope for islands with a small 
area of water has been observed in species- 
area curves before (Fig. 3). Williams (1981) 
gives a similar plot for birds on the Solomon 
Islands, and Lassen (1975) drew another for 
freshwater snails in small eutrophic lakes in 
Denmark. This steeper slope for smaller 
ponds Lassen (1975) explained by lower im- 
migration and an increased extinction rate 
with decreasing area. Birds carrying immi- 
grant snails are less likely to visit small ponds, 
and small ponds are more likely to freeze. 

Similarly, a steeper slope was obtained for 
Pacific islands with small water area, because 
the extinction rate increases due to ponds 
and lower courses drying up and because the 
survival rate of immigrants is low due to rela- 
tively few available habitats. 

Most investigations into which factors de- 
termine the number of species on islands 
have involved plants or birds. Johnson & 
Raven (1970) found that island area, latitude 
and soil types were important in the species 
diversity of plants on the British Isles and on 
California islands. Harris (1973), using multi- 
variate analysis, established that the vari- 
ables that contributed to the variation of num- 
bers of breeding land birds on the Galapagos 
Islands were total plants and altitude (87.7%) 
and distance (90.5%). Power (1972) found by 
multivariate analysis that the variation in the 
numbers of bird species on California islands 
was caused by tfie interaction of these vari- 
ables: numbers of native plant species and 
distance from other islands and from the 
mainland. The variation in numbers of plant 
species was mainly explained by island area 
and latitude. 



GASTROPODS ON PACIFIC ISLANDS 



247 



In this investigation, island area and height 
were important because they determine the 
diversity and size of the freshwater habitats 
available. The habitat diversity is best ex- 
pressed as water area for purposes of multi- 
ple regression analysis. Distance from a pos- 
sible source of new immigrants is also 
important in determining species numbers, 
probably because of the high rate of extinc- 
tion caused by water drying up and some- 
times by whole populations being washed 
away duhng tropical cyclones. 



ACKNOWLEDGEMENTS 

I wish to thank the University of the South 
Pacific for providing a research grant for this 
work and Dr. J. Macioiek and Mr. J. Ford for 
making available their collections of freshwa- 
ter gastropods from Ponepe. 



LITERATURE CITED 

AHO, A., 1984, Relative importance of hydrochem- 
ical and equilibrial variables on the diversity of 
freshwater gastropods in Finland. Pp. 198-206, 
in SOLEM, A. & A.C. VAN BRUGGEN, eds., 
World-wide snails: Blogeograpliical studies on 
non-marine Mollusca. Brill/Backhuys, Leiden. 

BLISS, С I., 1970, Statistics in Biology, Vol. 2. Mc- 
Graw-Hill, New York. 639 pp. 

BURCH, J. B. & С H. PATTERSON, 1971, Chro- 
mosome number of Hawaiian Lymnaeidae. Mal- 
acological Review, 4(2): 209-210. 

CARLOUIST, S., 1974, islano biology. Columbia 
University Press. New York. 266 pp. 

DIAMOND, J., 1984, Biogeographic mosiac in the 
Pacific. Biogeography of the Tropical Pacific: 
Proceedings of a Symposium. Bishop Museum 
Special Publication, 72: 1-14. 

FORD, J. I., 1979, Biology of a Hawaiian fluvial 
gastropod Neritina granosa Sowerby (Proso- 
branchia: Neritidae). M.S. Thesis, University of 
Hawaii. 

GOVINDAN, K. & R. NATARAJAN 1972, Studies 
on Neritidae (Neritacea: Prosobranchia) from 
peninsular India. Indian National Science Acad- 
emy Proceedings, Part В 225-239. 

HARRIS, M. P., 1973, The Galapagos avifauna. 
Condor, 75: 265-278. 

HAYNES, A., 1 984, Guide to the brackish and fresh 
water gastropods of Fiji. Institute of Natural Re- 
sources, University of the South Pacific. 37 pp. 

HAYNES, A., 1985, The ecology and local distribu- 
tion of non-manne aquatic gastropods in Viti 
Levu, Fiji. The Veliger, 28(2): 204-210. 



HAYNES, A., 1988a, The gastropods in the 
streams and rivers of five Fiji islands (Vanua 
Levu, Ovalau, Gau, Kadavu and Taveuni). The 
Veliger 30(4). 377-383. 

HAYNES, A., 1988b, A population of the Fijian 
freshwater thiarid gastropod Fijidoma maculata 
(Mousson). Archiv für Hydrobiologie. 1 1 3(1 ): 27- 
39. 

JOHNSON, M. P. & P. H. RAVEN, 1970, Natural 
regulation of plant species diversity. Evolutionary 
Biology. 4: 127-162. 

LASSEN, H. H., 1975, The diversity of freshwater 
snails in view of the equilibnum theory of island 
biogeography. Oecology. (Berl) 19:1-8. 

MACARTHUR, R. H. & E. O. WILSON, 1967, The 
theory of island biogeography. Princeton Univer- 
sity Press, New Jersey 203 pp. Monographs in 
Population Biology 1 . 

MACIOLEK, J. A., 1978, Shell character and habi- 
tat of nonmarine Hawaiian neritid snails. Micron- 
esica. 14 (2): 209-214. 

MACIOLEK, J. A. & J. I. FORD, 1987, Macrofauna 
and environment of the Nanpil-Kiepw river, 
Ponepe, Eastern Caroline Islands. Bulletin of 
Manne Science. 4(12): 623-632. 

MORRISON, J. P. E., 1954, The relationships of old 
and new world melai'ians. Proceedings of the 
United States National Museum, 103(3325): 
357-394. 

POWER, D. M., 1972, Numbers of bird species on 
the California Islands. Evolution, 26: 451-463. 

RIECH, е., 1937, Systematische, anatomische, 
ökologische und tiergeographische Unterschun- 
gen über die Susswassermollusken Papuasiens 
und Melanesiens. Archiv für Naturgeschichte 
(N.F.) 6(36): 40-101. 

SOLEM, A., 1959, Systematics and zoogeography 
of the land and freshwater Mollusca of the New 
Hebndes, Fieldiana: Zoology. 43: 1-359. 

SOUTHERN, W., J. ASH, J. BRODIE & P. RYAN, 
1986, The flora, fauna and water chemistry of 
Tagimaucia crater, a tropical highland lake and 
swamp in Fiji. Freshwater Biology, 16: 509- 
520. 

STARMUHLNER, f., 1970, Études hydrobi- 
ologiques en Nouvelle-Calédonie. О. RS. Т.О. M., 
Ser Hydrobiologie, 4(3/4): 3-127. 

STARMUHLNER, f., 1976, Beiträge zur Kenntnis 
der Süsswasser-Gastropoden pazifischer Inseln. 
Annalen des Naturhistorischen Museum in Wien, 
80: 473-656. 

STARMUHLNER, f., 1982, Occurence, distribution 
and geographical range of the freshwater gastro- 
pods of the Andaman Islands. Malacologia. 22 
(1/2): 455-656. 

STARMUHLNER, F., 1986, The fresh- and brack- 
ishwater gastropods of the Tongan and Samoan 
Islands. 9th International Malacological Con- 
gress, Edinburgh 12 pp. 

THORNE, R. F., 1963, Biotic distnbution patterns in 
the tropical Pacific. Pp. 311-350 in GRESSITT, 
J. L., ed., Pacific basin biogeography. Bishop 
Museum Press, Honolulu. 



248 HAYNES 

WALKER, J. C, 1984, Geographical relationships WILLIAMS, M., 1981, Island populations. Oxford 
of the buliniform planorbids of Australia. Pp. University Press, Oxford. 286 pp. 

189-197 in SOLEM, A. & A. С VAN BRUGGEN, 
World-wide Snails: Biogeographical studies on 
non-manne Mollusca. Brill/Backhuys, Leiden. Revised Ms. accepted 18 August 1989 



MALACOLOGIA. 1990, 31(2): 249-257 



ANALYSIS OF LYMNAEACEAN RELATIONSHIPS USING 
PHYLOGENETIC SYSTEMATICS 

Donald L. Swiderski 

Department of Geological Sciences. Michigan State University, East Lansing. l\/lichigan. 

U.S.A. 48824 



ABSTRACT 

Currently, evolutionary studies of lymnaeacean pulmonates are heavily dependent on a small 
number of classical morphological studies for family-level phylogenetic relationships. These 
classical studies are in general agreement on the relationships of the lymnaeacean families. 
Unfortunately, all of the previous studies infer relationships from a priori arguments for character 
evolution based on assumptions of the evolutionary or adaptive significance of the characters in 
question. Considering the widespread convergence in pulmonates, the assumptions may not be 
justified and the phylogenetic inferences derived from them are probably suspect. 

The present study employs outgroup analysis and component analysis to test the phyloge- 
netic implications of previously published character-state distributions. The purpose of this study 
is to determine whether the morphological descriptions reported in the literature support either 
the currently accepted phylogeny, or an alternative interpretation. The results of the outgroup 
analysis indicate that only a few of the characters described in the literature permit lym- 
naeaceans to be discriminated from all related pulmonates. The resuuc of the component 
analysis indicate that the few informative characters provide weak support for accepting a 
revised lymnaeacean phylogeny, but strongly support rejection of the classical interpretation. 

Key words: Lymnaeacea, phylogeny, component analysis, outgroup analysis. 



INTRODUCTION 

The pulmonate superfamily Lymnaeacea 
includes six families of freshwater snails: Chil- 
inidae, Latiidae, Acroloxidae, Physidae, Lym- 
naeidae and Planorbidae (Hubendick, 1978). 
(Although ICZN Recommendation 29A sug- 
gests superfamily names end in -oidea, -acea 
is conventional for this group and is the end- 
ing used in this paper.) The lymnaeaceans 
have a nearly global distribution (Hubendick, 
1 978) and occupy a wide variety of freshwater 
habitats (cf. Clarke, 1973). Associated with 
their large ecological range is tremendous 
morphological diversity, resulting in several 
hundred named species (cf., Clarke, 1973). 
This high level of diversify makes phyloge- 
netic studies of the Lymnaeacea difficult. 

Despite the difficulty of resolving relation- 
ships within the superfamily, interest in the 
problem persists. One motivation results from 
their ecological diversity; the Lymnaeacea are 
useful as indicators of ecological conditions, 
both in Recent (Aho, 1966; Clarke, 1979) and 
fossil (Ayyasamy et al., 1985; Good, 1987; La 
Rocque, 1966-1970) habitats. Phylogenetic 
studies may be useful in identifying traits or 



taxonomic groups associated with particular 
environments. Another motivation results 
from the role of many species as intermediate 
hosts for trematode parasites (Gomez et al., 
1986; Mandahl-Barth, 1957). Here, phyloge- 
netic studies may be relevant to analyses of 
host-parasite co-evolution. 

Interest in lymnaeacean phylogeny tends to 
focus on lower taxonomic levels: genera or 
species (e.g. Jelnes, 1986, Bulinus; Meier- 
Brook, 1983, Gyraulus). Studies at lower lev- 
els necessarily take the family-level phylog- 
eny as given. The few previous family-level 
analyses (Hubendick, 1947, 1978; Star- 
obogatov, 1967) support the phylogeny 
shown in Figure 1 . The concordance of these 
studies would normally be taken as a sign of 
reliability and robustness, but these studies 
are based on similar material and share a 
common approach. The authors argue that 
the gonad (Hubendick, 1978; Starobogatov, 
1967), prostate (Hubendick, 1947, 1978; 
Starobogatov, 1967) and male copulatory or- 
gans (Starobogatov, 1967) are more impor- 
tant than any other traits for phylogeny recon- 
struction because reproductive-tract char- 
acters are crucial to reproductive success. 



249 



250 



SWIDERSKI 




FIG. 1. Phylogenetic relationships supported by 
previously published analyses. 

Therefore, a sequence of improvements in 
these structures should reflect phylogenetic 
history. However, reproductive success de- 
pends on many factors in addition to gamete 
production and mating ability. Survival and 
the opportunity to mate depend, in part, on 
respiratory and digestive abilities. Therefore, 
respiratory and digestive structures are also 
crucial to reproductive success, and should 
not be given less weight than reproductive 
structures in phylogenetic inference. 

Harry (1964) uses Dollo parsimony to infer 
directions of character transformations from a 
phylogeny similar to Figure 1. Dollo parsi- 
mony assumes that acquisition of a new char- 
acter state is rare, but that loss of a derived 
state, reverting to a more primitive state, is 
much more frequent. Harry's results indicate 
that reproductive traits, as well as digestive 
and respiratory traits, are convergent. In ad- 
dition, Harry's results indicate that support for 
Figure 1 rests largely on shared primitive 
traits. 

The criticisms of previous work should not 
be construed to mean that morphological 
traits used in previous studies of the Lymnae- 
acea provide no basis for phylogenetic infer- 
ence. These criticisms are only intended to 
point out that inferred patterns of morpholog- 
ical change of all characters should be tested 
as hypotheses. Component analysis (Nelson 
& Platnick, 1981) is a cladistic approach to 
phylogeny reconstruction designed to test hy- 
potheses of character evolution. Since a phy- 
logenetic branching pattern can be repre- 
sented as a series of nested sets of taxa, 
component analysis tests whether the sets of 
taxa implied by hypotheses of character evo- 
lution do nest. A character may have two 
states implying a transformation from a pre- 
existing, primitive state to a new, derived 
state. The state hypothesized to be derived is 



expected to define a monophyletic group 
composed of all descendants of the species 
in which the character state transformation 
occurred. A component is the set of taxa that 
share the state hypothesized to be dehved; it 
represents a hypothesis of monophyly. If two 
components nest, then both hypotheses of 
monophyly are consistent with the same phy- 
logenetic pattern, and corroborate both hy- 
potheses of character evolution (Le Quesne, 
1969; Nelson & Platnick, 1981). In a special 
case, two derived states define identical, rep- 
licated components, providing the strongest 
possible corroboration of the two character 
transformation hypotheses (Nelson & Plat- 
nick, 1981). There are two cases in which 
components do not nest. In one case, the 
components are intersecting sets represent- 
ing conflicting hypotheses of monophyly, and 
at least one of the character hypotheses must 
be rejected (Le Quesne, 1969). In the other 
case, the components are mutually exclusive, 
and the character hypotheses they represent 
need not be rejected (Le Quesne, 1969), but 
they are not corroborated either (Nelson & 
Platnick, 1 981 ). The treatment of mutually ex- 
clusive components distinguishes clique or 
compatibility analysis from component analy- 
sis. In clique analysis, a clique is a set of com- 
ponents that do not conflict, and the largest 
clique is chosen as the best estimate of the 
phylogeny (Estabrook et al., 1977). In com- 
ponent analysis, only nested and replicated 
components are used, and the largest set of 
nested components, representing the largest 
set of mutually corroborated character hy- 
potheses, is chosen as the best estimate of 
the phylogeny (Nelson & Platnick, 1981). 

Component analysis, and all other cladistic 
methods, are critically dependent on the 
sources used to generate hypotheses of char- 
acter transformation. Several sources have 
been used; three common ones are outgroup 
analysis, ontogenetic analysis and paleonto- 
logical (stratigraphie) analysis (cf. Eldredge & 
Cracraft, 1980; Nelson & Platnick, 1981). For 
this study, I chose to use outgroup analysis 
because it is an extension of component anal- 
ysis (Eldredge & Cracraft, 1 980; Wiley, 1 981 ). 
Outgroup analysis assumes that the study 
group (ingroup) is monophyletic and treats 
the ingroup as a component. Related taxa 
(outgroups) are members of larger compo- 
nents that include the ingroup. Outgroup anal- 
ysis sorts character states into two sets: (1) 
those shared by both ingroup and outgroups, 
and (2) those restricted to the ingroup. Char- 



PHYLOGENETIC SYSTEMATICS OF LYMNAEACEA 



251 



acter states shared by the ingroup and any 
outgroup may be either primitive or conver- 
gent (Maddison et al., 1984; Wiley, 1981). In 
either case, these character states cannot be 
used for phylogeny reconstruction. Only de- 
rived states define components that are sub- 
sets of the ingroup; this is the essence of cla- 
distic methodology (Estabrook et al., 1977; 
Eldredge & Cracraft, 1980; Nelson & Platnick, 
1981; Wiley, 1981). Outgroup analysis is a 
method that eliminates character transforma- 
tion hypotheses inconsistent with the hypoth- 
esis of ingroup monophyly. When the consis- 
tent character hypotheses are used in 
component analysis, the resulting phylogeny 
would be based on the largest set of mutually 
corroborated character hypotheses consis- 
tent with the initial hypothesis of ingroup 
monophyly. The results could be refined iter- 
atively by using the ingroup taxa found to be 
primitive as functional outgroups to order the 
states of characters not found in the original 
outgroups (Eldredge & Cracraft, 1980; Wa- 
trous & Wheeler, 1981). 

The study presented here is a re-evaluation 
of the phylogenetic relationships of the six lym- 
naeacean families, using outgroup analysis 
and component analysis. Parsimony algo- 
rithms were not used because previous stud- 
ies of the lymnaeaceans indicated consider- 
ably homoplasy; under these circumstances, 
parsimony algorithms become unreliable 
(Felsenstein, 1978). Because the purpose of 
this study is to test the phylogenetic conclu- 
sions of earlier studies, I have used published 
morphological descriptions. The principal 
sources of morphological descriptions of lym- 
naeaceans are the four previous studies of 
lymnaeacean phylogeny (Harry, 1 964; Huben- 
dick, 1947, 1978; Starobogatov, 1967). Addi- 
tional studies of more limited taxonomic or 
morphological scope were examined if their 
data were prominently featured in the phylo- 
genetic studies cited above (e.g. Demian, 
1962, radula; Duncan, 1960a, 1069b, oviduct; 
Hubendick, 1 964, ancylids). The superfamilies 
Amphibolacea and Ellobiacea were used as 
the outgroups, with descriptions provided pri- 
marily by Hubendick (1978). 

I have provisionally accepted the family and 
superfamily taxonomy of Hubendick (1978), 
in which Lymnaeacea Rafinesque (1815) is 
equivalent to Hygrophila Ferussac (1821). 
Hubendick's (1978) six major subdivisions of 
the Lymnaeacea do not differ from the subdi- 
visions of Hygrophila recognized by Harry 
(1964) or Starobogatov (1967). The differ- 



ences among these three authors are pnma- 
rily nomenclatural. Brief descnptions of the 
contents of the six lymnaeacean families are 
given in the Appendix. 



OUTGROUP ANALYSIS 

Outgroup analysis was performed by com- 
panng the descnptions of the lymnaeaceans to 
the descriptions of the outgroups. Because of 
the large amount of convergence in the order 
Basommatophora, which includes the Lym- 
naeacea, the nearest relatives of the lym- 
naeaceans cannot be identified confidently 
(Duncan, 1960a; Hubendick, 1978; Tillier, 
1 984). While the inability to identify the nearest 
relative may be a problem for parsimony al- 
gorithms (Maddison et al., 1984), it need not 
be a problem for components analysis (El- 
dredge & Cracraft, 1 980). The purpose of out- 
group analysis is to identify the derived states 
of the ingroup. The nearest outgroup is likely 
to provide the best estimate (Wiley, 1 981 ), but 
any outgroup will provide a partial estimate 
(Eldredge & Cracraft, 1980), and no single 
outgroup is likely to provide a completely ac- 
curate estimate (Maddison et al., 1984). 
Therefore, any non-lymnaeacean basom- 
matophoran could be used as an outgroup. 
The outgroups used in this study were the 
Ellobiacea and Amphibolacea, which encom- 
pass most of the non-lymnaeacean basom- 
matophorans. Any trait shared by lymnae- 
aceans with either of these other superfamilies 
was considered either primitive or convergent 
and rejected from the phylogenetic analysis. 
Only those traits unique to the Lymnaeacea 
were considered derived character states and 
subjected to further study. 

As shown in Table 1, only five characters 
have derived states that are both unique to 
the Lymnaeacea and shared by at least two 
lymnaeacean families. These are the only 
character states that might reflect phyloge- 
netic relationships (Nelson & Platnick, 1981). 
Each of the five characters in Table 1 are dis- 
cussed below. For each character, the vari- 
ous states are briefly described, and the de- 
rived state is identified. 

1 — Prostate morphology. The prostates of 
the outgroups Ellobiacea and Amphibolacea 
are comprised of a smooth glandular epithe- 
lium along the male duct or groove. This type 
of prostate is present in most chilinids. Three, 
more complex morphologies are found in the 
lymnaeaceans; (la) a patch of alveoli in some 



252 



SWIDERSKI 



TABLE 1 . Character-state distributions across lymnaeacean families. Only characters with derived states 
shared by at least two families are listed. Derived states are italicized. 











Ciliated 






Prostate 


Stomach 


Pneumostomal 


pulmonary 


Radular 




morphology 


muscles 


lappet 


ridge 


row 


Outgroups 


smooth/pocket 


cylinder/diverticula 


single 


present 


horizontal 


Chilinidae 


smooth alveoli 


bilobed 


single 


present 


chevron 


Latiidae 


smooth 


cylinder 


single 


present 


chevron 


Acroloxidae 


smooth 


absent 


single 


present 


chevron 


Lymnaeidae 


folds 


bilobed 


siphon 


absent 


horizontal 


Physidae 


diverticula 


reduced 


siphon 


absent 


chevron 


Planorbidae 


diverticula 


cyUnäer/ bilobed 


single's/p/ion/other 


present/absent 


horizontal 



chilinids, (lb) series of elongate digitiform di- 
verticula in Physidae and Planorbidae, (1c) a 
dilation of the vas deferens with invaginated 
folds in lymnaeids. The invaginated folds are 
unique to the Lymnaeidae and do not require 
further discussion in this paper. The compar- 
ison between alveoli and diverticula does 
merit further discussion. Starobogatov (1967) 
argues that all of the more complex morphol- 
ogies arise in response to a need for more 
efficient packing of prostate tissue and that 
these morphologies increase secretory sur- 
face area relative to the total volume occupied 
by the prostate. He also argues that evagina- 
tion and invagination are fundamentally dis- 
tinct approaches to the packing problem. Fol- 
lowing Starobogatov's argument, alveoli and 
diverticula are homologs, both are evagina- 
tions and the component defined by evagina- 
tion includes Physidae, Planorbidae and 
some chilinids. Alternatively, alveoli and di- 
verticula are regarded as completely separate 
traits, so that diverticula define a component 
that includes only Physidae and Planorbidae. 
Because available literature does not provide 
sufficient information to resolve this issue, two 
components are listed in Table 2: 1 ', defined 
by evagination (some Chilinidae, Physidae 
and Planorbidae); and 1", defined by diver- 
ticula (Physidae and Planorbidae). Because 
there is no definitive evidence that diverticula 
are derived from alveoli, component 1" can- 
not be considered a subset of 1'. Instead, 1' 
and 1 " are considered alternative interpreta- 
tions of a single character, and their relation- 
ships to other components were examined in- 
dependently. 

2 — Stomach muscle arrangement. Ellobla- 
cea and Amphibolacea have well-developed 
stomach muscles in one of two arrangements: 
a cylindrical band around the stomach, or a 
muscular diverticulum. The muscular divertic- 



ulum is found only in the outgroups; but the 
cylindrical band is found in some lym- 
naeaceans. Two new states are found in lym- 
naeaceans: (2a) reduction and loss of stom- 
ach muscle, and (2b), organization of muscles 
into two lobes. Reduction and loss are shared 
by Acroloxidae and Physidae. The bilobed ar- 
rangement is shared by Chilinidae, Lymnaei- 
dae and most Planorbidae. Because not all 
planorbids are included in component 2b, the 
family name is enclosed in brackets in Table 2. 

3 — Pneumostomal lappet. The pneumo- 
stomal lappet is a fold external to the pulmo- 
nary opening. In both outgroups, the lappet is 
a single lobe bisected by the rectum and may 
function as a gill. In Lymnaeidae, Physidae 
and most Planorbidae, the region anterior to 
the rectum is converted to a siphon, a tube 
which appears to function as a snorkel (Harry, 
1964). In Lymnaeidae and Physidae, the re- 
gion posterior to the rectum is absent. The 
posterior region is often reduced, but rarely 
absent, in planorbids with a siphon (Huben- 
dick, 1955). Other variations of the pneumo- 
stomal lappet occur in those planorbids with- 
out a siphon (Hubendick, 1964). 

4 — Ciliated pulmonary ridge. This is an in- 
ternal structure of the pulmonate lung that ex- 
tends from the pulmonary opening to the apex 
of the lung on both dorsal and ventral sur- 
faces. The ridge appears to facilitate gas ex- 
change by regulating water flow through the 
lung (Pilkington et al., 1984; Sullivan & 
Cheng, 1974). The ridge is present in both 
outgroups and many lymnaeaceans, but is 
absent from Lymnaeidae, Physidae and sev- 
eral planorbids. 

5 — Radular tooth arrangement. The geo- 
metric arrangement of teeth in rows appears 
to be the only radular trait that is not highly 
variable and frequently convergent (Demian, 
1962; Hubendick, 1978). Straight transverse 



PHYLOGENETIC SYSTEMATICS OF LYMNAEACEA 



253 



TABLE 2. Components defined by shared derived 
character states. Brackets indicate families m 
which not all species possess the derived state. 

1' [Chilinidae], Physidae, Planorbidae 

1" Physidae, Planorbidae 

2a Acroloxidae, Physidae 

2b Chilinidae, Lymnaeidae, [Planorbidae] 

3 Lymnaeidae, Physidae, [Planorbidae] 

4 Lymnaeidae, Physidae, [Planorbidae] 

5 Chilinidae, Latiidae, Acroloxidae, Physidae 



tooth rows are present in all outgroups and 
some lymnaeaceans. Chevron-shaped rows 
distinguish four lynnnaeacean families: Chilin- 
idae, Latiidae, Acroloxidae and Physidae. 



COMPONENT ANALYSIS 

All characters states unique to Lymnaeacea 
are hypothesized to be derived. Each charac- 
ter state defines a set of taxa that is hypoth- 
esized to represent a monophyletic group. 
Two states listed in Table 1 (folds in the lym- 
naeid prostate, alveoli in the chilinid prostate) 
define components that include only the mem- 
bers of individual families. Because these 
traits are not shared by two or more families, 
they cannot support inferences of relation- 
ships between families (Nelson & Platnick, 
1 981 ). These states are included in Table 1 for 
completeness in the lists of character-state 
distributions but are excluded from the com- 
ponent analysis. Other derived states unique 
to single families (e.g. preputial gland of Phys- 
idae; Te, 1 975) were omitted from Table 1 and 
are not considered further. For each remaining 
derived state listed in Table 1 , the component, 
the set of taxa sharing that state, is listed in 
Table 2. 

Component analysis is performed by in- 
specting all possible pairs of components to 
determine whether there are pairs of nested 
sets. Nested pairs are consistent with a single 
phylogenetic branching pattern and therefore 
represent mutually corroborated hypotheses 
of monophyly. Next, all possible combinations 
of nested pairs are assembled. The largest 
combination, the largest number of mutually 
corroborated hypotheses of monophyly, is 
considered the best estimate of the actual 
phylogeny. 

In a few cases, not all members of a family 
possess a particular derived state. Such com- 
ponents conflict with the hypothesis that the 
given family is a monophyletic group. Nor- 



mally, the monophyly of a family would not be 
challenged by a study of family-level relation- 
ships, and the traits that conflict with the fam- 
ily definition would be rejected as homoplas- 
tic. However, several cases involve one 
family, the Planorbidae. Therefore, I exam- 
ined the distribution of this set of traits across 
genera to determine whether they consis- 
tently divide the planorbids into two or more 
smaller groups. 

Two traits co-occur in most planorbids: loss 
of the ciliated pulmonary ridge (character 3), 
and formation of a siphon (character 4). How- 
ever, there are snails with a siphon that have 
not lost the ciliated ndge, and snails that have 
lost the ciliated ridge without acquinng a si- 
phon (Hubendick, 1955, 1964, 1978). Thus, 
the presence of the siphon and the loss of the 
ridge do not coincide in all planorbids and do 
not support an argument for dividing the fam- 
ily. Furthermore, since these two traits have 
conflicting distributions, one or both must be 
homoplastic. Hubendick (1955) shows differ- 
ent patterns of partial ridge loss in the coiled 
planorbids: some losing the dorsal portion of 
the ridge, others losing the ventral portion of 
the ridge. This diversity in intermediate states 
may mean that the terminal state, complete 
ridge loss, can be reached by at least two 
different evolutionary routes. This is not proof 
that ridge loss is convergent, but it does sup- 
port the argument of Hecht & Edwards (1976) 
that losses are more likely than new additions 
to be convergent. 

The third trait dividing the Planorbidae is 
the organization of stomach muscles into two 
lobes (character state 2b). The published 
data on this trait are sparse but indicate that 
only the planorbid limpets lack the derived 
state (Hubendick, 1964, 1978). Thus, the trait 
represents a third way of dividing the Planor- 
bidae. Only one, if any, can be right. The cor- 
rect trait can be recognized only if it defines a 
component that nests with one of the compo- 
nents that include all of the Planorbidae. 

Because the brackets do not represent 
identical or nested sets of genera, none of the 
components in Table 2 are identical: no two 
derived states independently support the 
same phylogenetic inference. Components 3 
and 4 appear similar; both include Lymnaei- 
dae, Physidae and some planorbids; but they 
divide the Planorbidae into different groups, 
representing conflicting hypotheses of rela- 
tionship. In fact, most pairs of components 
represent conflicting hypotheses of relation- 
ships. One exception is composed of compo- 



254 



SWIDERSKI 




Physidae 



FIG. 2. Phylogenetic relationships supported by 
component analysis. 

nents 2a (Acroloxidae and Physidae) and 2b 
(Chilinidae, Lymnaeidae and Planorbidae). 
These two components are mutually exclu- 
sive, representing two independent hypothe- 
ses of character evolution in separate lin- 
eages. The components are not contradictory 
and could be used in clique analysis, but they 
do not corroborate each other. Consequently, 
the relationship between 2a and 2b does not 
contribute to the component analysis solution. 
The only nested components, representing 
mutually corroborated character transforma- 
tion hypotheses, are 2a (Acroloxidae and 
Physidae) and 5 (Chilinidae, Latiidae, Ac- 
roloxidae and Physidae). Therefore, the only 
phylogenetic relationships supported by com- 
ponent analysis are those shown in Figure 2. 
The branching pattern in Figure 2 is not 
completely resolved: there are two trichoto- 
mies. Each trichotomy indicates that the rela- 
tionship of three lineages remains unre- 
solved. Each trichotomy has three possible 
solutions (Eldredge & Cracraft, 1980; Nelson 
& Platnick, 1981), so there are nine fully re- 
solved trees consistent with Figure 2. How- 
ever, Figure 2 does indicate that within Lym- 
naeacea there is a monophyletic group 
characterized by a unique, denved, chevron 
radular row (component 5). This group in- 
cludes four families: Chilinidae, Latiidae, Ac- 
roloxidae and Physidae. The relationships of 
Lymnaeidae and Planorbidae to each other 
and to the group remain unresolved. Within 
the group defined by component 5, the rela- 
tionships of Chilinidae and Latiidae are un- 
clear, but Physidae and Acroloxidae (compo- 
nent 2a) appear to represent a distinct lineage 
characterized by reduced stomach muscles. 

DISCUSSION 

The relationships shown in Figure 2 are 
based on only two derived character states. 



With so little support, this phylogeny could be 
rejected rather easily. Only one additional de- 
rived state defining a component that nests 
with any of the other components in Table 2 
would produce an equally well-supported al- 
ternative phylogeny. Because phylogenies 
based on few traits are highly susceptible to 
revision, Figure 2 can only be regarded as a 
tentative hypothesis of lymnaeacean relation- 
ships. 

Although outgroup analysis of the charac- 
ters available in the lymnaeacean literature did 
not produce a strongly supported new phylog- 
eny, the results indicate that the generally ac- 
cepted phylogeny (Figure 1) should be re- 
jected. Only one component of Figure 1 was 
confirmed by outgroup analysis: Physidae + 
Planorbidae, defined by the prostate divertic- 
ula. None of the other monophyletic groups 
implied by Figure 1 is listed as a component in 
Table 2. Furthermore, half of the components 
in Table 2 conflict with the basal dichotomy 
shown in the conventional phylogeny. Com- 
ponents 3 and 4, defined by respiratory traits, 
are close to matching one branch of Figure 1 , 
but they conflict over which planorbids belong 
to that lineage. Thus, the outgroup analysis 
provides an argument for rejecting the con- 
ventional lymnaeacean phylogeny, indepen- 
dent of the support it provides for a specific 
revision. 

These results also cast doubt on the use of 
most reproductive tract characters for phylog- 
eny reconstruction. This is significant be- 
cause reproductive traits have been the prin- 
cipal characters considered in previous 
phylogenetic studies of the Lymnaeacea 
(separation of male and female ducts, Harry, 
1964, Hubendick, 1947, 1978; prostate, 
Hubendick, 1947, 1978, Starobogatov, 1967; 
preputium, Harry, 1964, Hubendick, 1947). 
Based on outgroup analysis, most evolution- 
ary changes in the lymnaeacean reproductive 
system are not unique to that superfamily. 
Only one reproductive character, prostate 
morphology, has a unique derived state listed 
in Table 1. Considering the number of sepa- 
rate, identifiable reproductive structures, the 
lack of unique derived states implies tremen- 
dous homoplasy. This large amount of ho- 
moplasy is not a surprise, however; because 
phylogenetic studies within lymnaeacean 
families frequently conflict when different re- 
productive traits are used (Hubendick, 1951, 
1955; Meier-Brook, 1983; Te, 1975, 1978). In 
addition, the patterns of character evolution 
described for reproductive traits conflict with 



PHYLOGENETIC SYSTEMATICS OF LYMNAEACEA 



255 



the few patterns that have been described for 
respiratory and radular characters (Huben- 
dick, 1955, 1978; Meier-Brook, 1983; Te, 
1978). Evidently, evolutionary changes of the 
reproductive system do not unambiguously 
reflect lymnaeacean relationships. 

While the results of the current study reject 
the use of reproductive characters for phylog- 
eny reconstruction, this study does not reject 
the possible evolutionary significance of 
these traits. In fact, evolutionary significance 
may account for the lack of phylogenetic sig- 
nificance. A partial explanation of conver- 
gence is that the functional or adaptive impor- 
tance of a trait ensures strong selective 
pressures favoring any changes that might 
improve function. This can be only a partial 
explanation of frequent convergence because 
the genetic and developmental sources of the 
necessary variation are not considered. Still, 
lymnaeacean reproductive characters may be 
examples of traits with great functional impor- 
tance that are not unique innovations, but the 
results of frequent convergence. 

Evolutionary significance and the causes of 
convergence are outside the scope of this pa- 
per. The focus of this study is that outgroup 
analysis of published data demonstrates that 
confidence in any interpretation of lymnaea- 
cean relationships is misplaced. Failure to 
produce a well-corroborated phylogeny of the 
Lymnaeacea is due, in part, to the uncertainty 
concerning the outgroup phylogeny. If out- 
group relationships were understood, it might 
be possible to determine which outgroup 
traits are most likely to be primitive and which 
traits are probably convergent (Maddison et 
al., 1984). Discriminating between primitive 
and convergent traits would help focus efforts 
aimed at identifying and discriminating among 
separate convergence events. Ultimately, 
recognition of the unique aspects of individual 
convergence events may enable the identifi- 
cation of separate monophyletic groups 
among convergent taxa (e.g. Lombard & 
Wake, 1986). 

Resolving outgroups and reproductive con- 
vergence provides only a partial solution to 
the problem of lymnaeacean phylogeny. The 
results of this study show that the lymnaea- 
cean phylogeny cannot be fully resolved us- 
ing currently available, morphological data. 
New data should be generated both from de- 
scriptions of morphological traits in other or- 
gan systems and from analysis of molecular 
traits. Granted, new data may be as prone to 
convergence as the traits discussed in this 



report, but there is no possibility of a solution 
until these other avenues are explored. 



ACKNOWLEDGMENTS 

This paper is based on analyses performed 
as part of a M.S. thesis for the Department of 
Geological Sciences, Michigan State Univer- 
sity. I would like to thank my advisory com- 
mittee: R. L. Anstey, D. F. Sibley and D. O. 
Straney. I would also like to thank D. O. 
Straney, M. L. Zelditch, A. С Carmichael, and 
the reviewers for their comments on earlier 
versions of this paper. 



LITERATURE CITED 

AGEITOS DE CASTELLANO, Z. J. & S. E. 
MIQUEL, 1980, Notas complementarias al gen- 
ero Chilina Gray (Mollusca Pulmonata). Neotro- 
pica. 26: 171-178. 

AHO, J., 1966, Ecológica! basis of the distribution 
of the freshwater molluscs in the vicinity of Tam- 
pere, South Finland. Annales Zoologici Fennici. 
3: 287-322. 

AYYASAMY, K., V. GANESAN & С V. L. 
NARASIMHAM, 1985, A new species of the ge- 
nus Physa from the Deccan infra-trappean beds 
of Andhra Pradesh, India. Journal Geological So- 
ciety of India. 26: 744-750. 

BRACE, R. C, 1983, Observations on the morphol- 
ogy and behaviour of Chilina fluctuosa Gray 
(Chilinidae), with a discussion on the early evo- 
lution of pulmonate gastropods. Phiilosopfiical 
Transactions of the Royal Society of London, В 
300: 463-491. 

BURCH, J. B. & С M. PATTERSON, 1964, Cyto- 
taxonomic studies of freshwater limpets (Gas- 
tropoda: Basommatophora) II. The New Zealand 
river limpet, Latia neritoides. Malacologia, 1 : 
313-318. 

CLARKE, A. H., 1973, The freshwater molluscs of 
the Canadian Interior Basin, t^alacologia, 13: 1- 
492. 

CLARKE, A. H., 1979, Gastropods as indicators of 
trophic lake stages. Nautilus, 93: 138-142. 

DEMIAN, E. S., 1962, Comparative morphological 
studies on the buccal mass of the Lymnaeacea. 
Göteborgs Kungliga Vetenskaps- ocfi Vitterhiets- 
Samtialles Handligar. Sjatte Foljden, Series B, 
Band 9, No. 3, 30 pp. 

DUNCAN, С J., 1960a, The evolution of the pul- 
monate genital system. Proceedings of thie Zoo- 
logical Society of London, 134: 601-609. 

DUNCAN, С J., 1960b, The genital system of the 
freshwater Basommatophora. Proceedings of tfie 
Zoological Society of London, 135: 339-356. 

ELDREDGE, N. & J. CRACRAFT, 1980, Pfiyloge- 



256 



SWIDERSKI 



netic patterns and the evolutionary process. Co- 
lumbia University Press, New York, 349 pp. 

ESTABROOK, G. F., J. G. STRAUCH, JR. & K. L. 
FIALA, 1 977, An application of compatibility anal- 
ysis to the Biacklths data on orthopteroid Insects. 
Systematic Zoology. 26: 269-276. 

FELSENSTEIN, J., 1978, Cases in which parsi- 
mony or compatibility methods will be positively 
misleading. Systematic Zoology. 27: 401-410. 

GOMEZ, J.. M. VARGAS & E. A. MALEK, 1986, 
Pseudosuccinea columella, new record (Lym- 
naeidae) in the Dominican Republic and trans- 
mission of fasclollasis In the Caribbean region. 
Nautilus. 100: 66-78. 

GOOD, S. C, 1987, Mollusc-based Interpreta- 
tions of lacustrine paleoenvironments of the 
Sheep Pass Formation (Latest Cretaceous to 
Eocene) of East Central Nevada. Palaios. 2: 
467-478. 

HARRY, H. W., 1964, The anatomy of Chilina fluc- 
tuosa Gray reexamined, with prolegomena on 
the phylogeny of the higher llmnic Basommato- 
phora (Gastropoda: Pulmonata). Malacologla. 1: 
355-385. 

HECHT, M. К. & J. L. EDWARDS, 1976, The de- 
termination of parallel or monophyletic relation- 
ships: the proteld salamanders — a test case. 
American Naturalist, 110: 653-677. 

HUBENDICK. В., 1947, Phylogenese relations be- 
tween higher llmnic Basommatophora. Zoolo- 
giska Bidrag fran Uppsala. 25: 141-164. 

HUBENDICK, В., 1948a, Studies on Bulinus. Arkiv 
for Zoologi. Band 40A, No. 16, 63 pp. 

HUBENDICK, B. 1948b, The anatomy of Bulinus. 
with a discussion of the term prostate and its 
sense in the Basommatophora. Proceedings of 
ttie Malacological Society of London. 27: 186- 
196. 

HUBENDICK, В., 1951, Recent Lymnaeidae. 
Kungliga Svenska Vetenskapsacademien Han- 
dligar, Ser. 4, 3: 1-223. 

HUBENDICK, В., 1955, Phylogeny In the Planor- 
bldae. Transactions of thie Zoological Society of 
London. 28:453-542. 

HUBENDICK, В., 1 962, Studies on Acroloxus (Moll. 
Basomm.). Göteborgs Kungliga Vetenskaps- ocfi 
Vitterhets- Samhialles Handligar. Sjatte Foljden, 
Series B, Band 9, No. 2, 68 pp. 

HUBENDICK, В., 1964. Studies on Ancylidae. 
Göteborgs Kungliga Vetenskaps- ochi Vitterhets- 
Samfialles Handligar, Sjatte Foljden, Series B, 
Band 9, No. 6, 72 pp. 

HUBENDICK, В., 1978, Systematlcs and compar- 
ative morphology of the Basommatophora. In 
FRETTER, V. & J. PEAKE, eds., Pulmonates, 
Vol. 2A, pp. 1-47. Academic Press, London. 

INTERNATIONAL COMMISSION ON ZOOLOGI- 
CAL NOMENCLATURE, 1985, International 
code of zoological nomenclature, Third Edition. 
British Museum (Natural History) London (for In- 
ternational Trust for Zoological Nomenclature), 
338 pp. 

JELNES, J. E. 1986, Experimental taxonomy of Bu- 



linus (Gastropoda: Planorbidae): the West and 
North African species reconsidered, based on an 
electrophoretlc study of several enzymes per in- 
dividual. Zoological Journal of the Linnean Soci- 
ety, 87: 1-26. 

LA ROCQUE, A., 1966-1970, Pleistocene Mol- 
lusca of Ohio. Ohio Division of the Geological 
Sun/ey. Bulletin 62, 800 pp. 

LE QUESNE, W. J., 1969, A method of selection of 
characters in numerical taxonomy. Systematic 
Zoology, 18: 201-205. 

LOMBARD, R. E. & D. B. WAKE, 1986, Tongue 
evolution In the lungless salamanders. Family 
Plethodontldae IV. Phylogeny of plethodontid 
salamanders and the evolution of feeding dy- 
namics. Systematic Zoology, 35: 532-551. 

MADDISON, W. P., M. J. DONOGHUE & D. R. 
MADDISON, 1984, Outgroup analysis and parsi- 
mony. Systematic Zoology, 33: 83-103. 

MANDAHL-BARTH, G., 1957, Intermediate hosts 
of Schistosoma, African Biomphalaria and Buli- 
nus: I. Bulletin of the World Health Organization, 
16: 1103-1163. 

MEIER-BROOK, C, 1983, Taxonomic studies on 
Gyraulus (Gastropoda: Planorbidae). Malacolo- 
gla, 24: 1-113. 

NELSON, G. & N. PLATNICK, 1981, Systematlcs 
and biogeography: Cladistics and vicariance. Co- 
lumbia University Press, New York, 567 pp. 

PILKINGTON, J. В., С. LITTLE & P. E. STIRLING, 
1 984, A respiratory current in the mantle cavity of 
Amphibola crenata (Mollusca, Pulmonata). Jour- 
nal of the Royal Society of New Zealand, 14: 
327-334. 

STAROBOGATOV, Y. I., 1967, On the systemati- 
zation of freshwater pulmonale molluscs. Trudy 
Zoologicheskogo Instituía Akademia Nauk 
SSSR, 42: 280-304 (In Russian). 

SULLIVAN, J. T. & T. С CHENG, 1974, Structure 
and function of the mantle cavity In Biomphalaria 
glabrata. Transactions of the American f\/licro- 
scopical Society, 93: 416-420. 

ТЕ, G. A., 1975, Michigan Physidae, with system- 
atic notes on Physella and Physodon (Basom- 
matophora: Pulmonata). Malacological Review, 
8: 7-30. 

TE, G. A., 1978, A systematic study of the family 
Physidae (Basommatophora: Pulmonata). Ph.D. 
thesis, University of Michigan. University Micro- 
films International, Ann Arbor, Michigan. 

TILLIER, S., 1984, Relationships of gymnomorph 
gastropods (Mollusca: Gastropoda). Zoological 
Journal of the Linnean Society, 82: 345-362. 

WATROUS, L. E. & Q. D. WHEELER, 1981, The 
outgroup comparison method of character anal- 
ysis. Systematic Zoology, 30: 1-11. 

WILEY, E. O., 1981, Phylogenetics: The theory and 
practice of phylogenetic systematlcs. J. Wiley & 
Sons, New York, 439 pp. 



Revised Ms. accepted 31 July 1989 



PHYLOGENETIC SYSTEMATICS OF LYMNAEACEA 



257 



APPENDIX 

The superfamily Lymnaeacea (Pulmonata: 
Basommatophora) is composed of six fami- 
lies. These families share the following traits: 
a glandular oviduct divided into histologically 
and functionally distinct regions (Duncan, 
1 960a, 1 960b), freshwater habitat, and loss of 
the operculum in both adult and embryo 
(Hubendick, 1978). None of these traits are 
unique to lymnaeaceans, but together these 
traits serve to distinguish the Lymnaeacea 
from other basommatophorans. Conse- 
quently, the Lymnaeacea is generally consid- 
ered to be a monophyletic group. Brief de- 
scriptions of the six lymnaeacean families are 
given below: 

Chilinidae H. & A. Adams, 1855— This is a 
monogeneric family erected to accommodate 
Chilina Gray, 1828. Several species have 
been reported from the rivers and estuaries of 
Chile, Argentina and Paraguay (Ageitos de 
Castellano & Miquel, 1980; Brace, 1983). The 
variation reported for Chilina is limited to shell 
shape and pigmentation, and radular tooth 
shape. 

Latiidae Hutton, 1882 — This family includes 
only one species, Latia neritoides Gray, 1 849, 
a freshwater limpet restricted to the streams 
of New Zealand (Burch & Patterson, 1964; 
Hubendick, 1962). 

Acroloxidae Thiele, 1931 — This family of 
limpets is usually considered to comprise a 
single genus, Acroloxus, but the subgenus 
Pseudancylastrum is sometimes elevated to 
the generic level. The species A. coloradensis 
is found in Colorado, but the remaining spe- 
cies are found in Europe and northern Asia 
(Hubendick, 1978). The principle variations in 
this family are shell shape and radular tooth 
shape (Hubendick, 1962). 



Physidae Fitzinger, 1833— Te (1978) pre- 
sents a recent revision of this family, in which 
he recognizes 48 species in four genera. Te 
differentiated physids on the basis of shell 
shape, pigmentation patterns, mantle edge 
shape, kidney and gizzard shape, and the 
structures of the bursa copulatorix and the pe- 
nial complex. This family is globally distnb- 
uted. 

Lymnaeidae Rafinesque, 1815— This fam- 
ily exhibits tremendous variation in shell mor- 
phology, supporting a large number of nomi- 
nal genera. There also is considerable 
variation in anatomical traits, especially the 
male reproductive organs. Hubendick (1951) 
was unable to discern any clear pattern 
among anatomical traits or between anatom- 
ical and conchological traits. Therefore, he 
concluded that this family includes only two 
genera, the helicoid Lymnaea and the patel- 
liform Lanx. The family is globally distnbuted. 

Planorbidae Gray, 1840 — This family rep- 
resents the merging of three classical fami- 
lies: helicoid Bulinidae Hermansen, 1846; dis- 
coid Planorbidae Gray, 1840; and patelliform 
Ancylidae Brown, 1844. In a series of papers, 
Hubendick (1947, 1948a, 1948b, 1955, 1964, 
1978) demonstrated that these three groups 
are not clearly separable, but have complex 
interrelationships. Hubendick (1978) coined 
the name Ancyloplanorbidae to indicate the 
synthetic nature of the group. However, under 
Article 23 of the ICZN, Ancyloplanorbidae 
must be considered a junior synonym Planor- 
bidae Gray, 1840; the oldest of three names 
from the merged families. The Planorbidae 
are globally distributed, and there is consid- 
erable variation in both shell morphology and 
internal anatomy. 



MALACOLOGIA, 1990, 31(2): 259-295 

LONGEVITY IN MOLLUSCS 

Joseph Heller 

Department of Zoology, The Hebrew University, Jerusalem, Israel 

ABSTRACT 

This paper compares longevity throughout an entire phylum, the Mollusca, in order to reveal 
common patterns underlying modes of reproduction. The comparison is based upon data 
gleaned from existing literature on the life durations of 547 species from mahne, freshwater and 
terrestrial habitats. 

Life-spans of molluscs range from two months to two hundred years. Molluscs living up to two 
years, or molluscs living more than two years but reproducing during only one season, are here 
defined as short-lived. 

Many molluscs are long-lived, and bivalves are the most long-lived of molluscs. In the terres- 
trial and manne habitat, a short-lived mode of life is often correlated with: (1) Lack of an external 
shell. (2) Possession of an external shell that is semitransparent. (3) Dwelling in a microenvi- 
ronment that is exposed to high solar radiation and to high temperatures. (In cold environments, 
on the other hand, the semelparous cycle of molluscs without external shells may be stretched, 
over two years or more.) (4) Dwelling in an environment that is predictable to such an extent that 
conditions favourable for reproduction occur (for an annual species) at least once a year. 
(5) Very minute size (in gastropods). 

These generalizations apply almost fully to terrestrial and manne habitats and are partly valid 
in freshwater habitats. 

The correlation between shell absence and longevity accounts for the greatest number of 
short-lived molluscs. This correlation may be explained in adaptive terms: shell absence may 
affect age-specific mortality via growth rates; or shell-less molluscs may utilise transient food 
resources. The difficulty in accepting any of these adaptive explanations stems from the ubiquity 
of the relation between shell-lessness and a short life span: almost every single shell-less 
mollusc, over a wide range of habitats in the sea and on the land, is short-lived. 

The correlation may be also explained in non-adaptive terms: shell and longevity covary, so 
that an initial, adaptive change in the shell engenders a secondary, automatic change in the 
life-span. If this non-adaptive explanation is indeed valid, then the short life span of many 
molluscs may be a byproduct of selection on the shell rather than an independently selected 
trait. One major difficulty in accepting this non-adaptive explanation is that it lacks evidence at 
the genetic level. 

Whatever the explanation for these correlations, they can be used to calculate the approxi- 
mate number of short-lived gastropods. On a very broad and rough estimate, about one half of 
the land snail genera of western Europe may be short-lived. 

Key words: longevity, molluscs, reproductive strategies, morphology, adaptation, non-adap- 
tation, size, radiation. 



INTRODUCTION 

In this paper, I compare longevity through- 
out an entire phylum, the Mollusca. To the 
best of my knowledge this topic has previ- 
ously been examined only twice, by Comfort 
(1957, 1964). On a more limited taxonomic 
range however, Zolotarev (1980) described 
the life spans of many bivalves from the Sea 
of Japan, and Powell & Cummins (1985) sur- 
veyed the longevities of some marine benthic 
prosobranchs and bivalves. 

Records of the length of time a mollusc 



lives are occasionally documented in ecolog- 
ical studies devoted to exploring the life his- 
tories of single species, or of several species 
within one genus. Frequently these data are 
presented in an incidental manner that does 
not relate to any larger evolutionary trends. 
Here, I collate data from existing literature to 
examine whether any general patterns of lon- 
gevity can be traced throughout the entire 
mollusc phylum. 

Methods of determining the age of molluscs 
include counting of growth checks in the shell 
(sometimes by shell-sectioning to reveal in- 



259 



260 



LONGEVITY IN MOLLUSCS 



terna! growth lines), population sampling, and 
the recapturing of marked animals. Infre- 
quently the age is also determined by isotopic 
analysis of the shell (e.g. Turekian et al., 
1975), and very infrequently by use of spec- 
tral analysis and flame photometry, or com- 
plex ionometric titration (Krasnov et al., 
1975). The assembling of longevity data de- 
termined by any of these methods into com- 
prehensive tables Is a simple, straightfonл^ard 
process. 

For the purposes of this study, I have ag- 
gregated all molluscs into two categories: 
short-lived (SL) and long-lived (LL). Short- 
lived molluscs are all those species that live 
up to two years, and also all those species 
that, regardless of how long they live, breed 
only over one season in their lifetimes. In con- 
trast, long-lived molluscs are all those species 
that live for more than two years and breed 
over two seasons at least. These two cate- 
gories can be compared to Kirkendall & 
Stenseth's (1985) classification of reproduc- 
tive strategies. All their uniparous, unisea- 
sonal and biseasonal molluscs fit into my 
short-lived category, whereas all their multi- 
seasonal iteroparous molluscs that live for 
more than two years are included in my long- 
lived category. 

The present classification overcomes sev- 
eral entanglements ahsing from the fact that 
in some semelparous molluscs the life-span 
is variable, being annual in one habitat but 
stretching over several years in another be- 
cause of a colder environment. When beahng 
in mind that life spans of molluscs range from 
several months to over two hundred years, 
the differences that this classification over- 
looks, in longevity amongst molluscs living up 
to two years, are minor. On the other hand, 
one of the disadvantages of this classification 
is that it lumps together, within the short-lived 
group, iteroparous molluscs that produce only 
a dozen progeny with semelparous species 
that produce many millions. I shall return to 
discuss this point later. 

It is suggested that when analysing the 
data, the lowest group of long-lived molluscs, 
those with life-spans of 2-3 years, be sepa- 
rated as an intermediate category, to be ex- 
cluded from later calculations. By doing so we 
avoid a situation whereby molluscs living 
three years but reproducing twice are classi- 
fied as long-lived, whereas those living three 
years and reproducing once (such as certain 
cephalopods) are classified as short-lived. I 
do not think that this intermediate category 



bears any biological uniqueness as compared 
to the short- or the long-lived categories. 

With 60,000 Recent species, molluscs form 
the second-largest phylum within the animal 
kingdom. Longevity determines the number of 
seasons in which many of these species (the 
iteroparous ones) will reproduce. What are 
the life histories of molluscs? Which are the 
short-lived ones? Why is it that of two mollusc 
species living in the same environment and in 
very close proximity, feeding in a similar way 
and predated by similar enemies, one is 
short-lived and the other long-lived? These 
are the questions addressed in this paper. 



METHODS 

The available literature was searched, and 
each species classified as short- or long- 
lived. 

Comfort (1957) reviewed the literature on 
the life duration of 135 molluscs. However, 
many of his records are of observations on 
captive specimens. This present paper there- 
fore considers his data only to the extent that 
they refer to natural populations, to taxa 
traceable to the generic level (at least) in to- 
day's taxonomy, and to those for which no 
more recent records could be found. Some of 
the literature on freshwater gastropods 
(Calow, 1978; Browne & Russel-Hunter, 
1978) bears conflicting evidence in that spe- 
cies listed as semelparous by one may be 
listed as iteroparous by the other. This re- 
flects the fact that within a species, some 
populations may be semelparous, others 
iteroparous or "quasi-iteroparous" (a gener- 
ally semelparous population that develops 
iteroparity under special circumstances). To 
overcome this difficulty, I have arbitrarily clas- 
sified in this paper as short-lived any species 
that is enlisted as semelparous or quasi- 
iteroparous by at least one of these authors. 

Tables 1-6, in the Appendix, present the 
maximum number of years a species lives, as 
recorded in the literature. Many authors de- 
scribe various species as living "at least" a 
certain number of years. These minimum es- 
timates of longevity are here presented as the 
life-span of the species, without further com- 
ment. The life-spans of short-living species 
are presented in these tables as SL, without 
further detail. 

As there is a close relation between the age 
at maturity and longevity among limpets 
(Branch, 1 981 ),opisthobranchs (Todd, 1981), 



HELLER 



MARINE 
BIVALVES 




FRESHWATER 
BIVALVES 






Шшм. 



MARINE 

PROSOBRANCHS 
8 PULMONATES 



St 2 4 6 8 Ю 12 



261 

^ OPISTHOBRANCHS 

I 



6 a 10 12 



CEPHALOPODS 



ja_ 



1 

1 
1 



FRESHWATER 
SNAILS 



SL 2 4 6 8 10 12 



LAND SNAILS 




FIG. 1. Life-span frequencies in various mollusc groups. SL, short-lived category. In each histogram the 
highest (right-most) life-span category includes all species that live 14 years or more. Data from Tables 1-6. 



land snails (Baur & Bengtsson, 1987) and, 
presumably, other mollusc groups, this trait is 
not presented in this paper. 



LONGEVITY DATA 

Fig. 1 illustrates the frequency of life spans 
in some various mollusc groups. 

Chitons 

Chitons are exclusively marine. Very few 
reliable data are available on their longevity, 
especially since Glynn (1970) severely criti- 
cised methods used by early workers in de- 
termining life spans. The only acceptable 
records I could find are of Cryptochiton stel- 
len, which lives for at least 25 years (MacGin- 
Itie & MacGinitie, 1968), and of Chaetopleura 



apiculata, which lives up to 4 years (Grave, 
1932). 

Gastropods 

Marine snails occur in two main groups, the 
prosobranchs and the opisthobranchs. Proso- 
branchs (20,000 species) are usually long- 
lived, whereas most opisthobranchs (about 
2,000 species) are short-lived. They repro- 
duce continuously once they reach sexual 
matunty, the frequency of their egg laying 
varying from several times per day to once 
every three weeks. They eventually die as a 
result of a senescent syndrome, typified by 
the shrinkage and breakdown of the digestive 
gland (Thompson 1976; Hadfield & Switzer- 
Dunlap, 1984). For nudibranchs (the largest 
group among the opisthobranchs), Todd 
(1981) distinguished three somewhat arbi- 



262 



LONGEVITY IN MOLLUSCS 



trary life-history patterns: (1) subannual spe- 
cies (life spans of a few weeks to a few 
months) are mostly small aeolids that feed on 
ephemeral prey, mainly hydrozoans; (2) an- 
nual species are larger (e.g. mostly donds) 
and eat animals that persist in time, such as 
sponges, barnacles, bryozoans; and (3) bien- 
nial species are large animals (a few den- 
dronotaceans and dorids) that feed on large, 
long-lived prey, such as alcyonahans. How- 
ever, Hadfield & Switzer-Dunlap (1984) sug- 
gest that there is probably a continuum in the 
distribution of opisthobranch life spans, from 
species with life spans of a few weeks, 
through those with intermediate life spans of 
months, to others living one year or more. A 
few marine snails belong to the pulmonales, 
here represented by one family, the Sipho- 
nariidae, which are long-lived (Powell & Cum- 
mins, 1985). 

Considering only the prosobranchs from 
among the marine gastropods, I found (Table 
1) records for 105 species belonging to 52 
genera and to 30 families. This amounts to 
about 2% of the 2,900 Recent genera (see 
Taylors Sohl, 1962). 

For the opisthobranchs, I found records of 
63 species belonging to 37 genera and 25 
families. This amounts to about 7% of the 500 
Recent genera of opisthobranchs (see Taylor 
& Sohl, 1962). All are short-lived. 

Of the marine pulmonales, I found records 
of three species, all belonging to one genus. 

Freshwater snails belong to one of two ma- 
jor groups, the pulmonale basommatophorans 
and the prosobranchs. Most basommatopho- 
rans are short-lived. They are annual and 
semelparous, with complete replacement of 
generations after breeding in 1а*е spring or 
early summer. However, although this is the 
basic pattern, closer observation shows much 
variation. One such deviation from the basic 
pattern is the production of a second summer 
breeding generation without replacement of 
one generation by the other. Another deviation 
is production of two generations per year, with 
complete replacement. Sometimes there can 
also be three generations, again, with or with- 
out replacement. Lastly, a perennial, often bi- 
ennial, pattern also occurs, with each gener- 
ation capable of reproducing in two successive 
years. Such patterns of intraspecific variation 
in life histones are common amongst fresh- 
water pulmonales and might be due either to 
ecological effects, genetic divergence, or to a 
combination of these factors (Russell-Hunter, 
1978). In the freshwater pulmonale Lymnaea 



elodes for example, intraspecific variation in 
life histories appears to be the result of phe- 
notypic plasticity rather than of genetic differ- 
ences (Brown, 1985). Freshwater proso- 
branchs tend to be more long-lived than 
freshwater pulmonales (Calow, 1978; Brown, 
1983; Geraerts & Joosse, 1984). Freshwater 
prosobranchs also have smaller clutch sizes, 
lower growth rates, smaller shell sizes at ma- 
tunty and larger shell sizes at death (Brown, 
1983). 

Both Calow (1978) and Geraerts & Joosse 
(1 984) suggest that semelparity is a response 
to the harsh freshwater conditions that make 
it necessary to confine the whole embryonic 
development within the protecting egg mass. 
This procedure demands an increased repro- 
ductive effort but increases the chance of 
embryonic survival, and hence diminishes 
the need for a long adult phase as an insur- 
ance policy. Both studies suggest that this 
semelparous condition is associated with re- 
productive recklessness, in that the parents 
continue low reproductive activity under ad- 
verse conditions despite fatal effects, where- 
as in iteroparous species reproduction stops 
quickly and the available energy is saved for 
survival. Both also comment that freshwater 
snails with an iteroparous strategy are those 
that inhabit small, closed water bodies, where 
there is more competition, more density-de- 
pendent control, and hence a greater pre- 
mium on the survival of a large, "experi- 
enced" adult. 

I have found records of 60 species belong- 
ing to 29 genera and 1 1 families (Table 2). 
Both prosobranch and pulmonale freshwater 
snails have, on the whole, short life-spans as 
compared to marine snails (Table 1) and ter- 
restrial ones (Table 3). 

Land snails belong to one of two groups, 
the prosobranchs (mostly confined to the 
tropics) and the (mostly stylommatophoran) 
pulmonales, which are distributed world-wide. 
I could not find records concerning the lon- 
gevities of terrestrial prosobranchs, but for the 
vague statement that Pomatias elegans is 
"said to live 4-5 years" (Fretter & Graham, 
1978). The reproductive strategies of terres- 
trial pulmonales have been reviewed by Dun- 
can (1975), who emphasized that maturation 
and growth of stylommatophorans are tem- 
perature-dependent, and suggested that 
small helicids mature more quickly and have 
a shorter life span than large species. 

I found records of 75 species belonging to 
57 genera and 30 families (Table 3). With an 



HELLER 



263 



overall estimate of 2,200 stylommatophoran 
genera (Taylor & Sohl, 1962), this is 3%. 

Bivalves 

Marine bivalves are represented by a very 
large array of groups. I found records of 150 
species belonging to 90 genera and 37 fam- 
ilies (Table 4). This amounts to about 6% of 
the 1,400 Recent genera (data from Vokes, 
1967). Many of these records come from the 
study of Zolotarev (1 980), who found that half 
of the species examined from the Sea of Ja- 
pan have life spans of more than 20 years. 

Most freshwater bivalves belong to two 
families, the unionids (in which the juveniles 
undergo a parasitic stage in fish) and the 
sphaeriids (in which the young are brooded in 
the mother's body, emerging as miniatures of 
the adult). I found records for 52 species be- 
longing to 17 genera and to five families 
(Table 5). This amounts to about 4% of the 
400 genera of Recent freshwater bivalves 
(data from Vokes, 1967). 

Cephalopods 

There are about 650 species of cephalo- 
pods, belonging to 140 genera (Voss 1977). I 
found (Table 6) records for 27 species be- 
longing to 1 7 genera and to nine families. This 
amounts to about 12% of the 140 genera of 
Recent cephalopods. Except for one (Nau- 
tilus, which lives for well over 20 years; Saun- 
ders, 1984), all cephalopods are short-lived, 
reproducing through one season only, and 
death is the typical consequence of egg lay- 
ing or mating (Arnold & Williams-Arnold, 
1977; Wells & Wells, 1977; Calow, 1987). 

According to Calow (1987), the failure of 
cephalopods to take advantage of the wide 
variety of reproductive tactics used by other 
mollusc groups is a consequence of selection 
for fast growth rates in the juveniles. He sug- 
gested that rapid growth would reduce the 
likelihood of juvenile mortality due to préda- 
tion, because juveniles would be small and 
vulnerable for a shorter time. This, in turn, 
would make high adult investments in repro- 
duction and semelparity less risky because 
the probability of offspring survival would be 
high. They seem to "live fast and die young." 
Theoretical predictions that the level of invest- 
ment in reproduction by semelparous organ- 
isms should be high, that reduced levels of 
investment in reproduction should extend the 
lives of parents, and that the survival of juve- 



niles should generally be good, are probably 
not valid in cephalopods (Calow, 1987). Moy- 
nihan & Rodaniche (1982) have suggested 
that semelpanty, when it is followed by the 
death and disappearance of breeders, may 
be an effective discouragement to specializa- 
tion by predators; it also leaves more re- 
sources for the offspring. 



DATA ANALYSIS 

Tables 1-6, with data on 547 species, 
clearly demonstrate the enormous variability 
in longevity of the molluscs: from several 
months to well over two hundred years. 

These tables also demonstrate that short 
life spans are a very common strategy 
amongst molluscs. Our present state of 
knowledge is ripe to discuss the short-lived 
category, because it is usually based upon 
clear-cut, firm evidence that gives the species 
a definite life span, plus or minus one or two 
years (at the very most). Our knowledge on 
the long-lived group is still insufficient to en- 
able analysis of variation within this category, 
because many of the data refer to information 
on minimum life spans rather than to actual 
longevity in nature. Our data are sufficient, 
however, to analyse and compare the short- 
lived group, as a whole, to the long-lived one, 
as a whole. 

The intermediate category (consisting of 
species with life-spans of 2-3 years, as de- 
fined in the introduction, and amounting to 7% 
of the species listed) can now be excluded 
from calculations. The resulting picture is 
summarised in Table 7. At the level of both 
species and genus, almost half of the records 
are of short-lived molluscs. 

As a rule, all species within one genus are 
either long- or short-lived. (Exceptions to this 
rule are the marine prosobranchs Acmaea, 
Littorina and Cerithium and the marine bi- 
valve Donax.) This fact enables a stepping-up 
of the taxonomic level to that of genera. By 
doing so, we gain a firmer taxonomic ground. 
We also overcome the danger of distortion 
due to the fact that in certain genera very 
many species have been studied, and are 
thus over-represented in the literature. 

Are there any morphological or environ- 
mental factors in which the short-lived genera 
differ from the long-lived ones? Should this 
high frequency of short-lived genera be con- 
sidered as a representative picture of the mol- 



264 



LONGEVITY IN MOLLUSCS 



TABLE 7. Number of short- and long-lived molluscs 



SHORT-LIVED 



LONG-LIVED 



36 genera, 73 species 



CHITONS — — 2 genera, 2 species 
MARINE SNAILS 

Prosobranchs 16 genera, 25 species 

Opisthobranchs 37 genera, 63 species 

Pulmonales — — 1 genus, 2 species 

FRESHWATER SNAILS 23 genera, 48 species 3 genera, 6 species 

LAND SNAILS 30 genera, 44 species 22 genera, 26 species 

MARINE BIVALVES 1 1 genera, 21 species 73 genera, 116 species 

FRESHWATER BIVALVES 4 genera, 19 species 12 genera, 28 species 

CEPHALOPODS 16 genera, 26 species 1 genus, 1 species 

TOTAL: 137 genera, 246 species 150 genera, 254 species 



Notes to Table 7: 

1. "Long-lived" refers only to molluscs that live four years or more. "Short-lived refers to molluscs that live up to two years, 
and also to those that reproduce throughout one season only, regardless of their life-span. 

2. The data are from Tables 1-6, and from the text. 

3. In the manne prosobranch category, the mixed" genera Acmaea. Notacmea. Littonna and Centhium are counted twice: 
as short- and as long-lived. 



luscs in general? The following sections ex- 
plore these questions. 

Longevity and shell morphology 

Could it be that the life-span of a mollusc is 
associated with the presence or absence of a 
well-calcified external shell? 

To answer this question, each genus was 
classified into one of three shell categohes: 
shell fully calcified and opaque; shell consist- 
ing mainly of conchiolin, with very little cal- 
cium in it and semi-transparent; shell reduced 
to such an extent that the snail cannot retract 
into it, or that it is internal or totally absent. 
Most terrestrial and marine molluscs fall eas- 
ily into either of these three categories, but 
there is, of course, a continuum between the 
opaque and semitransparent shells, and the 
distinction between the two is, to a certain 
extent, arbitrary. 

The results are given in Table 8, which pre- 
sents the shell morphology in long-lived (ex- 
cluding the intermediate category) and short- 
lived molluscs. 

In marine prosobranchs, only two genera, 
Enteroxenos and Thyonicola, lack an external 
shell, and only Lacuna and Patina have a 
semitransparent one. All other marine proso- 
branchs have opaque, well-calcified shells. It 
is unfortunate that no data are available on 
the Lamellariidae and the Heteropoda, two 
other groups of shell-less prosobranchs. 

The majority of opisthobranchs listed in Ta- 
ble 1 эге shell-less. Genera with semi-trans- 
parent shells are Limacina. Cavolina. Clio. 



Creseis. Cuvierina and Diarca. The only ge- 
nus with an opaque, external calcified sfiell is 
Pupa. 

Land snails belonging to the shell-less cat- 
egory include Arion. Bielzia. Catinella. Dero- 
ceras. Eucobresia. Umax. Milax. Omalonyx, 
Parmacella. Semilimax. Testacella. Vaginulus 
and Vitnna. Those belonging to the interme- 
diate category, with semitransparent shells, 
include Aegopinella. Carychium. Elona. Mo- 
nacha and Oxychilus. All other genera have 
opaque, external shells. Comfort (1957) men- 
tions Geomalacus as living seven years, 
based upon animals studied in captivity. If this 
observation does indeed reflect longevities in 
natural populations and if, on the other hand, 
the weak evidence for a short life span in 
Veronicella is valid, then for shell-less snails 
the ratio between short-lived and long-lived 
genera would be 15:2 (as compared to gen- 
era with opaque shells, where the ratio is 12: 
21). Cochlicopa and Euconulus are not in- 
cluded in Table 2 because our present 
knowledge of their longevities places them 
within the intermediate group. If they do in- 
deed live more than three years, then for 
semitransparent snails the ratio between 
short-lived and long-lived genera would be 
4:2, an intermediate position between the 
shell-less and the opaque-shelled landsnails. 

Most marine bivalves are well-calcified. 
The only totally naked marine bivalve is 
Chlamydoconcha (Chlamydoconchidae), in 
which the shell is completely enclosed by the 
mantle. No data on its longevity were found. 
Shipworms (Teredinidae) are virtually shell- 



HELLER 

TABLE 8. Relation between shell and life span in mollusc genera 



265 



A. MARINE SNAILS (PROSOBRANCHS) 

SHELL opaque 

semitransparent 
no external shell 

B. MARINE SNAILS (OPISTHOBRANCHS) 



SHELL 



С LAND SNAILS 
SHELL 



D. MARINE BIVALVES 
SHELL 



E. CEPHALOPODS 
SHELL 



opaque 

semitransparent 
no external shell 



opaque 

semitransparent 
no external shell 



opaque 

semitransparent 
no external shell 



opaque 

semitransparent 
no external shell 



short-lived 



12 
2 
2 



1 

6 

30 



12 

4 

12 



9 
2 

16 



LIFE-SPAN 



long-lived 



36 



21 
1 



74 
1 



less, with a body that resembles a worm: The 
shell is greatly reduced, has lost its protective 
function and become an effective drilling tool 
for boring into wood. Soft-shelled clams (Mya 
and Panopea) have large siphons that are 
permanently extended, being much too large 
to be accommodated within the shell. How- 
ever, their valves are large, opaque and cal- 
cified to such an extent that I have placed 
them in the opaque category. The major 
semitransparent family is the Pinnidae (fan 
mussels), in which the valves consist largely 
of flexible organic conchiolin. Although Pinna 
atropurpúrea ( = P. bicolor) may perhaps be 
annual in Hong Kong (Wu, 1985), in Australia 
it lives substantially more than three years, 
and may well reach 12 years of age (Butler & 
Brewster, 1979). 

There are about 150 genera of cephalo- 
pods (Voss, 1977). Except for one {Nautilus. 
which has an external, calcified shell), the 
shell of all cephalopods is internal and re- 
duced (squids), or absent (octopuses). 

The statistical analysis of the data was car- 
ried out twice, and Fischer's exact test for in- 
dependence in 2 X 2 contingency tables 
(Sokal & Rohlf, 1981) was applied in both 
cases. First, for simplicity, the three shell cat- 
egories were lumped into two: shell present 



(categories 1 and 2) and shell absent (cate- 
gory 3). The frequency of the shell-less gen- 
era among the short-lived molluscs was sig- 
nificantly higher than their frequency among 
long-lived ones, among marine gastropods 
(prosobranchs alone, or prosobranchs and 
opisthobranchs combined), land snails and 
cephalopods. The frequency of shell-less 
genera that are short-lived is significantly 
higher than those that are long-lived (P = 5.1 
X ^0^\ Fisher's exact test). 

Next, categories 1 and 2 were separated 
and the Fisher's exact test again applied. The 
frequency of genera that have semitranspar- 
ent shells among the short-lived molluscs is 
significantly higher than among the long-lived 
ones, in marine gastropods (P = 0.0405) and 
land snails (P = 0.0380) separately, and for 
marine and land snails combined (P = 
0.0013). Amongst marine bivalves, there are 
no significant differences. 

To sum up Table 8, out of 49 mollusc gen- 
era without an external shell, 99% are short- 
lived; of 13 mollusc genera with a shell that is 
external but poorly calcified, 92% are short- 
lived; and of 165 genera with an external, 
well-calcified shell, only 21% are short-lived. 

The only freshwater shell-less molluscs I 
know of are the acochlidiacean genera Aco- 



266 LONGEVITY IN MOLLUSCS 

TABLE 9. Relation between shell morphology and habitat type 



HABITAT 



fully-exposed 



not fully-exposed 



LONG-LIVED LAND SNAILS 

shell solid 

shell transparent or absent 

SHORT-LIVED LAND SNAILS 

shell solid 

shell transparent or absent 



20 
1 



5 
16 



chlidium and Tantulum, found on a few is- 
lands in the Pacific and on one island in the 
Caribbean (Rankin, 1979). I found no data on 
their longevity. Shipworms, though normally 
requiring marine conditions for successful 
spawning, are occasionally recorded from in- 
land waters (Nair & Saraswathy, 1971). 
Freshwater molluscs are not represented in 
this calculation since their classification into 
calcified versus semitransparent genera runs 
into difficulties. Apparently some individuals 
within a genus may be opaque and others 
semitransparent. Whereas in Israel many 
genera are semi-transparent (Valvata, Bithy- 
nia, Hydrobia. Semisalsa. Pseudamnlcola. 
Galba, Stagnlcola, Radix, Ancylus, Ferrlssia, 
Bullnus, Planorbis, Segmentlna. Gyraulus. Bl- 
omphalarla, Hellsoma, Physella, PIsldlum), in 
Europe or North America these same genera 
may be opaque. It is unfortunate that much of 
the taxonomic literature does not refer to this 
trait in sufficient detail. A very welcome ex- 
ception is the study of Fretter & Graham 
(1978), who describe the following freshwater 
prosobranchs as semitransparent: Pota- 
mopyrgus, Pseudamnlcola, BIthynla. As for 
VIvlparus, V. contectus is described as semi- 
transparent and V. vivlparus as opaque. As 
for VIvlparus ater, snails from Lake Maggiore 
have partially dissolved shells whereas those 
of Lake Zuhch do not (Ribi & Gebhardt, 
1986). Since I am not confident that the clas- 
sification of genera or even species into 
opaque and semitransparent categories is 
consistent amongst freshwater molluscs, they 
(and the amphibious genus Succlnea) are 
omitted from the present analysis. Omitting 
the snails is not very significant because most 
of them (88% of the species) live less than 4 
years anyway, regardless of whether they be- 
long to the first shell category or the second. 
As concerning bivalves, however, this is 
rather unfortunate, because their longevities 
in freshwater range from less than one year to 



well over a century. It should at least be noted 
that all long-lived genera of freshwater bi- 
valves are unionids, and are well-calcified — 
what one would indeed expect from the lon- 
gevity pattern in the marine and terrestrial 
environments. 

Land snails: Life span and habitat 

Whereas amongst marine molluscs the ma- 
jority of short-lived genera are shell-less or 
with semitransparent shells, amongst terres- 
trial molluscs over 35% of the short-lived gen- 
era have well-calcified shells. Further infor- 
mation concerning these genera is gained 
when their habitat is considered. To examine 
whether the life-span of a terrestrial snail is 
associated with the environment in which it 
lives, and whether short-lived genera occupy 
a different micro-habitat than that of the long- 
lived ones, each genus was classified into 
one of two habitat categories (as described in 
literature): (1) Genera frequently exposed to 
heavy solar radiation. This includes molluscs 
that sit out on the tips of the vegetation, where 
they are fully exposed to the sun even when 
aestivating. (2) Genera not exposed to solar 
radiation, or found in habitats with intermedi- 
ate exposure to the sun. This includes all gen- 
era that are crevice-dwellers, litter-dwellers, 
or that sit in the more concealed, shady parts 
of vegetation or on shady parts of trees. 

The results are given in Table 9. 

Long-lived land snail genera that sit out on 
the vegetation include Cerlon. Intermediate 
genera (not presented in Table 8) include Tro- 
choldea. Short-lived genera include Brephu- 
lopsls, Bullmulus, Catlnella. Cernuella, Hell- 
cella. Monacha, Theba and Xeroplcta. 

Statistical analysis of the data given in Ta- 
ble 9 reveal that the frequency of the species 
that are both exposed and calcified among 
the short-lived land snails is significantly 
higher than the frequency of the species that 



HELLER 



267 



are both exposed and calcified among the 
long-lived ones (P = 0.0672, by Fisher's ex- 
act test). 



Life span and shell size 



Ten of the gastropods with opaque shells 
surveyed in this study are very minute (i.e. the 
reproducing adult is less than 4 mm). 
Amongst the land snails, Carychium is less 
than 2 mm. Vertigo less than 3 mm, and 
Punctum less than 2 mm. Amongst mahne 
prosobranchs, Rissoa parva is 3-4 mm, Ske- 
neopsis reachs 2 mm, Omalogyra 1 mm, Ris- 
soella 2 mm, Barleeia 3 mm, Littorina ne- 
glecta reaches 2-3 mm, and Littorina 
acutispira usually up to 2 mm. All are short- 
lived. 



DISCUSSION 

The data presented so far allow for state- 
ments about several patterns of longevity 
among molluscs. 

One general pattern concerns the associa- 
tion between the loss of a mollusc's calcified 
shell on the one hand and its short life-span 
on the other. Molluscs in which the shell has 
become internal or lost, and frequently also in 
those in which the shell is external but has 
lost its calcification or becomes rudimentary, 
are short-lived. This relationship holds true 
whether the mollusc is a gastropod (proso- 
branch, opisthobranch or pulmonate) or a 
cephalopod; whether it lives in the sea or on 
the land; whether its mode of reproduction in- 
volves gonochorism or hermaphroditism, 
planktonic larvae or hatchlings that resemble 
adults; whether it moves by crawling, jet-pro- 
pulsion or is sedentary; and whether it feeds 
as a herbivore, carnivore or omnivore. 

Describing correlations is one thing, ex- 
plaining them is another matter. Correlations 
can be explained in many ways, and the ap- 
proach may be adaptive or non-adaptive, 
each with its drawbacks. 

The relation between shell absence and 
longevity may be explained in adaptive terms, 
in that shell absence affects age-specific mor- 
tality directly. It enables high growth rates and 
juveniles of shell-less molluscs grow to adult 
size quicker than shelled ones, speeding 
through the vulnerable juvenile phase. Once 
they reach adult size their survival chances 
are similar to those of their parents, and since 
semelpahty is favoured whenever the survival 



chances of the parents and offsphng are sim- 
ilar (Calow, 1981), semelpahty will eventually 
indeed develop. 

Differences in growth rates between 
shelled and shelless molluscs do exist. 
Among terresthal molluscs for example, a 
slug such as Deroceras reticulatus matures 
within the first year, breeds in the second and 
then dies (Runham & Hunter, 1970), whereas 
a shelled landsnail such as Arianta arbusto- 
rum matures within two years, breeds and 
may then live on for another ten (Baur & 
Raboud, 1988). Similarly among cephalo- 
pods, a 120 mm-long squid can mature at the 
age of six months (Moynihan & Rodaniche, 
1982), whereas Nautilus matures within sev- 
eral years. From these aspects, the consis- 
tently faster growth rates of the shell-less mol- 
luscs may indeed be an advantageous trait. 

Whether these rapid growth rates should 
always and consistently lead to semelpahty is 
another question. Such an argument would 
imply ubiquity in the ecology of entire groups 
of shell-less molluscs, which is difficult to ac- 
cept. It is not reasonable to assume that all 
2,500 molluscan species — which have had 
different taxonomic origins ever since the 
Palaeozoic, which live in environments as dif- 
ferent as a whole spectrum of habitats in the 
sea and on land, which practice a wide scope 
of reproductive strategies ranging from plank- 
tonic veligers to direct development (with or 
without parental caring of eggs), which are 
either hermaphroditic or gonochoristic — 
should always and consistently practice a 
semelparous reproductive strategy only be- 
cause, since they enjoy a faster growth rate, 
their survival chances come to resemble 
those of their parents at an earlier age. The 
advantages to be gained from semelparity 
must surely be overwhelming if such a gen- 
eral correlation, cutting through an entire an- 
imal phylum, should be explained on its se- 
lective basis. 

Another, somewhat similar adaptive ap- 
proach to the relation between shell absence 
and longevity could be that when extrinsic 
mortality nsks (such as starvation, accident, 
disease or prédation) are higher for parents 
than for offspring, it pays the parent to in- 
crease its investment in the reproduction of 
many offspring. This increase would eventu- 
ally lead to a semelparous reproductive strat- 
egy (Calow, 1981, 1984). As applied to mol- 
luscs, this means that shell absence directly 
affects age-specific mortality: If the shell-less 
mollusc (slug, octopus or opisthobranch) 



268 



LONGEVITY IN MOLLUSCS 



were to live on after reproduction, then its sur- 
vival chances would be very low as compared 
to its progeny, due to such environmental fac- 
tors. 

This age-specific-mortality argument can 
definitely be applied to many opisthobranchs, 
in which the parent feeds upon food that is 
transient, whereas the juveniles feed upon 
another source. Thus in Aplysiamorpha and 
Sacoglossa, the adult feeds upon seasonally 
abundant green seaweed, whereas the juve- 
nile is a planktonic veliger that feeds upon 
unicellular algae (Kandel, 1979; Carefoot, 
1987). Among the bivalves, shipworms offer 
another excellent example of a mollusc utiliz- 
ing a transient habitat. They have rapid 
growth rates, reach an early maturity within 
3-6 weeks and have very high reproductive 
rates (Nair & Saraswathy, 1971; Turner, 
1973). 

Again however, whether shell-less mol- 
luscs always and consistently feed upon tran- 
sitional prey is another question. Opistho- 
branchs feed upon a wide variety of prey 
(hydrozoans, sponges, polychaetes, gastro- 
pods, bivalves, ascidians, sessile barna- 
cles — see Thompson, 1976), and many of 
these food resources are rather stable and 
not of a transient nature. Octopuses and cut- 
tlefish are opportunistic carnivores that feed 
upon shrimps, prawns, crabs, polychaetes, 
bivalves, gastropods and fishes (Boucaud- 
Camou & Boucher-Rodoni, 1983), food re- 
sources that are stable rather than transient. 
Slugs eat dead leaves, stems, bulbs, tubers, 
fungi, lichens and algae (Runham & Hunter, 
1970), a diet similar to that of shelled land- 
snails. It is questionable whether all of these 
food resources are indeed transient, but even 
if they are, this does not explain the question 
but merely rephrases it into "why are the 
shell-less molluscs, whether herbivores, om- 
nivores or carnivores, capable of feeding only 
upon transient, rather than stable re- 
sources?" This is back almost to the starting 
point. 

Extrinsic mortality risks include also préda- 
tion, and it could perhaps be argued that 
groups in which the extent of prédation in- 
creases with adulthood are likely to be short- 
lived. To make such a claim acceptable, some 
sort of evidence must be presented that 
shows that in nature, prédation pressures on 
adult shell-less molluscs are indeed greater 
than those on their progeny, thereby lowering 
their survival chances. Together with such 
data, additional evidence must also be pre- 



sented that adult shelled molluscs do not face 
such severe prédation risks. At present, I do 
not know of such evidence. 

To conclude, it should be re-emphasized 
that the question emerging from the data 
analysis is not whether some molluscs are 
short-lived, but why all shell-less molluscs are 
short-lived. The disadvantage of the adaptive 
approach is that it does not cope with the 
ubiquity of the relation between shell and lon- 
gevity, and when the whole spectrum of shell- 
less molluscs is considered, it loses much of 
its attractiveness. 

The ubiquity of the relation may be ex- 
plained in non-adaptive terms; A short life 
span may be a byproduct of selection on the 
shell, rather than an independently selected 
trait. Shell and longevity may covary so that 
an adaptive change in the shell engenders an 
automatic switch in longevity, the latter being 
irrelevant to adaptation and not under imme- 
diate control of the environment. 

Loss of the shell occurred independently in 
several molluscan lineages, as a result of a 
wide variety of selective forces that, at least 
as considered today, have very little to do with 
life cycles. In mahne prosobranchs, predatory 
pressure by crabs and fish has resulted in the 
survival of heavy, ridged or spiny shells (Ver- 
meij, 1978). Pressure on marine cephalopods 
to form a very light, buoyant animal capable of 
swimming actively in the water body (rather 
than passively drifting with the currents in a 
flying-balloon, Nautilus-s\y\e) has led to the 
persistence of those with an internal shell, or 
with no shell at all. Opisthobranchs' initial ex- 
ploitation of the infaunal (burrowing) environ- 
ment by the primitive order Bullomorpha, 
combined with their development of chemical 
defence supplied by the integument to re- 
place the mechanic defense supplied by the 
shell, led to the reduction of the shell and its 
eventual loss (Thompson, 1976). In such 
planktonic opisthobranchs as the Euthecoso- 
mata, a (transparent) shell is retained, how- 
ever, and functions as a retreat into which the 
animal withdraws, so as to sink and thereby 
rapidly avoid predators (Be & Gilmer, 1977). 
In terrestrial molluscs, the ability for deeper 
penetration into the ground and, in addition, 
the invasion of calcium-deficient, moisture- 
rich environments has been the outcome of 
developing a shell-less slug form in several 
unrelated taxonomic families (Solem, 1978). 
Alternatively, the slug form may have devel- 
oped through the habit of climbing up trees 
(Cain, pers. comm.). Once the shell is lost, in 



HELLER 



269 



any of these mollusc lineages and for any of 
these selective reasons, a mollusc will auto- 
matically become short-lived. 

Within the severe limits of a short life span, 
life history strategies vary in evolutionary re- 
sponse to different environmental conditions. 
For example, some of the species of British 
nudibranchs have fully annual life cycles with 
one breeding period, whereas others pass 
through numerous generations a year. The 
purely annual species feed on organisms that 
are abundant and stable throughout all sea- 
sons of the year, whereas those passing 
through a number of generations a year are 
species that feed, upon seasonal, transitory 
prey (Thompson, 1976). Seasonal food short- 
age may thus determine the relatively short 
life span of the one, as compared to a stable 
food supply that determines the slightly longer 
life span of the other. However, both have an 
overall short life span that does not exceed 
one year, two at the most. When we consider 
them together, as one single category, in 
comparison to the long-lived (and shell-pos- 
sessing) prosobranchs, these differences be- 
tween them seem trivial. This non-adaptive 
approach may suggest that shell absence is 
the overriding factor in determining whether a 
mollusc will be short- or long-lived. Once this 
major factor is set and the mollusc becomes 
short-lived, then environmental factors deter- 
mine the fine tuning. 

The non-adaptive approach may suggest (in 
a very schematic and over-simplified manner) 
that many molluscs, short-lived because they 
lost their shells, invaded the "transient food 
niche" where there is less competition from the 
long-lived (shelled) molluscs. However, a tran- 
sient food niche is not a prerequisite of shell- 
lessness, and many short-lived (shell-less) 
molluscs may enjoy a stable food niche, their 
short life-spans bearing no direct relevance to 
their food resources (and vice-versa). 

The advantage of this non-adaptive ap- 
proach is that it copes well with the ubiquity of 
the relationship between shell absence and 
short longevity. 

One severe weakness of the non-adaptive 
approach is that it requires a nearly single- 
gene linkage between shell-lessness and lon- 
gevity. There is, as yet, no direct evidence for 
any such link. 

A further weakness is that it cannot explain 
exceptional records, of shell-less molluscs 
that are not short-lived. This includes the ter- 
restrial slug Testacella, if we restrict ourselves 
to longevity records based upon evidence 



from natural populations. If also records of 
molluscs reared in captivity are accepted, 
then this also includes the landsnail Geoma- 
lacus (Comfort, 1957). Data on the longevity 
of slugs from additional pulmonate families, 
such as the Helicarionidae, Charopidae, 
Athoracophondae and Endontidae, could also 
help, since long life spans in these families 
would weaken the non-adaptive approach 
considerably, at least as a phenomenon that 
sweeps through the entire phylum. 

It should be emphasized that whereas for 
gastropods we have data on 45 shell-less gen- 
era, for bivalves we have data for only two. 
Both are short-lived, but this is obviously not 
nearly enough to enable generalisations about 
the entire class. A reminder as to why (in this 
respect) the bivalves should be approached 
cautiously comes from Nausitora fusticula. a 
large oviparous teredinid of tropical man- 
groves. Collected as a fully grown adult (age 
unknown), a specimen of this species lived for 
two and a half years in an aquarium at Harvard 
University (R. Turner, pers. comm.). 

To conclude, the disadvantages of the non- 
adaptive approach is that it lacks, as yet, ge- 
netic support and that it relies very heavily 
upon the ubiquity of the association, so that it 
cannot explain exceptions. 

A second pattern to emerge from the data 
in this study is that the life history of well- 
calcified molluscs is influenced by the tem- 
perature of their environment. A short life- 
span may occur in well-calcified molluscs that 
live in very hot environmental conditions. The 
rate of gamete development is directly depen- 
dent on temperature, and high temperatures 
increase the rate of gonad maturation. The 
scallop Argopecten irradians, for example, 
matures within 12 months in its natural habi- 
tat, but maturing can be accelerated by labo- 
ratory exposure to higher temperatures, and It 
then reaches reproductive stage within six 
months (Sastry, 1979). A warm environment 
will also enable the rapid growth of juvenile 
gastropods (see Runham & Hunter, 1970, for 
slugs; Geraerts & Joosse, 1984, for freshwa- 
ters gastropods). In the marine environment, 
the term hot applies to geographically wide- 
spread species, where populations from trop- 
ical waters complete their life-span in a much 
shorter time (Cerrato, 1980; Hadfield & Swit- 
zer-Dunlap, 1984). In the freshwater environ- 
ment, it applies to snails and bivalves that live 
in relatively warm waters (Geraerts & Joosse, 
1984; Mackie, 1984). In the terrestrial envi- 
ronment, my study suggests that it refers to 



270 



LONGEVITY IN MOLLUSCS 



those micro-habitats in which the land snails 
sit out on the vegetation, where they are fully 
exposed to solar radiation. Low temperature, 
on the other hand, can stretch a semelparous 
cycle that is annual in the warmer parts of a 
species' range, into a biennial one in the 
colder parts. (Theba pisana from Israel as 
compared to that of England is one such 
case; see Heller, 1982, and Cowie, 1984. Ari- 
anta arbustorum of the lower Alps as com- 
pared to that of higher altitudes is another 
example; Baur & Raboud, 1988.) The short 
life-span of molluscs of hot micro-habitats 
may be a phenotypic response to the environ- 
ment, or it may be a genetically controlled 
trait, subject to selection. 

In land snails, the relation between longev- 
ity and exposure may be explained in adap- 
tive terms, in that ionizing radiation increases 
the rate of ageing and reduces the average 
life-span of animals. Experimental evidence 
reviewed by Comfort (1978) showed that 
large doses of hard radiation (gamma and 
fast neutrons) shorten life considerably. 
These conclusions should be approached 
with caution, however, since the experiments 
made use of extremely heavy doses that ex- 
ceed natural quantities reaching the earth by 
several orders of magnitude. The effects of 
ultraviolet radiation on molluscs are as yet un- 
known, but upon entering a reptile's body ul- 
traviolet radiation can cause a breakdown of 
molecules and thereby alter vital biochemical 
processes (Porter, 1967). A short life span 
could thus be enforced in landsnails dwelling 
on the tips of vegetation, where they are sub- 
ject to heavier ultraviolet radiation than snails 
dwelling underneath stones. Even visual day- 
light radiation may be an important environ- 
mental factor that influences the gonad of 
land snails. Continuous illumination of the 
slug Deroceras reticulatum for five weeks in- 
creases the thickness of the germinal epithe- 
lium, rate of meiosis and also the numbers of 
Sertoli cells, male gametes, multinuculated 
spermatids, and it upsets in cytokinesis (Pari- 
var, 1978, and references therein). 

An exception to this generalisation that land 
snails of warm, strongly radiated environ- 
ments are short-lived concerns snails that in- 
habit environments that, in addition to being 
hot, are also extremely unpredictable, such 
as the shadeless vegetation of deserts. In 
such hot surroundings, where conditions for 
growth and reproduction are both very infre- 
quent and unpredictable, semelparous an- 
nual populations would quickly become ex- 



tinct. Molluscs of these habitats may be 
expected to be more long-lived than their 
close relatives from more favourable condi- 
tions. Short-lived molluscs are restricted, ac- 
cordingly, to environments in which there is a 
predictable weather. 

A third pattern to emerge from the data in 
this study is that among the shell-possessing 
gastropods, longevity is related to size. 
Though the linear relationship between body 
size and life-span found in mammals (Kohn, 
1971) definitely does not occur in molluscs, 
short life spans appear to occur more fre- 
quently among very minute gastropods than 
among larger ones. Every single one of the 
very minute gastropods found in this study are 
short-lived. (This does not apply to the bi- 
valves, in which minute genera, such as My- 
sella, may live for six years.) Furthermore, to 
the extent that the short-lived group can be 
divided into semelparous animals on the one 
hand and iteroparous animals on the other, 
most of these small snails belong to the latter 
group: They mature within several weeks early 
in the season, lay several eggs (up to about 
30) throughout the remainder of the season, 
and gradually die off by the end of the season 
(Fretter, 1948; Morton, 1954; Southgate, 
1982; Hughes, 1986; Baur, 1987; Pokryszko, 
pers. comm). Littorina acutispira (see Under- 
wood & McFadyen, 1983) and Rissoa repro- 
duce by planktonic veligers, and consequently 
have more juveniles than the rest of this group. 
Marine gastropods that are both well-calcified, 
short-lived and produce a very large number of 
offspring (such as Cerithium scabridum: see 
Ayal, 1978) are infrequent. 

A fourth pattern to emerge is that bivalves 
are more long-lived than other groups. 
Though they form only 40% of the species 
listed in the longevity tables, they constitute 
88% of the species that live over 25 years, 
92% of the species that live over 50 years, 
and they are the only molluscs that live over a 
century. Short life-spans are not a very com- 
mon strategy amongst bivalves, and only 
15% of the bivalve genera are short-lived (as 
compared to short life spans in 63% of the 
gastropods). A sedentary mode of life appar- 
ently bears the potential for a long life-span. 



CONCLUSIONS 

This paper paints the longevity pattern of 
molluscs in very broad strokes. To conclude, 
we can to a certain extent generalize about 



HELLER 



271 



the relationship between a mollusc's longev- 
ity, its morphology and its environment. 

Bivalves are the most long-lived of mol- 
luscs. 

Amongst bivalves, prosobranchs and pul- 
monates, short life-spans are more common 
in the freshwater than in the marine or terres- 
trial environment. 

In the terresthal and mahne habitat, a 
short-lived mode of life is often correlated 
with: 

1. Lack of an external shell. 

2. Possession of an external shell that 
is semitransparent. 

3. Dwelling in a micro-environment that 
is exposed to high solar radiation and 
to high temperatures. (In cold 
environments, on the other hand, the 
semelparous cycle of molluscs 
without external shells may be 
stretched, over two years or more.) 

4. Dwelling in an environment that is 
predictable to such an extent that 
conditions favourable for reproduction 
occur (for an annual species) at least 
once a year. 

5. Very minute size (in gastropods). 

Of these generalizations, the correlation 
between shell absence and longevity ac- 
counts for the greatest number of short-lived 
molluscs. 

When combined together, these correla- 
tions, of shell-morphology exposure and 
minute size, account for 84% of the short- 
lived marine snails, and 93% of the land 
snails mentioned in this study. Confining our- 
selves to the gastropods, we can now apply 
these correlations to predict, very approxi- 
mately, the number of short-lived genera. In 
the terresthal habitat, present information 
concerns mainly Europe. If we combine, from 
Kerney & Cameron (1 979), all the genera that 
are shell-less, have semitransparent shells, 
dwell in exposed habitats where they sit out 
on the vegetation, or are very minute (less 
than 4 mm) and assume that all these snails 
are short-lived, then as a very rough and 
broad estimate, half of the genera of Bhtain 
and northwestern Europe may be short-lived. 
Similar calculations reveal that about half of 
the 45 genera of the Mediterranean region of 
Israel may be short-lived, whereas in the Ne- 
gev Desert, where slugs and snails with semi- 
transparent shells cannot survive because of 
the dangers of desiccation, only one of the 



nine genera is short-lived and another one is 
intermediate. 

The empiric rules proposed in this paper 
are based upon evidence from 547 mollusc 
species. Future research will probably modify 
them considerably. 



ACKNOWLEDGMENTS 

Dr. B. Baur, Dr. M. Lazahdou-Dimithadou, 
Dr. B. Pokryszko, Dr. N. Runham, Dr. Y. 
Steinberger and Prof. R. Turner kindly permit- 
ted me to quote from as-yet unpublished data. 
I am endebted to Dr. U. Metro for carrying out 
the statistics. 

In writing a paper that discusses longevity 
pattern throughout an entire phylum and re- 
views data of five hundred species, it is im- 
possible not to make errors. I thank Dr. B. 
Baur, Prof. A. J. Cain, Prof. P. Calow and 
Prof. S. Stearns for criticizing earlier versions 
and weeding out many mistakes. The remain- 
ing mistakes are all, of course, my own re- 
sponsibility. 

I owe special thanks to Mr. M. Hallel, for 
assisting me in the whting of this paper. 



LITERATURE CITED 

ALDRIDGE, D. W., 1982. Reproductive effort in re- 
lation to life-cycle bioenergetics in three natural 
populations of the freshwater snail Leptoxis car- 
inata. Ecology, 63: 196-208. 

ANDERSON, R. C, 1987. Field aspects of the se- 
piolid squid Rossia pacifica Berry, 1911. Annual 
Report of thie Western Society of f\/lalacologists, 
20: 30-32. 

ANSELL, A. D., 1983. The biology of the genus 
Donax. In: McLachlan, A. & T. Erasmus, eds., 
Sandy beaches as ecosystems. Junk, The 
Hague. 

ANSELL, A. D. & A. H. PARULEKAR, 1978. On the 
rate of growth of Nuculana minuta (Müller) (Bi- 
valvia: Nuculanidae). Journal of tJlolluscan Stud- 
ies. 44: 71-82. 

APLEY, M. L., 1970. Field studies on life-history, 
gonadal cycle and reproductive periodicity in 
r\/lelampus bidentatus (Pulmonata: Ellobiidae). 
t\/lalacologia. 10: 381-397. 

APPLETON, С. С, 1978. Review of literature on 
abiotic factors influencing the distribution and life 
cycles of bilharziasis intermediate host snails. 
Malacological Review, 1 1 : 1-26. 

ARNOLD, J. M. & L. WILLIAMS-ARNOLD, 1977. 
Cephalopoda: Decapoda. Pp. 227-241 in 
GIESE, A. & J. PEARSE, eds., Reproduction of 



272 



LONGEVITY IN MOLLUSCS 



marine invertebrates, vol. 4. Academic Press, 
N.Y. & London. 

AVOLIZI, R. J., 1976. Biomass turnover in popula- 
tions of viviparous sphaeriid clams: comparisons 
of growth, fecundity, mortality and biomass pro- 
duction. Hydrobiologia. 51: 163-180. 

AVAL, Y., 1978. Geographical distribution, ecolog- 
ical niche and the strategy of reproduction of the 
colonizer Cerithiium scabridum Phil. (Gas- 
tropoda: Cerithiidae) as compared with those of 
some other sympatric non-colonizing species. 
Ph.D. thesis, the Hebrew University, Jerusalem 
(Hebrew, English summary). 

BAAS, J. A., 1979. Growth and fecundity of Pisid- 
ium amnicum (Müller) (Bivalvia: Sphaendae) in 
the Tandoll Brook, Dorset, England. Journal of 
Conchology. 30: 129-134. 

BAKER, R. E., 1965. Catinella arenaria Bouchard 
Chantereux at the Braunton Burrows National 
Reserve, Devon. Proceedings of tf^e Malacolog- 
ical Society of London, 36: 259-265. 

BAKER, R. E., 1968. The ecology of the wnnkled 
snail Helicella caperata Mont, on the Braunton 
Burrows sand system. Proceedings of the f\/lala- 
cological Society of London, 38: 41-54. 

BAUER, G., 1987. Reproductive strategy of the 
freshwater pearl mussel t^argantifera margaritif- 
era. Journal of Animal Ecology, 56: 691-701. 

BAUR, В., 1987. The minute land snail Punctum 
pygmaeum (Draparnaud) can reproduce in the 
absence of a mate. Journal of f\/lolluscan Studies, 
53: 112-113. 

BAUR, B. & J. BENGTSSON, 1987. Colonizing 
ability in land snails on the Baltic uplift archipel- 
agos. Journal of Biogeography, 14: 329-341 . 

BAUR, B. & C. RABOUD, 1988. Life history of the 
land snail Arianta arbustorum along an altitudinal 
gradient. Journal of Animal Ecology, 57: 71-88. 

BE, A. W. & R. W. GILMER, 1 977. A Zoogeographie 
and taxonomic review of euthecostomatous 
Pteropoda. Pp. 733-808 in RAMSEY, T. S., ed.. 
Oceanic micropaleontology, vol. 1. Academic 
Press, London. 

BERGSTROM, В., & W. С. SUMMERS, 1983. 
Sepietta oweniana. Pp. 75-91 in BOYLE, P. R., 
ed., Cephalopod life cycles, vol. 1. Academic 
Press, London. 

BERRIE, A. D., 1965. On the life cycle of Lymnaea 
stagnalis (L.) in the west of Scotland. Proceed- 
ings of thie Malacological Society of London, 37: 
83-92. 

BERRIL, N. J., 1931. The natural history of Bulla 
tiydatis Linn. Journal of thie Marine Biological As- 
sociation of the United Kingdom, 1 7: 567-571 . 

BERRY, A. J. & A. H. KADRI, 1974. Reproduction 
in the Malayan freshwater cerithiacean gastro- 
pod Melanoides tuberculata. Journal of Zoology, 
172: 369-381. 

BETT, J., 1960. The breeding season of slugs in 
gardens. Proceedings of the Zoological Society 
of London. 135: 559-568. 

BLAIR, D., & C. M. FINLAYSON, 1981. Observa- 
tions on the habitat and biology of a lymnaeid 



snail, Austropelpa vinosa (Gastropoda: Pulmo- 
nata), an intermediate host for avian schisto- 
somes in tropical Australia. Australian Journal of 
Mahne & Freshwater Research, 32: 757-767. 

BLANKLEY, W. O. & G. M. BRANCH, 1985. Ecol- 
ogy of the limpet Nacella delesserti (Philippi) at 
Marion Island in the sub-Antarctic Southern 
Ocean. Journal of Experimental Marine Biology 
& Ecology, 92: 259-285. 

BLINN, W. C, 1963. Ecology of the land snails Me- 
sodon thyroidus and Allogona profunda. Ecology, 
44: 498-505. 

BOAG, D. A. & W. D. WISHART, 1982. Distnbution 
and abundance of terrestrial gastropods on a 
winter range of bighorn sheep in southwestern 
Alberta, Canada. Canadian Journal of Zoology, 
60: 2633-2640. 

BOATO, A. & M. B. RASOTTO, 1987. Functional 
protandry and seasonal reproductive cycle in So- 
latopupa similis (Bruguiere) (Pulmonata Chon- 
drinidae). Bollettino di Zoológica. 54: 119-125. 

BOLETZKY, S. v., 1983a. Sepia officinalis. Pp. 31- 
52 in BOYLE, P. R., ed., Cephalopod life cycles. 
vol. 1. Academic Press, London. 

BOLETZKY, S. v., 1983b. Sepiola robusta. Pp. 53- 
67 in BOYLE, P. R., ed.. Cephalopod life cycles, 
vol. 1. Academic Press, London. 

BOUCAUD-CAMOU, E. & R. BOUCHER-RODONI, 
1983. Feeding and digestion in cephalopods. Pp. 
149-187 in SALEUDDIN, A. S., & K. M. WILBUR, 
eds.. The Mollusca, vol. 5. Academic Press, N.Y. 

BOYLE, P. R., 1983. Eledone cirrhosa. Pp. 365- 
386 in BOYLE, P. R., ed., Cephalopod life cycles, 
vol. 1. Academic Press, London. 

BOYLE, P. R., 1987. Molluscan comparisons. Pp. 
307-327 in BOYLE, P. R., ed., Cephalopod life 
cycles, vol. 2. Academic Press, London. 

BRANCH, G. M., 1981. The biology of limpets: 
Physical factors, energy flow, and ecological in- 
teractions. In: BARNES, M., ed.. Oceanography 
and Marine Biology, 19: 235-380. 

BRETOS, M., 1980. Age determination in the key- 
hole limpet Fissurella crassa Lamarck (Archaeo- 
gastropoda: Fissurellidae) based on shell growth 
nngs. Biological Bulletin. 159: 606-612. 

BROWN, A. C, 1982. The biology of sandy-beach 
whelks of the genus Bullía. In Barnes, M., ed.. 
Oceanography and Manne Biology. 20: 309- 
362. 

BROWN, D. S., 1980. Freshwater snails of Africa 
and their medical importance. Taylor & Francis, 
London. 

BROWN, K. M., 1983. Do life history tactics exist at 
the intraspecific level? data from freshwater 
snails. The American Naturalist. 121: 871-879. 

BROWN, K. M., 1985. Intraspecific life history vari- 
ation in a pond snail: The roles of population di- 
vergence and phenotypic plasticity. Evolution, 
39: 387-395. 

BROWNE, R. A. & W. D. RUSSELL-HUNTER, 
1978. Reproductive effort in molluscs. Oecologia, 
37: 23-27. 

BRYNE, M., 1985. The life history of the gastropod 



HELLER 



273 



Thyonicola americana Tikassingle, endoparasitlc 
in a seasonally eviscerating holothurian host. 
Ophelia. 24: 91-101. 

BUCKLEY, D. е., 1986. Bioenergetics of age-re- 
lated versus size-related reproductive tactics in 
female Viviparus georgianus. Biological Journal 
of the Linnean Society. 27: 293-309. 

BUTLER, A. J. & F. J. BREWSTER, 1979. Size 
distribution and growth of the fanshell Pinna bi- 
color Gmelin (Mollusca: Eulamellibranchia) in 
South Australia. Australian Journal of toarme & 
Freshwater Research. 30: 25-39. 

CALOW, P., 1978. The evolution of life-cycle strat- 
egies in fresh-water gastropods. t\Aalacologia. 1 7: 
351-364. 

CALOW, P., 1981. Resource utilization and repro- 
duction. In: TOWNSEND, С R. & P. CALOW, 
eds., Physiological ecology, an evolutionary ap- 
proach to resource use. Blackwell Scientific Pub- 
lications, Oxford. 

CALOW, P., 1984. Explonng the adaptive land- 
scapes of invertebrate life cycles. In: ENGELS, 
W., ed.. Advances in invertebrate reproduction. 
Elsevies, Amsterdam. 

CALOW, P., 1987. Fact and theory — an overview. 
Pp. 351-365 in: BOYLE, P. R., ed., Cephalopod 
life cycles, Vol. 2. Academic Press, London. 

CAMERON, R. A. D., 1982. Life histones, density 
and biomass in a woodland snail community. 
Journal of l^olluscan Studies. 48: 159-166. 

CAREFOOT, T. H., 1987. Aplysia: Its biology and 
ecology. In Barnes, M., ed., Oceanography and 
toarme Biology. 25: 167-284. 

CERRATO, R. M., 1980. Demographic analysis of 
bivalve populations. Pp. 417-468 in RHOADS, 
D., & R. LUTZ, eds., Skeletal growth of aguatic 
organisms. Plenum Press, N.Y. & London. 

CHATFIELD, J., 1968. The life history of the helicid 
snail t^onacha cantiana (Montagu), with refer- 
ence also to /И. cartusiana (Müller). Proceedings 
of the Malacological Society of London. 38: 233- 
245. 

CHILDRESS, J., H. FELDBECK & G. SOMERO, 
1987. Symbiosis in the deep sea. Scientific 
American. 256: 106-112. 

CHIU, S. & K. CHOU, 1962. Observations on the 
biology of the carnivorous snail Euglandina rosea 
Ferussac. Bulletin of the Institute of Zoology, 
Academia Sínica (Taipei), 1: 17-24. 

CHOAT, J. H. & R. BLACK, 1979. Life histones of 
limpets and the limpet-lamminarian relationship. 
Journal of Experimental Marine Biology & Ecol- 
ogy, 41: 25-50. 

CHRISTENSEN, J. M. & S. P. DANCE, 1980. Sea- 
shells. Penguin, Harmondsworth. 

CHUKHCHIN, V. D., 1978. Taxonomic position and 
ecology of Black Sea Hydrobiidae. Malacological 
Review. 11:114. 

COMFORT, A., 1957. The duration of life in mol- 
luscs. Proceedings of the Malacological Society 
of London. 32: 219-240. 

COMFORT, A., 1964. Ageing, the biology of senes- 
cence. London, Routledge & Kegan Paul. 



COMFORT, A., 1978. The biology of senescence. 
Churchill Livingstone, Edinburgh. 

COMELY, С A., 1978. Modiolus modiolus from the 
Scottish west coast. I. Biology. Ophelia, 17- 167- 
193. 

COOK, L. M. & A. J. CAIN, 1980. Population size 
and morph frequency in experimental popula- 
tions of the snail Cepaea nemoralis (L). Biologi- 
cal Journal of the Linnean Society. 1 4: 259-292. 

COWIE, R. H., 1984. The life-cycle and productivity 
of the land snail Theba pisana (Mollusca: Heli- 
cidae). Journal of Animal Ecology, 53: 31 1-325. 

DAGUZAN, J., 1982. Contnbution à l'étude de la 
croissance et de la longévité de Elona quimipe- 
riana (de Férussac) (Gastéropode Pulmoné Sty- 
lommatophore) vivant en Bretagne Occidentale. 
Malacologia, 22: 385-394. 

DUDGEON, D., 1982. The life history of Brotia 
hainanensis (Brot, 1872) (Gastropoda: Proso- 
branchia: Thiaridae) in a tropical forest stream. 
Zoological Journal of the Linnean Society, 76: 
141-154. 

DUDGEON, D., 1986. The life cycle, population dy- 
namics and productivity of Melanoides tubercu- 
lata (Müller 1774) (Gastropoda: Prosobranchia: 
Thiaridae) in Hong Kong. Journal of Zoology, 
208: 37-53. 

DUNCAN, С J., 1975. Reproduction. Pp. 309-365 
in: FRETTER, V. & J. PEAKE, eds.. Pulmonales, 
vol. 1. Academic Press, London & NY. 

DUNDEE, D. S., 1977. Observations on the veroni- 
cellid slugs of the southern United States. Nau- 
tilus. 91: 108-114. 

DURRANT, P. M., 1980. The growth of a population 
of Ancylus fluviatilis Müller from the R. Lea, Hert- 
ford. Journal of Molluscan Studies. 46: 257-264. 

DUSSART, G. В., 1979. Sphaerium corneum (L.) 
and Pisidium spp. Pfeiffer — the ecology of fresh- 
water bivalve molluscs in relation to water chem- 
istry. Journal of Molluscan Studies, 45: 19-34. 

EDWARDS, D. С & J. D. HUEBNER, 1977. Feed- 
ing and growth rates of Polinices duplicatus prey- 
ing on Mya arenaria at Barnstable Harbour, Mas- 
sachusetts. Ecology, 58: 1218-1236. 

EGOROVA, E. N., 1978. Seasonal dynamics of the 
number and some bioenergetic indices of a pop- 
ulation of Falsimargarita iris (Smith) (Gas- 
tropoda: Trochidae) from the Davis Sea (East 
Antarctic). Malacological Review, 11: 106-107. 

EVERSOLE, A., 1978. Life cycles, growth and pop- 
ulation bioenergetics in the snail Helisoma trivolis 
(Say). Journal of Molluscan Studies, 44: 209- 
222. 

FALKNER, G., 1984. Das bayerische Weinberg- 
schnecken-Projekt (Untersuchungen an Helix 
pomatia L.). Mitteilungen der Deutschen Malakol- 
ogischen Gesellschan. 37: 182-197. 

FASHUYI, A., 1981. Observations on the ecology 
of Bulinus (Physopsis) globosus Morelet, B. (Bu- 
linus) forskalii (Ehrenberg) and Lymnaea natal- 
ensis (Kraus) in rice swamps in Sierra Leone. 
Journal of Molluscan Studies, 47: 328-333. 

FLETCHER, W. J., 1984. Intraspecific variation in 



274 



LONGEVITY IN MOLLUSCS 



the population dynamics and growth of the limpet 
Cellana tramosehca. Oecologia. 63: 110-121. 

FOCARDI, S. & D. QUATTRINI, 1972. Structure of 
the reproductive apparatus and life cycle of Milax 
gagates (Draparnaud). fvlollusca, Gastropoda, 
Puimonata. Bollettino di Zoológica. 39: 9-27. 

FORD, J. I., 1987. Biology, microhabitat and in- 
traspecific shell variation of Neritina granosa 
Sowerby (Prosobranchia: Neritidae) in Hawaiian 
streams. Bulletin of Marine Sciences. 41 : 635. 

FOTHERINGHAM, N.. 1 971 . Life history patterns of 
the littoral gastropods Shaskyus festivus (Hinds) 
and Ocenebra poulsoni Carpenter (Prosobran- 
chia: Muncidae). Ecology. 52: 742-747. 

FRANK, P. W., 1969. Growth rates and longevity of 
some gastropod molluscs on the coral reef at 
Heron Island. Oecologia. 2: 232-250. 

FRANZ, D. R., 1972. The ecology and reproduction 
of a marine bivalve, Mysella planulata (Er^'cina- 
cea). Biological Bulletin. 144: 93-106. 

FRETTER, v., 1948. The structure and life history 
of some minute prosobranchs of rock pools Ske- 
neopsis planorbis (Fabricius), Omalogyra ato- 
mus (Phillippi), Rissoella diaphana (Alder) and 
Rissoella opalina (Jeffreys). Journal of the Ma- 
rine Biological Association of the United King- 
dom. 27: 597-630. 

FRETTER, V. & A. GRAHAM, 1978. The proso- 
branch molluscs of Britain and Denmark, Part 3. 
Supplement 5, Journal of Molluscan Studies. 

GAGE, J., 1968. The mode of life of Mysella cu- 
néala, a bivalve commensal with Phascolion 
strombi (Sipunculidea). Canadian Journal of Zo- 
ology. 46: 919-934. 

GALE, W. F., 1977. Growth of the fingernail clam, 
Sphaerlum transversum (Say) in field and labo- 
ratory experiments. Nautilus, 91: 8-12. 

GERAERTS, W. P. M. & J. JOOSSE. 1984. Fresh- 
water snails (Basommatophora). Pp. 142-208 in 
TOMPA, A. S., N. H. VERDONK & J. A. M. 
BIGGELAAR, eds.. Reproduction. In WILBUR, 
K. M., chief ed.. The Mollusca, vol. 7. Academic 
Press. 

GEV, 8., Y. ACHITUV & A. J. SUSSWEIN, 1984. 
Seasonal determinants of the life cycle in two 
species of Aplysia found in shallow waters along 
the Mediterranean coast of Israel. Journal of 
Experimental Marine Biology & Ecology. 74: 67- 
83. 

GLYNN, P. W., 1970. On the ecology of the Carnb- 
bean chitons Acanthopleura granúlala Gmelin 
and Chiton tuberculatus Linne: Density, mortality, 
reproduction, and growth. Smithsonian Contribu- 
tions to Zoology. 66: 1-21 . 

GODAN, D., 1983. Pest slugs and snails, biology 
and control. (English translation, from German). 
Spnnger-Verlag, Berlin, Heidelberg, NY. 

GOODWIN, B. J., 1978. The growth and breeding 
cycle of Littorina obtusata (Gastropoda: Proso- 
branchiata) from Cardigan Bay. Journal of Mol- 
luscan Studies, 44: 231-242. 

GRAHAME, J., 1977. Reproductive effort and r- 
and K- selection in two species of Lacuna (Gas- 



tropoda: Prosobranchia). Marine Biology. 40: 
217-22. 

GRAHAME, J. & G. M. BRANCH, 1985. Reproduc- 
tive patterns of marine invertebrates. Annual Re- 
view of Oceanography & Marine Biology. 23: 
373-398. 

GRAVE, B. H, 1 932. Embryology and life history of 
Chaetopleura apiculata. Journal of Morphology, 
54: 153-160. 

GUILLOU, J. & P. G. SAURIAU, 1985. Some ob- 
servations on the biology and the ecology of a 
Venus striatula population in the Bay of 
Douarnenez, Brittanny. Journal of the Manne Bi- 
ological Association of the United Kingdom, 65: 
889-900. 

HADFIELD, M. G., & B. S. MOUNTAIN, 1980. A 
field study of a vanishing species, Achatinella 
mustelina (Gastropoda, Puimonata), in the Wa- 
ianae Mountains of Oahu. Pacific Science, 34: 
345-358. 

HADFIELD, M. G. & M. SWITZER-DUNLAP, 1984. 
Opisthobranchs. Pp. 209-350 in: TOMPA, A. S., 
N. H. VERDONK & J. A. M. BIGGELAAR, eds., 
Reproduction. In: WILBUR, K. L., chief ed.. The 
Mollusca, vol. 7. Academic Press. N.Y. 

HANLON, R. T., 1983a. Octopus briareus. Pp. 251- 
266 in BOYLE, P. R., ed., Cephalopod life cycles, 
vol. 1. Academic Press, London. 

HANLON, R. T., 1983b. Octopus joubini. Pp. 293- 
310 in BOYLE, P. R., ed., Cephalopod life cycles, 
vol. 1. Academic Press, London. 

HARTWICK, В., 1983. Octopus dofleini. Pp. 277- 
291 in BOYLE, P. R., ed., Cephalopod life cycles, 
vol. 1. Academic Press, London. 

HAYASHI, I., 1980a. Reproductive biology of the 
ormer, Haliotis tuberculata. Journal of the Mahne 
Biological Association of the United Kingdom, 60: 
415-430. 

HAYASHI, I., 1980b. Structure and growth of a 
shore population of the ormer Haliotis tubercu- 
lata. Journal of the Mahne Biological Association 
of the United Kingdom 60: 431-437. 

HEARD, W. H.. 1975. Sexuality and other aspects 
of reproduction in Anodonta (Pelecypoda: Union- 
idae). Malacologia, 15: 81-103. 

HEARD, W. H., 1977. Reproduction of fingernail 
clams (Sphaeriidae: Sphaerium and Musculium). 
Malacologia. 16: 421-455. 

HEATWOLE, H. & A. HEATWOLE, 1978. Ecology 
of Puerto Rican camaenid treesnails. Malacolo- 
gia. 17: 241-315. 

HELLER, J., 1982. Natural history of Theba pisana 
in Israel. Journal of Zoology. London. 196: 475- 
487. 

HELLER, J. & M. VOLOKITA, 1981. Shell-bandind 
polymorphism of the land snail Xeropicta vestalis 
along the coastal plain of Israel. Biological Jour- 
nal of the Linnean Society. 16: 279-284. 

HENDELBERG, J., 1960. The fresh-water pearl 
mussel. Margaritifera margaritifera (L.). Institute 
of Freshwater Research, Drottningholm, 41 : 
149-171. 

HIXON, R. F., 1983. Loligo opalescens. Pp. 95- 



HELLER 



275 



1 14 in BOYLE, P. R., ed., Cephalopod life cycles. 
vol. 1. Academic Press, London. 

HOAGLAND, K. A., 1986. Genetic vanation in 
seven wood-boring teredinid and pholadid bi- 
valves With different patterns of life history and 
dispersal. Malacologia. 27: 323-339. 

HODASI, J. K., 1979. Life history studies of Acha- 
tina (Achatina) achatina (Linné). Journal of Mol- 
luscan Studies. 46: 106-115. 

HOLME, N. A., 1974. The biology of Loligo forbesi 
Steenstrup (Mollusca: Cephalopoda) in the Ply- 
mouth Area. Journal of tfie Manne Biological As- 
sociation of thie United Kingdom. 54: 481-503. 

HOUBRICK, R., 1 974. Growth studies on the genus 
Cerithium (Gastropoda: Prosobranchia) with 
notes on ecology and microhabitats. Nautilus. 88: 
14-27. 

HOUBRICK, R. S., 1 980. Observations on the anat- 
omy and life history of l^/lodulus modulus (Proso- 
branchia: Modulidae). Malacologia. 20: 1 17-142. 

HUGHES, R. N., 1986. A functional biology of ma- 
rine gastropods. Croom Helm, London. 

HUGHES, R. N. & D. J. ROBERTS. 1980a. Growth 
and reproductive rates of Littohna neritoides 
(L.) in North Wales. Journal of the Manne Biolog- 
ical Association of ttie United Kingdom, 60: 591- 
599. 

HUGHES, R. N., & D. E. ROBERTS, 1980b. Re- 
productive effort of winkles (Littorina spp.) with 
contrasted methods of reproduction. Oecologia. 
47: 130-136. 

HUNTER, P. J., 1968. Studies on slugs of arable 
ground. 2. Life cycles. Malacologia. 6: 379-389. 

JOLL, L. M., 1983. Octopus tetricus. Pp. 325-334 
in BOYLE, P. R., ed., Cephialopod life cycles, vol. 
1. Academic Press, London. 

JONES, D., 1983. Sclerochronology: reading the 
record of the molluscan shell. American Scientist. 
71: 384-391. 

JONES, D., I. THOMPSON & W. AMBROSE, 1978. 
Age and growth determination for the Atlantic 
surf clam Spisula solidissima (Bivalvia: Mactra- 
cea), based on internal growth lines in shell cross 
sections. Marine Biology, 47: 63-70. 

KANDEL, E. R., 1979. Befiavioral biology of Aply- 
sia. Freeman, San Francisco. 

KENNISH, M. J., 1980. Shell microgrowth analysis. 
Mercenaria mercenaria as a type example for re- 
search in population dynamics. In RHOADS, D. & 
R. LUTZ, eds., Skeletal growtti of aquatic organ- 
isms. Plenum Press, N.Y. & London. 

KENNY, R., 1977. Growth studies of the tropical 
intertidal limpet Acmaea antillarum. Marine Biol- 
ogy. 39: 161-170. 

KERNEY, M. P. & R. A. CAMERON, 1979. A field 
guide to tfie land snails of Britain and nortfi-west 
Europe. Collins, London. 

KIRKENDALL, L. R., & N. С STENSETH, 1985. On 
defining "breeding once". Tfie American Natural- 
ist. 125: 189-204. 

KOHN, R. R., 1971. Principles of mammalian ag- 
ing. Prentice-Hall, Englewood Cliffs. 

KONDRATENKOV, A. P., 1 978. On the generations 



of Hydrobia ulvae (Pennant) (Gastropoda) in the 
White Sea. Malacological Review. 1 1 : 112. 

KOSINSKA, M., 1980. The life cycle of Deroceras 
sturanyi (Simroth, 1894) (Pulmonata, 
Limacidae). Zoológica Polonica, 28: 113-155. 

KRASNOV, E. v., V. N. ZOLOTAREV, A. V. IG- 
NATIEV, L. A. POZDNIAKOVA, A. V. SILINA, 
V. E. SHEJGUS & С V. JAVNOV, 1975. Physical 
and chemical methods for studying growth of ma- 
nne mollusks. Malacological Review. 11: 153- 
154. 

KRISTENSEN, T. K., 1983. Gonatus fabricii. Pp. 
159-173 in BOYLE, P. R., ed., Cephialopod life 
cycles, vol. 1. Academic Press, London. 

LAAN, K. L., 1971. The population ecology of the 
terrestrial snail Helmintfioglypta arrosa (Pulmo- 
nata: Helicidae). Ph.D thesis. University of Cal- 
ifornia, Berkeley: 235 pp. 

LAZARIDOU-DIMITRIADOU, M., 1981. Contnbu- 
tion à l'étude biologique des escargots Cernuella 
virgata (DaCosta) et Xeropicta arenosa Ziegler 
(Gastropoda, Pulmonata) vivant sur les micro- 
dunes de Potidea, Chalkidiki (Grèce du Nord). V 
Convegno Nazionale délia Societa Malacologica 
Italiana. 73-83. 

LEE, S. Y., 1985. The population dynamics of the 
green mussel, Perm viridis (L.) in Victoria Har- 
bour, Hong Kong — dominance in a polluted en- 
vironment. Asian Marine Biology, 2: 107-118. 

LILLY, M. M., 1953. The mode of life and the struc- 
ture and functioning of the reproductive ducts of 
Byttiinia tentaculata. Proceedings of the Malaco- 
logical Society of London. 30: 87-1 10. 

LIVSHITS, S., 1985. Ecology of the terresthal snail 
Brephulopsis bidens (Pulmonata: Enidae): mor- 
tality, burrowing and migratory activity. Malaco- 
logia, 26: 212-223. 

LIVSHITS, G. & A. SHILEIKO, 1978. Life cycle of 
the mollusk Brephulopsis bidens. Soviet Journal 
of Ecology, 9: 460-465. 

LOPEZ- JAMAR, е., G. GONZALEZ & J. MEJUTO, 
1987. Ecology, growth and production of Thya- 
sira flexuosa (Bivalvia, Lucinacea) from Ria de 
la Cornua, northwest Spain. Ophelia, 27: 111- 
126. 

LUTZ, R. A. & M. CASTAGNA, 1980. Age compo- 
sition and growth rate of a mussel (Geukensia 
demissa) population in a Virginia salt marsh. 
Journal of Molluscan Studies, 46: 106-115. 

LUTZEN, J., 1979. Studies on the life history of 
Enteroxenos bonnevie. a gastropod endopara- 
sitic in aspidochirote holothuhans. Ophelia, 18: 
1-51. 

MACGINITIE, G. E. & N. MACGINITIE, 1968. 
Notes on Cryptochiton stellen (Middendorff, 
1848). Veliger, 11: 59-61. 

MACKIE, G. L., 1979. Growth dynamics in natural 
populations of Sphaeriidae clams (Sphaerium, 
Musculium, Pisidium). Canadian Journal of Zool- 
ogy, 57: 441-456. 

MACKIE, G. L., 1984. Bivalves. Pp. 351-418 in: 
TOMPA, A. S., N. H. VERDONK & J. A. M. 
BIGGELAAR, eds.. Reproduction. In: WILBUR, 



276 



LONGEVITY IN MOLLUSCS 



K. M., chief ed., The Mollusca, vol. 7. Academic 
Press. 

MACKIE, G. L. & D. G. MUGGINS, 1976. Biological 
notes on Eupera cubensis (Bivalvia: Sphaeri- 
idae) from Kansas. Journal of the Fisheries Re- 
search Board Canada. 33: 1652-1656. 

MAES, V. & D. RAEIHLE, 1975. Systematics and 
biology of Thala floridana (Gastropoda: Vexill- 
idae). Malacologia. 15: 43-67. 

MANGOLD, K., 1983a. Octopus vulgaris. Pp. 335- 
364 in BOYLE, P. R., ed., Cephalopod life cycles. 
vol. 1. Academic Press, London. 

MANGOLD, K., 1983b. Eledone moschata. Pp. 
387-400 in BOYLE, P. R., ed., Cephalopod life 
cycles, vol. 1 . Academic Press, London. 

MASLIN, J. L. & Y. BOUVET, 1986. Population dy- 
namics of Corbula trígona, in Lake Ahe'me, a 
West African lagoon in Benin. Oikos, 46: 292- 
302. 

MATTESON, M. R., 1948. Life history of Elliptio 
complánala (Dilwyn 1817). American Midland 
Naturalist. 40: 690-723. 

MCGRATH, D. & D. O'FOIGHIL, 1986. Population 
dynamics and reproduction of hermaphroditic La- 
saea rubra (Montagu) (Bivalvia, Galeommata- 
cea). Ophelia. 25: 209-219. 

MEAD, A. R., 1 961 . The giant African snail: a prob- 
lem in economic malacology. Chicago University 
Press, Chicago. 

MEIER-BROOK, C, 1970. Investigations in the bi- 
ology of some Pisidium species (Mollusca; Eu- 
lamellibranchiata; Sphaeriidae). Archiv fur Hy- 
drobiologie. Supplement. 38: 73-150. 

MEYER, K. O., 1955. Naturgeschichte der Strand- 
schnecke Ovatella myosotis (Draparnaud). Ar- 
chiv für Molluskenkunde. 84: 1-43. 

MILLER, M. C, 1962. Annual cycles of some Manx 
nudibranchs, with a discussion of the problem of 
migration. Journal of Animal Ecology, 31 : 545- 
569. 

MORDAN, P. В., 1978. The life cycle oi Aegopinella 
nitidula (Draparnaud) (Pulmonata: Zonitidae) at 
Monks Wood. Journal of Conchology. 29: 247- 
252. 

MORTON, B. S., 1969. Studies on the biology of 
Dreissena polymorpha Pall. 3. Population dy- 
namics. Proceedings of the Malacological Soci- 
ety of London. 38: 471-482. 

MORTON, B. S., 1977. Population dynamics of 
Limnoperna fortunei (Dunker 1857) (Bivalvia: 
Mytilacea) in Plover Cove Reservoir, Hong Kong. 
Malacologia. 16: 165-182. 

MORTON, B. S., 1981. The biology and functional 
morphology of Mytilopsis sallei (Recluz) (Bi- 
valvia: Dreissenacea) fouling Visakhpatnam Har- 
bour, Andhra Pradesh, India. Journal of Mollus- 
can Studies. 47: 25-42. 

MORTON, B. S., 1986. The population dynamics 
and life history tactics of Pisidium clarkaneum 
and P. annandalie (Bivalvia: Pisidiidae), sympat- 
ric in Hong Kong. Journal of Zoology. 210: 427- 
449. 

MORTON, J. E., 1954. Notes on the ecology and 



annual cycle of Carychium tridentatum at Box 
Hill. Proceedings of the Malacological Society of 
London. 31 : 30-46. 

MOYNIHAN, M. & A. F. RODANICHE, 1982. The 
behaviour and natural history of the Caribbean 
Reef Squid Sepioteuthis sepioidea. Paul Parey, 
Berlin & Hamburg. 

NAIR, B. & M. SARASWATHY, 1971. The biology 
of wood-boring teredinid molluscs. Advances in 
Marine Biology. 9: 336-510. 

NEGUUS, С L., 1966. A quantitative study of 
growth and production of unionid mussels in the 
River Thames at Reading. Journal of Animal 
Ecology. 35: 513-532. 

NESIS, K. N., 1983. Dosidicus gigas. Pp. 217-231 
in BOYLE, P. R., ed., Cephalopod life cycles, vol. 
1. Academic Press, London. 

NIXON, M., 1983. Teuthowenia megalops. Pp. 
233-247 in BOYLE, P. R., ed., Cephalopod life 
cycles, vol. 1 . Academic Press, London. 

NYBAKKEN, J., 1974. Phenology of the smaller 
dendrononotatacean, arminacean and aeoli- 
dacean nudibranchs at Asiloma State Beach 
over a twenty-seven month period. Veliger, 16: 
370-372. 

NYBAKKEN, J., 1978. Abundance, diversity and 
temporal variability in a California intertidal nudi- 
branch assemblage. Manne Biology. 45: 129- 
146. 

OCKELMANN, K. W. & K. MUUS, 1978. The biol- 
ogy, ecology and behaviour of the bivalve My- 
sella bidentata (Montagu). Ophelia. 17: 1-93. 

ODOR, R. K., 1983. ///ex illecebrosus. Pp. 175- 
199 in BOYLE, P. R., ed., Cephalopod life cycles, 
vol. 1. Academic Press, London. 

ODOR, R. K. & E. G. MACALASTER, 1983. Bat- 
hypolypus arcticus. Pp. 401-410 in BOYLE, 
P. R., ed., Cephalopod life cycles, vol. 1. Aca- 
demic Press, London. 

OKUTANI, T., 1983. Todarodes pacificus. Pp. 201- 
214 in BOYLE, P. R., ed., Cephalopod life cycles, 
vol. 1. Academic Press, London. 

PARIVAR, K., 1978. A histological survey of go- 
nadal development in Arion ater L. (Mollusca, 
Pulmonata). Journal of Molluscan Studies, 44: 
250-264. 

PAULY, D. & H. CALUMPONG, 1984. Growth, re- 
production and mortality of the sea hare Dola- 
bella auriculaha (Gastropoda: Aplysiidae) in the 
central Vishayes, Philippines. Mahne Biology, 
79: 289-293. 

PETERSEN, G. H., 1978. Life cycles and popula- 
tion dynamics of marine benthic bivalves from 
Disko Bugt area of west Greenland. Ophelia, 17: 
95-120. 

PERRON, F. E., 1982. Inter- and intraspecific pat- 
terns of reproductive effort in four species of 
cone shells (Conus spp.). Mahne Biology, 68: 
161-167. 

PHILLIPS, D. W., 1981. Life history features of the 
marine intertidal limpet Notacmea scutum (Gas- 
tropoda) in central California. Marine Biology, 64: 
95-103. 



HELLER 



277 



PHILLIPS, В. F. & N. А. CAMPBELL, 1974. Mortal- 
ity and longevity m the whelk Dicathias órbita 
(Gmelin). Australian Journal of Manne & Fresh- 
water Research. 25: 25-33. 

PICKEN, G. В., 1980. The distribution, growth and 
reproduction of the Antarctic limpet Nacella (Pa- 
tinegra) concinna. Journal of Experimental !\/la- 
rine Biology & Ecology. 42: 71-85. 

PIECHOCKI, A., 1982. Life cycle and breeding bi- 
ology of Vestía elata (Rossm.) (Gastropoda, 
Clausiliidae). Malacologia. 22: 219-223. 

PINEL-ALLOUL, B. & E. MAGNIN, 1973. Observa- 
tions sur le cycle vital et la croissance d' Amnícola 
limosa (Say) (Mollusca, Gastropoda, Prosobran- 
chia) du Lac Saint Luis près de Montréal. Cana- 
dian Journal of Zoology. 51 : 311-313. 

PILSBRY, H., 1939. Land Mollusca of North Amer- 
ica. Vol. 1 . Monographs of the Academy of Nat- 
ural Sciences of Philadelphia, 3: 573 + ix pp. 

PLUMMER, J., 1982. Reproductive potential and 
longevity in Archachatina marginata ovum. Jour- 
nal of Molluscan Studies. 48: 366-367. 

POKRYSZKO, B. M., 1986. Life history of Verligo 
pusilla Müller (Gastropoda: Pulmonata: Vertig- 
inidae). Proceedings of the Ninth International 
l\Aalacological Congress. Edinburgh (abstracts): 
65. 

POMEROY, D. E., 1 969. Some aspects of the ecol- 
ogy of the land snail Helicella virgata. in South 
Australia. Australian Journal of Zoology. 17: 
495-514. 

PORTER, W., 1967. Solar radiation through the liv- 
ing body walls of vertebrates with emphasis on 
desert reptiles. Ecologicalf\Aonographs. 2,1: 273- 
296. 

POWELL, E. N. & H. CUMMINS, 1985. Are mollus- 
can maximum life spans determined by long-term 
cycles in benthic communities? Oecologia. 67: 
177-182. 

POWELL, E. N. & R. J. STANTON Jr., 1985. Esti- 
mating biomass and energy flow of molluscs in 
paleo-communities. Paleontology (London), 28: 
1-34. 

RA'ANAN, M., 1986. The systematics and ecolog- 
ical evolution of the genus l\/lelanopsis in Lake 
Kinneret. Ph.D. thesis, Hebrew University, Jerus- 
alem (Hebrew, English summary). 

RABOUD, C, 1986. Age determination of Arianta 
arbustorum (L.) (Pulmonata) based on growth 
breaks and inner layers. Journal of Molluscan 
Studies. 52: 243-247. 

RALPH, R. & J. G. MAXWELL, 1 977. Growth of two 
Antarctic lamellibranchs, Adamussium colbecki 
and Laternula elliptica. Marine Biology. 42: 171- 
175. 

RANDOLPH, P. A., 1973. Influence of environmen- 
tal vahability on land snail population properties. 
Ecology, 54: 933-955. 

RANKIN, J. J., 1979. A freshwater shell-less mol- 
lusc from the Caribbean: structure, biotics, and 
contribution to a new understanding of the Aco- 
chlidiodea. Life Sciences Contributions, Royal 
Ontario Museum, 116: 1-123. 



RHOADS, D. C, E. С LUTZ, E. С. REVELAS & 
R. M. CERRATO, 1981. Growth of bivalves at 
deep sea hydrothermal vents along the Galapa- 
gos Rift. Science. 214: 911-913. 

RIBI, G. & M. GEBHARDT, 1986. Age specific fe- 
cundity and size of offspring in the prosobranch 
snail, Viviparus ater. Oecologia, 71: 18-24. 

RICHARDOT-COULET, M. & L. ALFARO- 
TEJERA, 1985. The life cycle and ecology of the 
freshwater planorbidae Armiger chstata. Journal 
of Molluscan Studies, 51 : 35-51 . 

RUDMAN, W. В., 1972. A study of the anatomy of 
Pupa and Maxacteon (Acteonidae, Opisthobran- 
chia) with an account of the breeding cycle of 
Pupa kirki. Journal of Natural History, 6: 603- 
619. 

RUNHAM, N. W. & P. J. HUNTER, 1970. Terrestrial 
slugs. Hutchinson, London. 

RUNHAM, N. W. & A. A. LARYEA, 1968. Studies 
on the maturation of the reproductive system of 
Agholimax reticulatus (Pulmonata: Limacidae). 
Malacologia, 7: 93-108. 

RUSSELL-HUNTER, D.S., 1978. Ecology of fresh- 
water pulmonales. In: FRETTER, V. & J. PEAKE, 
eds., Pulmonates, vol. 2. Systematics, evolution 
& ecology. Academic Press, London & NY. 

SASTRY, A. N., 1979. Pelecypoda (excluding Os- 
treidae). Pp. 113-129 in GIESE, A. & J. 
PEARSE, eds.. Reproduction of marine inverte- 
brates, vol. 5. Academic Press, NY & London. 

SAUNDERS, W. В., 1984. The role and status of 
Nautilus in its natural habitat: evidence from 
deep-water remote camera photosequences. 
Palaeobiology, 10: 469-486. 

SELLMER, G., 1967. Functional morphology and 
ecological life history of the gem clam. Gemma 
gemma (Eulamellibranchia: Veneridae). Malaco- 
logia, 5: 137-223. 

SHEPHERD, S. A., G. P. KIRKWOOD & R. L. 
SANDLAND, 1982. Studies on southern Austra- 
lian abalone (genus Haliotis). 3. Mortality of two 
exploited species. Australian Journal of Marine & 
Freshwater Research, 33: 265-272, 

SHRADER, A. L,, 1974, Life parameters of an ex- 
perimental population of Omalonyx felina (Pul- 
monata: Succineidae), Malacological Review, 7: 
48. 

SINGLEY, С Y,, 1983, Euprymna scolopes. Pp. 
69-74 in BOYLE, P, R,, ed,, Cephalopod life cy- 
cles, vol, 1, Academic Press, London. 

SIRE, J, Y, & G, BONNET, 1984, Growth and struc- 
ture of the calcified operculum of the Polynesian 
gastropod Turbo setosus (Prosobranchia: Tur- 
binidae): determination of individual age. Marine 
Biology, 68: 161-167, 

SMITH, B, D., 1987. Growth rate, distribution and 
abundance of the introduced topshell Trochus ni- 
loticus Linnaeus on Guam, fvlarianna Islands, 
Bulletin of Marine Sciences, 41 : 466-474, 

SMITH, D, G,, 1976, Notes on the biology of Mar- 
garitifera margaritifera (Linn.) in central Massa- 
chusetts. American Midland Naturalist, 96: 252- 
256. 



278 



LONGEVITY IN MOLLUSCS 



SMOLENSKA, J., 1936. Observations biologiques 
sur la Bielizia coerulans. Annales Musei Zoo- 
logia Polonia. 1 1 ; 241-261 . 

SOKAL, R. R, & F, J. ROHLF, 1981. Biometry, 2nd 
ed. Freeman, San Francisco. 

SOLEM, Д., 1978. Classification of land fVlollusca. 
Pp. 49-97, in: FRETTER, V. & J. PEAKE, eds., 
The Pulmonata, vol. 2A: Systematics. evolution 
and ecology. Academic Press, London & NY. 

SOLEfVl, A. & С С. CHRISTENSEN. 1984. Ca- 
maenid land snail reproductive cycle and growth 
patterns in semiarid areas of north-western Aus- 
tralia. Australian Journal of Zoology, 32: 471- 
491. 

SOUTHGATE, T., 1982. The biology of Barlóela 
unifasciata (Gastropoda: Prosobranchia) in red 
algal turfs in SW Ireland. Journal of the Marine 
Biological Association of the United Kingdom, 62: 
461-468. 

SPOEL, S., 1958. Groei en ouderdom bij Viviparus 
contectus (Millet 1813) en Viviparus viviparus 
(Linne 1758). Basteria, 22: 77-90. 

STAIKOU, A., & M. LAZARIDOU-DIMITRIADOU, 
1986. Contributions to the biology and ecology of 
the edible snail Helix lucorum L. Ninth Interna- 
tional Malacological Congress, Edinburgh. 

STANCZYKOWSKA, A., A. MAGNIN & A. DU- 
MOUCHEL, 1971. Etude de trois populations de 
Viviparus malleatus (Reeve) (Gastropoda, 
Prosobranchia) de la région de Montréal: Crois- 
sance, fécondité, biomasse et production an- 
nuelle. Canadian Journal of Zoology. 49: 1431- 
1441. 

STRANDINE, E. G., 1941. Quantitative study of a 
snail population. Ecology, 22: 86-91. 

SUMMERS, W. С, 1983. Loligo pealei. Pp. 115- 
142 in BOYLE, P. R., ed., Cephalopod life cycles, 
vol. 1. Academic Press, London. 

SUTHERLAND, J. P., 1970. Dynamics of high and 
low populations of the limpet Acmaea scabra 
(Gould). Ecological Monographs, 40: 169-188. 

TAYLOR, D. W. & N. F. SOHL, 1962. An outline of 
gastropod classification. Malacologia, 1 : 7-32. 

TAYLOR, J. W., 1907. Monograph of the land and 
freshwater Mollusca of the British Isles, Vol. 2: 
Testacellidae, Limacidae, Arionidae. Taylor, 
Leeds. 

THOMPSON, T., 1976. Opisthobranchs. Ray Soci- 
ety, London. 

TODD, С D., 1981. The ecology of nudibranch 
molluscs. Oceanography and Marine Biology, 19: 
141-234. 

TONG, L. K. Y., 1986. The population dynamics 
and growth of Thais clavigera and Morula musiva 
(Gastropoda: Muricidae) in Hong Kong. Asian 
Marine Biology, 3: 145-162. 

TUREKIAN, K. K., J. K. COCHRAN, D. P. 
KHARKAR, R. M. CERRATO, J. R. VAISNYS, 
H. L. SANDERS, J. F. GRASSIE & J. A. ALLEN, 
1975. Slow growth rate of a sea-clam determined 
by 228Ra chronology. Proceedings of the Na- 
tional Academy of Science of the USA, 72, 7: 
2829-2832. 



TURNER, R. D., 1973. Wood-boring bivalves, op- 
portunistic species in the deep sea. Science. 
180: 1377-1379. 

TUSKES, P. M., 1981. Population structure and bi- 
ology of Liguus tree snails on Lignumvitae Key, 
Flonda. Nautilus. 95: 162-169. 

UMINSKI, T., 1975. Life cycles in some Vitnnidae 
(Mollusca, Gastropoda) from Poland. Annales 
Zoología. 33: 17-31. 

UMINSKI, T., 1979. Life history in Eucobresia niva- 
lis (Dumony et Mortillet) with notes on two other 
Vitrinidae (Mollusca, Gastropoda). Bulletin de Í- 
Academie Polonaise des Sciences. 28: 205- 
210. 

UMINSKI, T., & и. FOCHT, 1979. Population dy- 
namics of some land gastropods in a forest hab- 
itat in Poland. Malacologia. 18: 181-184. 

UNDERWOOD, A. J. & K. E. MCFADYEN, 1983. 
Ecology of the intertidal snail Littonna acutispira 
Smith. Journal of Experimental Manne Biology & 
Ecology. 66: 169-197. 

UVALIEVA, K. K., 1978. Life cycles and damage 
caused by the slugs Deroceras caucasicum 
(Simroth) and Parmacella rutellum Hutton in the 
Alma-Ata region (Kasakh SSR). Malacological 
Review. 1 1 : 82. 

VAHL, O., 1971. Growth and density of Patina pel- 
lucida on Laminaha hyperborea (Gunnerus) from 
Western Norway. Ophelia. 9: 31-50. 

VAN CLEAVE, H. & D. ALTRINGER, 1937. Studies 
on the life cycle of Campeloma rufum. a fresh- 
water snail. American Naturalist. 71: 167-184. 

VAN CLEAVE, H. & L. G. LEDERER, 1932. Studies 
on the life cycle of the snail Viviparus contec- 
toides. Journal of Morphology. 53: 499-522. 

VAN CLEAVE, H. & T FOSTER, 1937. The sea- 
sonal life history of a land snail, Polygyra thyroi- 
dus (Say). Nautilus. 51: 50-54. 

VAN HEUKELEM, W., 1983a. Octopus cyanea. Pp. 
267-276 in BOYLE, P. R., ed., Cephalopod life 
cycles, vol. 1. Academic Press, London. 

VAN HEUKELEM, W., 1983b. Octopus maya. Pp. 
311-323 in BOYLE, P. R., ed., Cephalopod life 
cycles, vol. 1. Academic Press, London. 

VENTILLA, R. F., 1982. The scallop industry in 
Japan. Advances in Marine Biology. 20: 310- 
382. 

VERMEIJ, G. J., 1978. Biogeography and adapta- 
tion: Patterns in mahne life. Harvard University 
Press, Cambndge, Mass. 

VOKES, H. E., 1967. Genera of the Bivalvia: a sys- 
tematic and bibliographic catalogue. Bulletins of 
American Paleontology. 51 (232): 111-394. 

VOSS, G. L., 1977. Present status and new trends 
in cephalopod systematics. In: NIXON, M. & J. B. 
MESSENGER, eds.. The biology of cephalo- 
pods. Symposia of the Zoological Society of Lon- 
don. 38: 49-60. 

VOSS, R. S., 1976. Observations on the ecology of 
the Florida tree snail Liguus fasciatus (Müller). 
Nautilus. 90: 65-69. 

WEFER, G. & J. S. KILLINGLEY, 1982. Growth 
histories of strombid snails from Bermuda re- 



HELLER 



279 



corded in their O-^^ and C-^^ profiles. Marine Bi- 
ology. 60: 129-135. 

WELLS, F. е., 1976. Growth rate of four species of 
euthecosomatous pteropods occurring off Barba- 
dos, West Indies. Nautilus. 90: 114-116. 

WELLS, M. J. & J. WELLS, 1977. Cephalopoda: 
Octopoda. Pp. 291-336 in GIESE, A. & J, 
PEARSE, eds., Reproduction of manne inverte- 
brates, vol. 4. Academic Press, N.Y. 

WILLIAMS, M. J. & M. DREDGE, 1981. Growth of 
the saucer scallop, Amusium japonicum balloti 
Habe in central eastern Queensland. Australian 
Journal of Marine & Freshwater Researcti. 32: 
657-666. 

WILLIAfVlSON, P. & M. A. KENDALL, 1981. Popu- 
lation structure and growth of the trochid Mon- 
odonta lineata determined by shell rings. Journal 
of ttie Manne Biological Association of thie United 
Kingdom. 61: 1011-1026. 

WINTERBOURN, M. J., 1970. Population studies 
on the New Zealand freshwater gastropod Pota- 
mopyrgus antipodarum (Gray). Proceedings of 
tfie Malacological Society of London. 39: 139- 
152. 

WOODRUFF, D., 1978. Evolution and adaptive ra- 
diation of Cerion: a remarkably diverse group of 



West Indian land snails. Malacologia, 17: 223- 
239. 

WORMS, J., 1983. Loligo vulgaris. Pp. 143-157 in 
BOYLE, P. R., ed., Cephalopod life cycles, vol. 1. 
Academic Press, London. 

WU, R. S., 1985. Some aspects of the ecology of 
Pinna atropurpúrea Sowerby (Bivalvia: Pin- 
nidae). With a note on the commensal pontoniine 
shrimp Anchistus custos (Forskal). In: MORTON, 
B. & D. DUDGEON, eds., The malacofauna of 
Hong Kong and Southern China. Hong Kong Uni- 
versity Press, Hong Kong. 

YOKLEY, P., 1972. Life history of Pleurobema cor- 
datum (Rafinesque 1820) (Bivalvia: Unionacea). 
Malacologia. 11: 351-364. 

YOM-TOV, Y., 1971. The biology of two desert 
snails, Trochoidea (Xerocrassa) seetzeni and 
Sphincterochila boissieri. Israel Journal of Zool- 
ogy. 20: 231-248. 

ZAIKA, V. E., 1973. Specific production of aquatic 
invertebrates. Halstead Press, N.Y. 

ZOLOTAREV, V. N., 1980. The life span of bivalves 
from the sea of Japan and Sea of Okhotsk. So- 
viet Journal of Marine Biology. 6: 301-308. 

Revised Ms. accepted 19 June 1989 



APPENDIX 



TABLE 1. Life spans in marine snails 



Species 



Lifespan 



Authority 



PROSOBRANCHIA. 

Haliotidae 
Haliotis cracherodii 
Haliotis iris 
Haliotis laevigata 
Haliotis ruber 
Haliotis rufescens 
Haliotis tuberculata 

Fissurellidae 
Fissurella barbadensis 
Fissurella crassa 
Montfortula rugosa 

Patellidae 
Cellana grata 
Ce I Ian a radiât a 
Cellana tramoserica 
Nacella concinna 
Nacella delesserti 
Patella aspersa 
Patella cochlear 
Patella granatina 
Patella granulans 
Patella longicosta 
Patella оси lus 
Patella vu Ig ata 
Patina pe I lucida 



51 Powell & Cummins, 1985 

10 Powell & Stanton, 1985 

10 Shepherd et al., 1982 

10 Shepherd et al., 1982 

13 Shepherd et al., 1982 

6 Hayashi, 1980a, b 

3 Hughes & Roberts, 1980a, b 

10 Bretos, 1980 

3 Powell & Cummins, 1985 

15 Comfort, 1957 

4 Powell & Cummins, 1985 

5 Fletcher, 1984 
21 Picken, 1980 

10 Blankley & Branch, 1985 

12 Powell & Stanton, 1985 

25 Powell & Cummins, 1985 

6 Powell & Cummins, 1985 
8 Powell & Cummins, 1985 

16 Grahame & Branch, 1985 
3 Powell & Cummins, 1985 

15 Comfort, 1957 

SL Vahl, 1971 



(continued) 



280 



LONGEVITY IN MOLLUSCS 



TABLE 1 . (continued) 



Species 



Lifespan 



Authority 



Acmaeidae 
Acmaea antillarum 
Acmaea digitalis 
Acmaea dorsuosa 
Acmaea insessa 
Acmaea paradigitalis 
Acmaea pelta 
Acmaea scabra 
Acmaea testudinalis 
Notacmea petterdi 
Notacmea scutum 
Patelloida alticostata 
Patelloida latistrigata 

Trochidae 
Falsimargarita iris 
Gibbula umbilicalis 
Margantes helicinus 
Monodonta lineata 
Tegula funebralis 
Trochus niloticus 

Turbidae 

Turbo setosus 

Neritidae 
Nerita albicilla 
Nerita atramentosa 
Nerita fulgurans 
Nerita japónica 
Nerita polita 
Nerita tesselata 
Nerita versicolor 

Littorinidae 
Littorina acutispira 
Littohna coccínea 
Littorina littorea 
Littorina neglecta 
Littorina neritoides 
Littorina nigrolineata 
Littorina obtusata 
Littorina rudis 
Littorina scabra 
Littorina sitkana 

Lacunidae 
Lacuna pallidula 
Lacuna vincta 

Skeneopsidae 
Skeneopsis planorbis 

Omalgyridae 
Omalgyra atomus 

Rissoellidae 
Rissoella diaphana 
Rissoella opalina 

Rissoidae 
Barleeia unifasciata 
Rissoa parva 
Rissoa splendida 



SL 


Kenny. 1977 


5 


Choat & Black. 1979 


15 


Comfort, 1957 


SL 


Choat & Black, 1979 


20 


Powell & Cummins, 1985 


4 


Powell & Cummins, 1985 


12 


Sutherland, 1970 


7 


Zaika, 1973 


10 


Powell & Cummins, 1985 


SL 


Phillips, 1981 


6 


Powell & Cummins, 1985 


4 


Powell & Cummins, 1985 


3 


Egorova, 1978 


5 


Williamson & Kendall, 1981 


SL 


Zaika, 1973 


15 


Williamson & Kendall, 1981 


30 


Williamson & Kendall, 1981 


15 


Smith, 1987 



3 Sire & Bonnet, 1984 

12 Frank, 1969 

6 Powell & Cummins, 1985 

6 Powell & Cummins, 1985 

3 Comfort, 1957 

4 Powell & Cummins, 1985 

6 Hughes & Roberts. 1980a, b 

8 Hughes & Roberts, 1980a 

SL Undenwood & McFadyen, 1983 

4 Comfort, 1957 

9 Hughes & Roberts, 1980b 
SL Hughes, 1986 

15 Hughes & Roberts, 1980b 

9 Hughes & Roberts, 1980b 

3 Goodwin, 1978 

11 Hughes & Roberts, 1980b 

5 Comfort, 1957 

SL Powell & Cummins, 1985 

SL Grahame, 1977 

SL Grahame, 1977 

SL Fretter, 1948 

SL Fretter, 1948 

SL Fretter, 1948 

SL Fretter, 1948 

SL Southgate, 1982 

SL Powell & Stanton, 1985 

SL Zaika, 1973 



HELLER 



281 



TABLE 1 . (continued) 



Species 



Lifespan 



Authority 



Entoconchidae 
Enteroxenos bonnevie 
Thyonicola amencana 

Modulidae 
Modulus modulus 

Cerithiidae 
Centhium coeruleum 
Centhium eburneum 
Centhium lutosum 
Centhium muscarum 
Centhium rupestre 
Centhium scabndum 

Diastomidae 
Diastoma varium 

Potamididae 
Batillana attramentaha 
Centhidea décollât a 

Calyptraeidae 
Calyptraea chinensis 

Trichotropidae 

Tnchotropis cancellatum 

Strombidae 
Strombus costatus 
Strombus gigas 
Strombus luhuanus 

Naticidae 
Conuber sórdida 
Polinices duplicatus 

Thaididae 
Dicathias órbita 
Morula musiva 
Nucella lamellosa 
Ocenebra poulsoni 
Shasl<yus festivus 
Thais clavigera 
Thais lapillus 
Urosalpinx cinerea 

Buccinidae 
Neptúnea antigua 

Nassahidae 
Bullia rhodostoma 
Nassahus obsoleta 
Nassarius reticulatus 

Mitridae 
Thala floridana 

Fasciolaridae 
Latirolagena smaragdula 

Vasidae 

Vasum turbinellus 

Conidae 
Conus arenatus 
Conus pennaceus 



SL 


Lutzen, 1979 


SL 


Bryne, 1985 


SL 


Houbrick, 1980 


10 


Ayal, 1978 


SL 


Houbrick, 1974 


SL 


Houbrick, 1974 


SL 


Houbrick, 1974 


5 


Ayal, 1978 


SL 


Ayal, 1978 



SL Powell & Stanton, 1985 

7 Powell & Cummins, 1985 

9 Powell & Stanton, 1985 

5 Comfort, 1964 

3 Comfort, 1964 

5 Wefer& Killingley, 1982 
7 Wefer& Killingley, 1982 

6 Frank, 1969 

5 Powell & Cummins, 1985 

4 Edwards & Huebner, 1977 

5 Phillips & Campbell, 1974 
9 Tong, 1986 

6 Hughes, 1986 

15 Fothenngham, 1971; Phillips & Campbell, 1974 

20 Fothenngham, 1971, Phillips & Campbell, 1974 

7 Tong, 1986 

10 Hughes & Roberts, 1980a,b 

4 Powell & Stanton, 1985 

10 Powell & Stanton, 1985 



20 


Brown, 1982 


3 


Comfort, 1957 


15 


Powell & Stanton, 1985 


6 


Maes & Raeihie, 1975 


12 


Frank, 1969 


10 


Frank, 1969 


19 


Powell & Cummins, 1985 


10 


Perron, 1982 



{continued) 



282 



LONGEVITY IN MOLLUSCS 



TABLE 1 . (continued) 



Species 



Lifespan Authority 



Terebridae 
Terebra gouldi 

OPISTHOBRANCHIA 
Acteonidae 
Pupa kirki 

Retusidae 
Retusa obtusa 

Limacinidae 
Limacina bulimoides 
Limacina inflata 
Limacina trochiformis 

Cavoliniidae 
Cavolinia gibbosa 
Clio pyramidata 
Creseis virgula 
С и vi eh na columella 
Diacria trispinosa 

Aplysiidae 
Aplysia californica 
Aplysia depilans 
Aplysia fasciata 
Aplysia juliana 
Aplysia kurodai 
Aplysia punctata 
Dolabella auricularia 
Phiyllaplysia taylori 

Limapontiidae 
Limapontia capitata 
Limapontia depressa 
Limapontia senestra 

Elysiidae 
Elysia viridis 

Tritoniidae 

Tritonia hombergi 

Dendronotidae 
Dendronotus frondosus 
Dendronotus subramosus 

Hancockiidae 
Hancockia californica 

Dotoidae 
Doto amyra 
Doto corónala 
Doto fragilis 
Doto kya 

Tethyidae 
Melibe leonia 

Goniodorididae 
Ancula cnstata 
Goniodoris nodosa 

Onchidorididae 
Adalana próxima 
Acanthodoris pilosa 
Onchidons bilamellata 
Onchidons depressa 



10 Powell & Cummins, 1985 



SL 


Rudman, 1972 


SL 


Thompson. 1976 


SL 


Wells, 1976 


SL 


Wells, 1976 


SL 


Wells, 1976 


SL 


Wells, 1976 


SL 


Wells, 1976 


SL 


Wells, 1976 


SL 


Wells, 1976 


SL 


Wells, 1976 


SL 


Gev et al., 1984 


SL 


Gev et al., 1984 


SL 


Gev et al., 1984 


SL 


Gev et al., 1984 


SL 


Gev et al., 1984 


SL 


Comfort, 1957 


SL 


Pauly & Calumpong, 1984 


SL 


Pauly & Calumpong, 1984 


SL 


Miller, 1962 


SL 


Comfort, 1957 


SL 


Miller, 1962 


SL 


Miller, 1962 


SL 


Miller, 1962 


SL 


Miller, 1962; Nybakken 1974 


SL 


Nybakken, 1974 


SL 


Nybakken, 1974 


SL 


Nybakken, 1974 


SL 


Miller, 1962 


SL 


Miller, 1962 


SL 


Nybakken, 1974 


SL 


Comfort, 1957 


SL 


Todd, 1981 


SL 


Miller, 1962; Thompson, 1976 


SL 


Thompson, 1976 


SL 


Miller, 1962; Todd, 1981 


SL 


Thompson, 1976 


SL 


Todd, 1981 



HELLER 



283 



TABLE 1 . (continued) 



Species 



Lifespan 



Authority 



Onchidohs muricata 
Onchidohs pusilla 

Triophidae 

Tnopha maculata 

Polyceridae 
Limacia clavigera 
Polycera quadhlineata 

Chromodorididae 
Chromodohs nodosus 
Chromodons zebra 

Archidorididae 
Archidoris pseudoargus 

Kentrodorididae 
Jorunna tomentosa 

Heroidae 
Hero formosa 

Coryphellidae 
Coryphella lineata 
Coryphella trilineata 

Facelinidae 
Facelina coronata 

Aeolidiidae 
Aeolidia papulosa 

Eubranchidae 
Eubranchus exiguus 
Eubranchus oliváceas 
Eubranchus pallidus 
Eubranchus rustyus 

Cuthonidae 
Catronia alpha 
Tergipes despectus 
Thnchesia abronia 
Trinchesia albocrusta 
Thnchesia amoena 
Thnchesia flavovulta 
Thnchesia fol lata 
Thnchesia lagunae 

PULMONATA 

Siphonariidae 
Siphonaha denticulata 
Siphonaria lessoni 
Siphonaha vi rg и lata 



SL 


Miller, 1962; Thompson, 


SL 


Miller, 1962; Todd, 1981 


SL 


Nybakken, 1978 


SL 


Miller, 1962 


SL 


Miller, 1962 


SL 


Comfort, 1957 


SL 


Comfort, 1957 


SL 


Thompson, 1976 


SL 


Miller, 1962 


SL 


Miller, 1962 


SL 


Miller, 1962 


SL 


Nybakken, 1974 


SL 


Todd, 1981 


SL 


Miller, 1962 


SL 


Miller, 1962 


SL 


Nybakken, 1974 


SL 


Miller, 1962; Todd, 1981 


SL 


Nybakken, 1974 


SL 


Nybakken, 1974 


SL 


Miller, 1962 


SL 


Nybakken, 1974 


SL 


Nybakken, 1974 


SL 


Miller, 1962 


SL 


Nybakken, 1974 


SL 


Todd, 1981 


SL 


Nybakken, 1974 



6 Powell & Cummins, 1985 

5 Powell & Cummins, 1985 

3 Powell & Cummins, 1985 



Notes to Table 1 : 

Acmaea Insessa lives on the kelp Egregia laevigata and must mature and reproduce within a year, before death of the alga 

(Choat & Black 1979). 

Margantes tielicinus is a small topshell of Arctic oceans. 

The rissoacean genera Skeneopsis. Omalogyra and Rissoella are minute (about 2 mm) herbivorous gastropods of rocky 

tide pools. They are hermaphrodites, and Omalogyra may practice self-fertilization (Fretter, 1948). Barleeia dwells amongst 

filamentous red algae, where it grazes upon diatoms. 

Enteroxenos is a genus of greatly modified, shelless prosobranchs that live as endoparasites in aspidochirote holothunans. 

The population breeds throughout the year, but each female produces only one egg batch, after which she dies (Lutzen, 

1979). 

Thyonicola americana is an endoparasite of holothurians. Evisceration in these holothunans is a seasonal (autumn) event 

which sheds the parasites, that then die. Minimal life span of the parasite is 6 months (Bryne, 1985). 



284 



LONGEVITY IN MOLLUSCS 



M. modulus lives upon angiosperm sea-grasses (Houbrick, 1980), 

Retusa obtusa feeds upon the marsh-dwelling prosobranch Hydrobia ulvae (Thompson, 1976). 

Limacinids and cavolimids are are euthecosomatous pteropods. a small group of planktomc gastropods occurring mainly in 

tropical oceans (Wells, 1976). 

Onchidoris bilamellata feeds upon barnacles (Thompson, 1976) 

Comfon (1957) m stating that the opisthobranch Haminea hydatis lives four years, quotes Berrill (1931) I could not find 

evidence for this m Berrill's paper. 

Comfort (1957) mentions Philine aperta as living 3-4 years, an exceptionally long life-span for an opisthobranch. As I could 

not reach the original reference and as I do not know whether this species is iteroparous or semelparous, P aperta is not 

included in this present list. 



TABLE 2. Life spans in freshwater snails 



Species 



Lifespan 



Authority 



PROSOBRANCHIA 
Neritidae 

Nehtina granosa 

Theodoxus fluviatilis 

Viviparidae 
Campeloma rufum 
Viviparus ater 
Viviparus contectoides 
Viviparus georgianas 
Viviparus malleatus 
Viviparus viviparus 

Hydrobiidae 
Amnicola limosa 
Falsihydrobia streletzl<iensis 
Hydrobia acuta 
Hydrobia pusilla 
Hydrobia ulvae 
Hydrobia ventrosa 
Potamopyrgus antipodarum 
Potamopyrgus jenl<insi 

Bithyniidae 
Bithynia leachi 
Bithynia tentaculata 

Valvatidae 

Valvata cristata 
Va I vat a liumeralis 
Valvata piscinalis 
Valvata pulchella 

Pleurocendae 
Leptoxis carinata 

Thiaridae 
Brotia hainanensis 
Melanoides tu bereu I at a 
Melanopsis costata 



1-3 
3-4 
5 
11 

SL 
SL 
SL 
SL 
SL 
SL 
SL 
SL 

SL 
2-3 

SL 

SL 
SL 
SL 

SL 

3 

SL 

6 



Ford, 1987 

Fretter & Graham, 1978 

Van Cleave & Aitnnger, 1937 
Ribi & Gebhardt, 1986 
Van Cleave & Lederer, 1932 
Buckley, 1986 
Stanzykowska et al., 1971 
Spoel, 1958 

Pinel-Alloul & Magnin, 1973 
Chukhchin, 1978 
Chukhchin, 1978 
Chukhchin, 1978 
Kondratenkov, 1978 
Chukhchin, 1978 
Winterbourn, 1970 
Winterbourn, 1970 

Fretter & Graham, 1978 
Lilly, 1953 

Fretter & Graham, 1978 
Calow, 1978 
Calow, 1978 
Zaika, 1973 

Aldridge, 1982 

Dudgeon, 1982 
Dudgeon, 1986 
Ra'anan, 1986 



PULMONATA: 

Lymnaeidae 
Acella haldemani 
Austropelpa vinosa 
Lymnaea elodes 
Lymnaea humilis 
Lymnaea natalensis 



SL 
SL 

SL 
SL 
SL 



Calow, 1978 
Blair & Finlayson, 1981 
Calow, 1978 
Calow, 1978 
Fashuyi, 1981 



HELLER 



285 



TABLE 2. Life spans in freshwater snails 



Species 



Lifespan 



Authority 



Lymnaea palustris 
Lymnaea peregra 
Lymnaea trunculata 
Lymnaea stagnalis 

Physidae 
Aplexa hypnorum 
Physa acuta 
Physa fontinalis 
Physa gynna 
Physa integra 
Physa virgata 

Planorbidae 
Anisus vortex 
Armiger cnstata 
Biomphalaria glabrata 
Biomphalana pfeiffen 
Bulinus forsl<alii 
Bulinus globosus 
Bulinus nasutus 
Helisoma tnvolis 
Planorbis albus 
Planorbis carinatus 
Planorbis con tort и s 
Planorbis corneus 
Planorbis planorbis 
Planorbis vortex 

Ancylidae 
Ancylus fluviatilis 
Ancylus lacustris 
Ferrissia rivularis 
Hebetancylus excentricus 
Laevapex fuscus 



SL 
SL 
SL 
SL 

SL 
SL 
SL 
SL 
SL 
SL 

SL 
SL 

SL 
SL 
SL 
SL 
SL 
SL 
SL 
SL 
SL 
SL 
SL 
SL 

SL 
SL 
SL 
SL 
SL 



Browne & Russell-Hunter, 1978 
Calow, 1978 
Calow, 1978 
Berne, 1965 

Calow, 1978 
Calow, 1978 
Calow, 1978 
Calow, 1978 
Calow, 1978 
Calow, 1978 

Zaika, 1973 

Richardot & Alfaro, 1985 
Appleton, 1978 
Appleton, 1978 
Fashuyi, 1981 
Fashuyi, 1981 
Brown, 1980 
Eversole, 1978 
Calow, 1978 
Calow, 1978 
Calow, 1978 
Calow, 1978 
Calow, 1978 
Calow, 1978 

Durrant, 1980 
Calow, 1978 
Calow, 1978 
Calow, 1978 
Calow, 1978 



Notes to Table 2: 

Nentina granosa is a rheophllic gastropod endemic to Hawaiian freshwater streams. The species is diadromus. The female 

reproduces thousands of planktivorous veligers. that accumulate at stream mouths (Ford, 1987). 

Campeloma is a freshwater snail of North America that breeds parthenogenetically (Van Cleave & Altringer, 1937). 

In Vivíparas contectoides the males live slightly longer than one year, but the females live about three years (Van Cleave 

& Lederer, 1932). 

In Viviparus georgianas males live for three years, females for four (Buckley, 1986). 

Btthynia tentaculata lives only up to 2 years in the Bielorussian lakes (Zaika, 1973). 

Falsihydrobia streletzkiensis is similar to IHydrobia in various aspects of its morphology, but differs in its genitalia. Its 

taxonomic assignment at the family level is still unclear (Chukhchin, 1978). 

Leptoxis carinata. a freshwater cerithiacean of North America, is a semelparous biennial (Aldridge, 1982). 

Melanoides tuberculata is an ovoviviparous, usually parthenogenetic snail. In Hong Kong, studies at the population level 

suggest that the life span is at least one year and at the most two, with a single peak in juvenile recruitment coinciding with 

the warmer months (Dudgeon, 1986). However, although release of hatchlings is strictly seasonal, fully developed larvae 

are found m the brood pouches throughout the year. In Malaysia (Berry & Kadri. 1974) snails reach a life-span of 3 1/2 years, 

as extrapolated from laboratory growth rates. 

Melanopsis is the most common freshwater snail in Israel. Isolated pairs of /W costata were kept by Ra'anan (1986) in 

captivity for SIX years. 



286 



LONGEVITY IN MOLLUSCS 



TABLE 3. Life spans in landsnails. 



Species 



Lifespan 



Authority 



Veronicellidae 

Vaginulus borellianus 
Veronicella ameghini 

Ellobiidae 
Carychium tridentatum 
Me I am pu s sp. 
Ovatella myosotis 

Achatineilidae 
Achatinella musteline 

Cochlicopidae 
Cochlicopa lubrica 

Vertiginidae 
Vertigo pusilla 

Chondrinidae 
Solatopupa similis 

Enidae 
Brephulopsis bidens 

Clausiliidae 
Cochlodina lam i nata 
Vestía elata 

Cerionidae 
Cerion spp. 

Achatinidae 
Achat i na achat i na 
Achatina fúlica 
Archachatina margínala 

Endodontidae 
Discus rot und at us 
Punctum pygmaeum 

Arionidae 
Arion ater 

Arion circumscriptus 
Arion hortensis 
Arion intermedius 
Arion subfuscus 

Succineidae 
Catinella arenaria 
Omalonyx felina 
Succinea ovalis 

Vitrinidae 
Eucobresia nivalis 
Semilimax kotulai 
Vitrina alaskana 
Vitrina pellucida 

Zonitidae 
Aegopinella nitidula 
Aegopinella nitens 
Oxychilus cellahus 
Oxychilus helveticus 



SL 
SL 


Runham & Hunter, 1970 
Dundee. 1977 


SL 

4 
4 


Morton, 1954 
Apley, 1970 
Meyer, 1955 


9 


Hadfield & Mountain, 1980 


3 


Unninski & Focht, 1979 


SL 


Pokryszko, 1986. 


8 


Boato & Rasotto, 1 987 


SL 


Livshitz & Shileyko, 1978; Livshitz, 1i 


2 

7 


Cameron, 1982 
Piechocki, 1982 


10 


Woodruff, 1978 


5 
8 

11 


Hodasai, 1979 
Mead, 1961 
Plummer, 1982 


SL 
SL 


Cameron, 1982 
Baur, pers. comm. 


SL 
SL 
SL 
SL 
SL 


Runham & Laryea, 1968 

Godan, 1983 

Bett, 1960; Hunter, 1968 

Godan, 1983 

Bett, 1960 


SL 
SL 
SL 


Baker, 1965 
Shrader, 1974 
Strandine, 1941 


SL 
SL 
SL 
SL 


Uminski, 1979 

Uminski, 1975 

Boag & Wishart, 1982 

Taylor, 1907; Uminski & Focht, 1979 


SL 
SL 
SL 
SL 


Mordan, 1978 
Mordan, 1978 
Mordan, 1978 
Cameron, 1982 



HELLER 



287 



TABLE 3. (continued) 



Species 



Lifespan 



Authority 



Euconulidae 
Euconulus fulvus 

Milacidae 
Milax budapestensis 
Milax sowerbll 
Milax gagates 

Limacidae 
Blelzia coerulans 
Deroceras caucaslcum 
Deroceras reticulatum 
Deroceras sturanyi 
Umax flavus 
Umax maximus 

Parmacellidae 
Parmacella rutellum 

Bulimulidae 
Bullmulus dealbatus 
Liguus fasclatus 

Elonidae 
Elona quimperlana 

Testacellidae 
Testacella sp. 

Polygyridae 
AI logon a profunda 
Mesodon roemen 
Polygyra thyroideus 

Oleacinidae 
Euglandina rosea 

Pleurodontidae 
Caracolus caracolas 

Camaenidae 
Ampllrhagada napierana 

Sphincterochilidae 
Sphincterochlla prophetarum 
Sphincterochila zonata 

Helminthoglyptidae 
Helmlnthoglypta arrosa 

Bradybaenidae 
Bradybaena fruticum 

Heiicidae 
Arlanta arbustorum 
Cepaea nemoralls 
Cernuella virgata 
Eobania vermlculata 
Hellcella caperata 
Helix aspersa 
Helix lucorum 
Helix pom at I a 
Levantina hierosolyma 
Monacha cartusiana 
Monacha haifaensis 
Theba pisana 



SL 



10 



15 
15 

10 



17 
9 

SL 
5 

SL 
5 
5 

15 
7 

SL 

SL 

SL 



Uminski & Focht, 1979 



SL 
SL 
SL 


Hunter, 1968 

Bett, 1960 

Focardi & Quattrini, 1972 


SL 
SL 
SL 
SL 
SL 
SL 


Smolenska, 1936 
Uvalieva, 1978 
Godan, 1983 
Kosinska, 1980 
N. Runham, pers. comm. 
N. Runham, pers. comm. 


SL 


Uvalieva, 1978 


SL 
6 


Randolph, 1973 

Voss, 1976; Tuskes, 1981 


SL 


Daguzan, 1982 


6 


Taylor, 1907 


4 
3 
4 


Blinn, 1963 

Randolph, 1973 

Van Cleave & Foster, 1937 



Chiu & Chou, 1962 

Heatwole & Heatwole, 1978 

Solem & Christensen, 1984 

Steinberger, pers. comm. 
Steinberger, pers. comm. 

Laan, 1971; Pilsbry, 1939 

Comfort, 1957 

Raboud, 1986 
Cook & Cain, 1980 
Lazaridou, 1981 
Lazaridou, pers. com. 
Baker, 1968 
Lazaridou, pers. com. 
Staikou & Lazaridou, 1986 
Falkner, 1984 
pers. observations 
Chatfield, 1968 
pers. observations 
Heller, 1982; Cowie, 1984 



(continued) 



288 LONGEVITY IN MOLLUSCS 

TABLE 3. (continued) 



Species Lifespan Authority 

Trichia hispida SL Cameron, 1982 

Trochoidea simulate 3 Yom-Tov, 1971 

Xeropicta arenosa SL Lazaridou, 1981 

Xeropicta vestalis SL Heller & Volokita, 1981 

Notes to Table 3: 

Vaginulus borellianus is an Argentinian slug that lives for about a year to 18 months^ Eggs are laid in a mucus envelope on 

the soil surface (Lanza & Quattrim, 1964; in Runham & Hunter, 1970). 

Veronicella ameghini is an introduced species in the southern USA, The suggestion that its longevity is 'likely around two 

years" (Dundee, 1977: 1 14) is a free estimate that is not based upon concrete facts. 

Achatinella mustelina is a tree-dwelling snail of Hawaii. 

Ovatella and Melampus dwell in salt marshes along sea coasts. They live above sea level, like land snails, but reproduce 

by veligers, as marine snails do. I arbitrarily classify them as terrestrial. Ovatella myosotis first develops the masculine 

system and functions as a male, then also the female system and continues to function as both male and female (Meyer, 

1955). 

Caryctiium tridentatum is a pnmitive pulmonate that lives m a saturated atmosphere under fallen leaves and logs. The snails 

change sex throughout their lifetime: a penod of 12 months is required for the completion of a single sperm-producing 

phase, followed by a single egg-producing one. Morton suggests that the snails appear to have a "double-phase" semelp- 

arous reproductive strategy. I accept Mortons semelparous interpretation, but with heavy doubts, as his fig. 2 suggests that 

at any time of the year there are not nearly enough juveniles in the population to replace the much larger adult group. His 

data may well suggest that Carychium is an iteroparous, long-lived species with a few juvenile snails joining the population 

and a few adults dying off each year. 

Omalonyx felina is a tropical succineid of Venezuela. 

Punctum pygmaeum is a minute (1.5 mm) snail that has a Holarctic distribution. Its biology m Sweden is currently being 

studied by B. Baur. 

Euconulus fulvus and Discus cronkhitei in Canada have, on their shell, "one or more varices which suggests that they 

survive one or more winters" (Boag & Wishart, 1982: 2636). 

Aegopinella nitidula has a bienniel life cycle with delayed maturity and overlapping generations, and Mordan (1978) 

suggests that this may be advantageous in unstable environmental conditions. 

Umax flavus and L. maximus are generally annual species and hence short-lived (N. Runham, personal communication). 

Comfort (1957) mentions them as living 5 years (based upon animals studied m captivity), and Godan (1983) suggests that 

they live three years. Since N. Runham has been personally involved in studying them, I prefer his evidence. 

Elona quimpehana matures within two years, and lives for another year and a half (Daguzan, 1982). Its classification as a 

short-lived species stretches the definition of "short-lived" to its limit. 

Liguas fasciatus is a tree snail of Antillean origin that is found in tropical hardwood trees and exhibits great variability m shell 

coloration. Voss (1976) suggests that reproduction occurs at the end of the fourth year, after which many snails die, and it 

should therefore be classified as exhibiting a semelparous strategy. However, the size distnbutions m his fig. 1-2 suggest 

an iteroparous cycle, with a few juveniles joining the adult population every year. In addition, his table 1 shows an increase 

in size of the yearly classes, and this can only be explained by the slow accumulation of individuals into the various size 

classes over several years, namely an iteroparous strategy, with a very long life-span. Also Tuskes (1981), when studying 

Liguus fasciatus. reached conclusions diffenng considerably from those of Voss. 

Euglandina rosea is a carnivorous snail that feeds mainly upon other land snails. Sphmcterochila zonata and S. 

proptietarum are found in the Negev Desert, where they were studied by Yom-Yov (1971), who suggested that S. zonata 

lives more than 8 years, and Y. Stemberger (unpublished data), who informs me that they live 15 years at least. 

Levantina fiierosolyma is found in Mediterranean to and habitats of the Middle East, where it dwells m rock-crevices and 

beneath stones. 

Cernuella virgata, a European species, maintains a short life span with an annual life cycle also in populations introduced 

into Australia (Pomeroy, 1969). 

Eobania vermiculata in Greece is found at the lower parts of the vegetation. Sexual maturity is reached in two years, and 

it may then live for another three years (Lazaridou, pers. com,). 



HELLER 



289 



TABLE 4. Life spans in marine bivalves 



Species 


Lifespan 


Authority 


Nuculidae 






Acila insignis 


9 


Zoiotarev, 1980 


Nucula annulata 


8 


Cerrato, 1980 


Nucula nucleus 


12 


Comfort, 1964 


Nucula sulcata 


17 


Comfort, 1964 


Nucula túrgida 


10 


Comfort, 1957 


Nuculanidae 






Nuculana minuta 


7 


Ansell & Parulekar, 1978 


Nuculana pernula 


9 


Zoiotarev, 1980 


Yoldia limatula 


4 


Powell & Cummins, 1985 


Malletiidae 






Tindana callistiformis 


100 


Turekian et al., 1975 


Arcidae 






Area boucardi 


20 


Zoiotarev, 1980 


Anadara broughtoni 


46 


Zoiotarev, 1980 


Senilia senilis 


9 


Powell & Stanton, 1985 


Glycymendae 






Glycymeris yessoensis 


64 


Zoiotarev, 1980 


Mytilidae 






Bathymodiolous thermophila 


19 


Rhoads et al., 1981 


Brachiodontes variabilis 


3 


Powell & Cummins, 1985 


Crenomytilus grayanus 


150 


Jones, 1983 


Geukensia demissa 


23 


Lutz & Castagna, 1980 


Modiolus demissus 


8 


Zaika, 1973 


Modiolus modiolus 


61 


Zoiotarev, 1980 


Mytilaster lineatus 


3 


Zaika, 1973 


Mytilus californiensis 


5 


Cerrato, 1980 


Mytilus coruscus 


39 


Zoiotarev, 1980 


Mytilus edulis 


15 


Zoiotarev, 1980 


Mytilus galloprovinciallis 


12 


Powell & Stanton, 1985 


Mytilus vanabilis 


5 


Comfort, 1957 


Perna viridis 


3 


Lee, 1985 


Septifer keenae 


15 


Zoiotarev, 1980 


Dreisseneidae? 






Mytilopsis sallei 


SL 


Morton, 1981 


Pinnidae , 






Pinna atropurpura 


12 


Butlers Brewster, 1979 


Pteriidae 






Pinctada martensii 


8 


Powell & Cummins, 1985 


Pinctada vulgaris 


7 


Comfort, 1957 



Pectinidae 
Adamusium colbecki 
Amusium balloti 
Amusium japonicum 
Argopecten gibbus 
Argopecten irradians 
Argopecten japonicum 
Chlamys albidus 
Chlamys islándica 
Chlamys opercularis 
Chlamys varia 
Notovola meridionalis 
Patinopecten caurinus 
Patinopecten yessoensis 
Pectén maximus 



10 


Ralph & Maxwell, 1977 


4 


Powell & Cummins, 


1985 


4 


Williams & Dredge, 


1981 


SL 


Williams & Dredge, 


1981 


SL 


Sastry, 1979 




SL 


Powell & Cummins, 


1985 


8 


Zoiotarev, 1980 




23 


Powell & Cummins, 


1985 


6 


Williams & Dredge, 


1981 


7 


Powell & Cummins, 


1985 


11 


Williams & Dredge, 


1981 


15 


Powell & Cummins, 


1985 


12 


Ventilla, 1982 




12 


Cerrato, 1980 





(continued) 



290 



LONGEVITY IN MOLLUSCS 



TABLE 4. (continued) 



Species 



Lifespan 



Authority 



Placopecten magellanicus 
Swiftopecten swiW 

Ostreidae 
Crassostrea madrasensis 
Crassostrea virginica 
Ostrea edulis 

Lucinidae 
Cavatidens omissa 

Thyasiridae 
Thyasira flexuosa 

Ungulinidae 
Felaniella usta 

Galeommatidae 
Lasaea rubra 

Montacutidae 
Mysella bidentata 
My sell a cunéala 
Mysella planulata 

Carditidae 

Venericardia crebricostata 

Cardiidae 
Cardium ciliatum 
Cardium corbis 
Cardium edule 
Cardium corbis 
Cerastoderma glaucum 
Clinocardium nuttallii 
Keenocardium californiese 
Serripes groenlandicus 

Mactridae 
Mactra sulcataria 
Mulinia lateralis 
Rangia cuneata 
Spisula sachalinensis 
Spisula solidissima 
Spisula voyi 
Tresus capax 

Mesodesmatidae 
Mesodesma ventricosum 

Solenidae 
Solan corneus 
Solen krustensterni 

Cultellidae 
Ensis siliqua 
Siliqua alta 
Siliqua patula 

Teiiinidae 
Cadella lubrica 
Gastrana contabulata 
Macoma balthica 
Macoma calcárea 
Macoma litoralis 
Macoma middendorffi 
Peronidia venulosa 



12 
15 

4 

6 

20 

SL 

SL 

9 



Williams & Dredge, 1981 
Zolotarev, 1980 

Powell & Cummins, 1985 
Comfort, 1957 
Chiristensen & Dance, 1980 

Powell & Cummins, 1985 

Lopez-Jamar et a!., 1987 

Zolotarev, 1980 

McGrath & O'Foighil, 1986 



7 


Ockelmann & Muus, 1978 


6 


Gage, 1968 


4 


Franz, 1972 


58 


Zolotarev, 1980 


25 


Petersen, 1978 


16 


Powell & Cummins, 1985 


7 


Cerrato, 1980 


10 


Cerrato, 1980 


7 


Powell & Stanton, 1985 


14 


Zolotarev, 1980 


11 


Zolotarev, 1980 


22 


Petersen, 1978 


12 


Zolotarev, 1980 


3 


Cerrato, 1980 


10 


Powell & Cummins, 1985 


55 


Zolotarev, 1980 


31 


Jones et al., 1978 


52 


Zolotarev, 1980 


16 


Powell & Cummins, 1985 



Comfort, 1957; Powell & Cummins, 1985 



5 


Powell & Cummins, 1985 


12 


Zolotarev, 1980 


12 


Comfort, 1964 


24 


Zolotarev, 1980 


17 


Cerrato, 1980 


17 


Zolotarev, 1980 


15 


Zolotarev, 1980 


18 


Zolotarev, 1980 


17 


Petersen, 1978 


6 


Powell & Cummins, 1985 


24 


Zolotarev, 1980 


31 


Zolotarev, 1980 



HELLER 



291 



TABLE 4. {continued) 



Species 



Lifespan 



Authority 



Peronidia zyonoensis 
Tellina alternata 
Tellina deltoidalis 
Tellina tenuis 

Donacidae 
Donax denticulatus 
Donax gouldii 
Donax incarnatus 
Donax hanleyanus 
Donax semistnatus 
Donax serra 
Donax sordidus 
Donax spiculum 
Donax trunculus 
Donax túmida 
Donax variabilis 
Donax venustus 
Donax vi tt at и s 

Psammobidae 
Gari kazunensis 
Nuttallia ezonis 
Nuttallia olivácea 

Scrobiculariidae 
Scrobicularia plana 

Semelidae 
Abra ovata 
Cumingia tellinoides 
Theora fragilis 

Solecurtidae 
Tagelus divisus 

Arcticidae 
Árctica islándica 

Vesicomyidae 
Calyptogena magnifica 

Veneridae 
Anomalocardia squamosa 
Callista brevisiphonata 
Callista chione 
Callithaca adamsi 
Dosinia angulosa 
Dosinia elegans 
Dosinia exoleta 
Dosinia hepática 
Dosinia japónica 
Gemma gemma 
Kate! y si a opima 
Mercenaria mercenaria 
Mercenaria stimpsoni 
Protothaca euglypta 
Protothaca jedoensis 
Protothaca staminea 
Tapes phillippinarum 
Tivela stultorum 
Venerupis japónica 
Venerupis pullastra 



61 


Zolotarev, 1980 




3 


Powell & Cummins, 


1985 


4 


Powell & Cummins, 


1985 


5 


Comfort, 1957 




SL 


Powell & Cummins, 


1985 


3 


Powell & Cummins, 


1985 


3 


Powell & Cummins, 


1985 


3 


Ansell, 1983 




SL 


Ansell, 1983 




SL 


Ansell, 1983 




SL 


Powell & Cummins, 


1985 


SL 


Powell & Cummins, 


1985 


3 


Ansell, 1983 




SL 


Powell & Cummins, 


1985 


SL 


Ansell, 1983 




SL 


Ansell, 1983 




7 


Ansell, 1983 




14 


Zolotarev, 1980 




40 


Zolotarev, 1980 




20 


Zolotarev, 1980 




18 


Comfort, 1957 




4 


Zaika, 1973 




4 


Comfort, 1957 




SL 


Powell & Cummins, 


1985 


3 


Powell & Stanton, 1985 


220 


Jones, 1983 




11 


Jones, 1983 




3 


Powell & Stanton, 1985 


76 


Zolotarev, 1980 




40 


Powell & Cummins, 


1985 


29 


Zolotarev, 1980 




26 


Zolotarev, 1980 




3 


Powell & Cummins, 


1985 


7 


Comfort, 1964 




6 


Powell & Cummins, 


1985 


27 


Zolotarev, 1980 




SL 


Sellmer, 1967 




3 


Powell & Cummins, 


1985 


9 


Kennish, 1980 




40 


Zolotarev, 1980 




14 


Zolotarev, 1980 




15 


Zolotarev, 1980 




13 


Powell & Cummins, 


1985 


SL 


Powell & Cummins, 


1985 


14 


Cerrato, 1980 




25 


Zolotarev, 1980 




9 


Cerrato, 1980 





(continued) 



292 



LONGEVITY IN MOLLUSCS 



TABLE 4. Life spans in marine bivalves 



Species 



Lifespan 



Authority 



Venus gallina 
Venus mercenaria 
Venus striatula 

Myidae 
Mya arenaria 
Mya japónica 
Mya priapus 
Mya truncata 

Corbulidae 
Aniscorbula venusta 
Corbula trigona 
Corbula vicaria 
Potamocorbula amurensis 

Hiatellidae 
Hiatella byssifera 
Panope generosa 

Teredinidae 
Bankia gouldi 
Teredo bartschi 
Teredo navalis 

Pandoridae 
Pandora pulchella 

Laternulidae 
Laternula elliptica 



Cerrato, 1980 



15 


Cerrato, 1980 


10 


Guillou & Sauriau, 1985 


28 


Jones, 1983 


42 


Zolotarev, 1980 


15 


Zolotarev, 1980 


18 


Petersen, 1978 


8 


Zolotarev, 1980 


SL 


Maslin & Bouvet, 1986 


4 


Powell & Cummins, 1985 


5 


Zolotarev, 1980 


15 


Petersen, 1978 


120 


Jones, 1983 


SL 


Hoagland, 1986 


SL 


Hoagland, 1986 


SL 


Hoagland, 1986 


11 


Zolotarev, 1980 


13 


Ralph & Maxwell, 1977 



Notes to Table 4: 

Tindaria callistiformis is a minute (8.6 mm) nuculanacean that lives on the sea bottom, at a depth of 3,800 m (Turekian et 

al., 1975). 

Comely 1978 suggests that M. modiolus lives only 35 years. 

Data for Perna viridis concern a polluted habitat where the mussels suffer precocious mortality due to unnaturally stressfull 

conditions (Lee, 1985). 

Mysella bidentata lives In association with the ophuroid Amphiura. In the second year of its life it functions as a male: from 

three years onwards it is a hermaphrodite. (Ockelmann & Muus, 1978). 

Mysella cuneata. a bivalve of minute size (up to 3 mm) is a commensal of a sipunculid which occupies discarded shells 

(Gage, 1968). 

Mysella planulata. of minute size (4 mm), lives in muddy sands. It is a simultaneous hermaphrodite (Franz, 1972). 

Lasaea rubra is an intertidal bivalve of minute size (3.2 mm). It is ovoviviparous (f\/lcGrath & O'Foighil, 1986). 

Calyptogena magnifica and Bathymodiolous thermica belong to the hydro-thermal vent fauna of Galapagos. The biology of 

these species is descnbed by Childress et al., 1987. 

Donax vifíatus has a life-span of 3 years at the soutnern end of its geographical range, but longevity increases at higher 

latitudes and may reach 7 years in northern populations (Ansell, 1983). 

Corbula trígona dwells in coastal lagoons in western Africa (tVlaslin & Bouvet, 1986). 

Teredo is a highly specialised bivalve adapted for bonng into wood. Its average life-span in Miami is about 10 weeks (Nair 

& Saraswathy, 1971). 



HELLER 



293 



TABLE 5. Lifespans of freshiwater bivalves 



Species 



Lifespan 



Authority 



Mytilidae 
Li m nope r па fort un ei 

Unionidae 
Amblema plicata 
Anodonta anatina 
Anodonta californiensis 
Anodonta corpulenta 
Anodonta gibbosa 
Anodonta imbecilis 
Anodonta minima 
Anodonta peggyae 
Anodonta piscinalis 
Anodonta woodiana 
Anatontoides subcylindraceus 
Elliptic complanata 
Elliptio dilatata 
Lampsilis anodontoides 
Lampsilis ovata 
Lampsilis recta 
Lampsilis siliquoidea 
Margaritifera margaritifera 
Pleurobema coccineum 
Pleurobema cordât и m 
Quadrula sp. 
Trifogonia verrucosa 
Unio crassus 
Unio pictorum 
Unio tumidus 

Dreissenidae 
Dreissena polymorpha 

Corbicuiidae 
Corbicula fluminea 
Corbicula cf. fluminalis 

Spfiaeriidae 
Byssanodonta cubensis 
Pisidium amnicum 
Pisidium annandalei 
Pisidium casertanum 
Pisidium clarkeanum 
Pisidium compressum 
Pisidium hibernicum 
Pisidium lilljeborgi 
Pisidium variabile 
Sphaerium corneum 
Sphaerium fabalis 
Sphaerium occidentale 
Sphaerium rivicola 
Sphaerium simile 
Sphaerium solidum 
Sphaerium striatinum 
Sphaerium transversum 
Sphaerium partumeium 
Musculium japonicum 
Musculium lacustre 
Musculium partumeium 
Musculium secuhs 
Musculium transversum 



SL 


Morton. 1977 




16 


Comfort, 1957 




10 


Neguus, 1966 




5 


Heard, 1975 




8 


Heard, 1975 




16 


Heard, 1975 




12 


Heard, 1975 




10 


Neguus, 1966 




15 


Heard, 1975 




15 


Comfort, 1957 




12 


Morton, 1986 




9 


Comfort, 1957 




12 


Matteson, 1948 




12 


Comfort, 1957 




8 


Comfort, 1957 




19 


Comfort, 1957 




18 


Comfort, 1957 




19 


Comfort, 1957 




116 


Hendelberg, 1960; Smith), 


1976; Bauer, 1987 


12 


Comfort, 1957 




30 


Yokley, 1972 




50 


Comfort, 1957 




11 


Comfort, 1957 




15 


Comfort, 1957 




15 


Neguus, 1966 




11 


Neguus, 1966 




5 


Morton, 1969 




4 


Morton, 1986 




10 


Morton, 1986 




3 


Mackie & Huggins, 1976 




SL 


Baas, 1979 




SL 


Morton, 1986 




SL 


Mackie, 1984 




SL 


Morton, 1986 




3 


Meier-Brook, 1970 




3 


Meier-Brook, 1970 




3 


Meier-Brook, 1970 




SL 


Mackie, 1979 




SL 


Dussart, 1979 




SL 


Mackie, 1979 




SL 


Heard, 1977 




SL 


Heard, 1977 




3 


Avolizi, 1976 




SL 


Heard, 1977 




SL 


Mackie, 1984 




SL 


Gale, 1977 




SL 


Gale, 1977 




SL 


Heard, 1977 




SL 


Morton, 1986 




SL 


Mackie, 1984 




SL 


Mackie, 1979 




SL 


Mackie, 1984 


(continued) 



294 LONGEVITY IN MOLLUSCS 

Notes to Table 5: iq77\ 

L/mnoperna tor/une/ is an inhabitant of freshwater rivers and streams in China and southeast Asia (Morton, 1977). 

Margantifera marganMera is a slow-growing mussel that takes about 20 years to reach sexual maturity. 

Within the Sphaeriidae, Heard (1 977) suggests that most P,s,d^um and Sphaenum inhabit permanent lentic and lotic waters. 

in contrast to most Muscu//um that are found in ephemeral habitats. 

P,s/d-um c/ar/ceanum is generally iteroparous, but may also be semelparous. It lives for 4-8 months (Morton^ 1986). 

Sphaenum corneum lives for about 4-8 months in Canada, one year in Germany and Russia, but may live 3-4 years in 

Sweden (Heard, 1977). 

Sphaenum simile in New York may live up to 4-5 years (Heard. 1977). 

Sphaenum transversum may reach densities of 10,000/m^ It can complete its life history in less than a month (Gale, 1977). 



HELLER 



295 



TABLE 6. Lifespans in cephalopods 



Species 



Lifespan 



Aufhority 



NAUTILOIDEA 
Nautilidae 
Nautilus pompilus 

COLEOIDEA. 
Spirulidae 
Spirula spirula 

Sepiidae 
Sepia officinalis 

Sepiolidae 
Euprymna scolopes 
Rossia pacifica 
Sepietta oweniana 
Sepiola robusta 

Loliginidae 
Loligo forbesi 
Loligo opalescens 
Loligo pealei 
Loligo vulgaris 
Sepioteutfiis sepiola 

Gonatidae 
Gonatus fabrica 

Ommastrephidae 
Dosidicus gigas 
lllex illecebrosus 
Tadarodes pacificus 

Cranchiidae 
Teuthowenia megalops 

Octopodidae 
Batfiypolypus arcticus 
Eledone cirrhosa 
Eledone moschiata 
Octopus briareus 
Octopus cyanea 
Octopus dofleini 
Octopus joubini 
Octopus maya 
Octopus tetricus 
Octopus vulgaris 



20 



SL 

SL 

SL 
SL 
SL 
SL 

SL 
SL 
SL 
SL 
SL 

SL 

SL 
SL 
SL 

SL 

SL 
SL 
SL 
SL 
SL 
SL 
SL 
SL 
SL 
SL 



Saunders, 1984 



Comfort, 1957 

Boletzky, 1983a 

Singley, 1983 
Anderson, 1987 
Bergstrom & Summers, 1983 
Boletzky, 1983b 

Holme, 1974 

Hixon, 1983 

Summers, 1983 

Worms, 1983 

Moynihan & Rodaniche, 1983 

Kristensen, 1983 

Nesis, 1983 
O'Dor, 1983 
Okutani, 1983 

Nixon, 1983 

O'Dor & tVlacalaster, 1983 

Boyle, 1983 

Mangold, 1983b 

Hanion, 1983a 

Van Heukelem, 1983a 

Hartwick, 1983a 

Hanion, 1983b 

Van Heukelem, 1983b 

Joli, 1983 

Mangold, 1983 



Notes to Table 6: 

Octopus dofleini is one of the largest octopod species, with an arm span of up to 9.6 m and a weight of 272 kg (Hochberg 

& Fields, 1 980, as cited in Boyle, 1 987, table 1 6. 1 ). It takes 2-3 years to reach maximum size. Both males and females stop 

eating and die after the reproductive period, but males may perhaps live 1 or 2 years longer if they don't reproduce 

{Hartwick, 1983), 

Bathypolypus arcticus is a deep sea octopus that lives at depths of 1000 m, in temperatures that rarely range above 6=C. 

It requires nearly 4 years to complete its life cycle: one year of embryonic development, one of growth, one of gametoge- 

nesis and one of brooding (O'Dor & Macalaster, 1983). 



MALACOLOGIA, 1990, 31(2): 297-312 



COMPARATIVE MORPHOLOGY OF LIVING NAUTILUS (CEPHALOPODA) 
FROM THE PHILIPPINES, FIJI AND PALAU 

Kazushige Tanabe\ Jyunzo Tsukahara^ & Shozo Hayasaka^ 



ABSTRACT 

Morphological features of Nautilus from the Philippines, Fiji and Palau are cornpared from a 
taxonomic viewpoint on the basis of live-caught animals. In spite of their widely separated 
distnbutions, animals from the three populations share quite similar overall shell morphology, 
ontogenetic shell variation, and radular and jaw structures. Shell coloration and sculpture, and 
the shape of radular teeth, all of which have been used in previous taxonomic studies, are also 
markedly variable even in specimens of individual populations, and their ranges of variation 
overlap among the three samples. The three samples can be distinguished mainly by adult 
features, such as the dimensions of the shells and total number of septa, which are probably 
attributed to the difference in their pre-reproductive ages. Judging from these observations and 
available genetic data, it is suggested that the Palau population, previously distinguished as N. 
belauensis and the other two populations belong to the same, wide ranging species, N. pom- 
pilius. or otherwise they are closely related sibling species, N. belauensis and N. pompilius 
respectively. 

Key words: Nautilus pompilius. Nautilus belauensis. southwest Pacific, morphology, taxon- 
omy. 



INTRODUCTION 

The superfamily Nautilaceae (Ceph- 
alopoda, Nautiloidea) first appeared in the Tri- 
assic, and flourished mainly during the Meso- 
zoic and Middle Tertiary. They suddenly 
declined after the Miocene, and at the present 
time only a few species of the genus Nautilus 
survive, in the relatively deep waters of the 
tropical southwestern Pacific. 

Although 1 1 species and seven variants of 
Nautilus have hitherto been proposed (see 
Saunders, 1987, table 1), their taxonomic va- 
lidity has long been obscured because of the 
seemingly morphological conservatism of the 
genus, extreme splitting of phenotypes based 
on small collections, and the lack of knowl- 
edge of the morphological and genetic varia- 
tion within individual populations. Recently, 
Saunders (1987) revised these "species" and 
variants into five or possibly six recognized 
species {Nautilus pompilius Linnaeus, 1758; 
N. macromphalus Sowerby, 1849; N. scrobic- 
ulatus [Lightfoot, 1786]; N. stenomphalus 
Sowerby, 1849; N. belauensis Saunders, 
1981 ; and possibly N. repeilus Iredale, 1944), 
but some malacologists (e.g. Habe, 1980; Ab- 



bott & Dance, 1983) regard the latter three 
species as geographic variants of N. pompil- 
ius. 

The species-level taxonomy of Nautilus 
should, therefore, be re-examined in view of 
recent biométrie and electrophoretic analyses 
of large live-caught collections (Ward et al., 
1977; Tanabe et al., 1983, 1985; Saunders & 
Davis, 1985; Tanabe & Tsukahara, 1987; Ma- 
suda & Shinomiya, 1983; Woodruff et al., 
1983, 1987; Swan & Saunders, 1987), for 
these works detected marked morphological 
and genetic variation even within individual 
populations. 

This paper considers the taxonomic rela- 
tionships of two closely allied morphospecies, 
N. belauensis and N. pompilius, on the basis 
of the comparative morphologic examination 
of large collections from several populations. 

MATERIAL AND METHODS 

Material 

The following three samples of Nautilus 
populations from widely separated areas 
were used in this study: (1) 34 specimens (10 



'Geological Institute, University of Tokyo, Tokyo 113, Japan. 

^Institute of Biology, Faculty of Science, Kagoshima University, Kagoshima 890, Japan. 

^Institute of Earth Sciences, Faculty of Science, Kagoshima University, Kagoshima 890, Japan. 



297 



298 



TANABE, TSUKAHARA & HAYASAKA 



males and 24 females) of N. pompilius cap- 
tured with baited traps from off Bindoy Village 
(depth of 120-310 m), Tañon Strait, the Phil- 
ippines, in September 1981 (specimens B1- 
B32, B41 and B52 among 52 animals listed in 
Hayasaka et al., 1982, table 10); (2) A total of 
280 specimens (245 males, 34 females and 
one unsexed juvenile) of N. pompilius cap- 
tured alive from off Suva (Kandavu Passage; 
depth of 290-450 m), Viti Lebu, Fiji, on two 
occasions (August-September 1983 and 
1986; see Tanabe, 1985, fig. 5, tables 1-3, 
and Tanabe, 1988, fig. 3, tables 1-4, for their 
locations and biological data), and (3) 94 
specimens (57 males, 36 females and one 
unsexed juvenile) of N. belauensis captured 
live from eastern Mutremdiu Bay ( = Mutrem- 
diu Point of Saunders, 1 981 a, figure 1 ); depth 
of 190-400 m, off Augulpelu Reef, Palau, in 
August-September 1988 and in January 
1989. In addition to the above three live- 
caught samples, two specimens caught from 
off Siquijor Island, Bohol Strait, the Philip- 
pines (provided by the courtesy of native fish- 
ermen; sp. nos. SQ 1-2 in Hayasaka et al., 
1982) were used for comparison of ontoge- 
netic septal growth and mature shell size. 

The specimens illustrated are kept at the 
University Museum, University of Tokyo 
(UMUT), and the remaining ones used for 
measurements are deposited at the Geology 
and Biology Institutes, Kagoshima University. 
Of three (Habe, 1980) or possibly six (Saun- 
ders, 1987) currently recognized Nautilus 
species, N. pompilius has the widest geo- 
graphic range, extending from the Philippines 
in the northwest to American Samoa in the 
southeast (Saunders, 1987). The two sam- 
ples from the Philippines (Tañon Strait) and 
Fiji thus represent the western and eastern 
marginal populations of this species. The two 
specimens from Bohol Strait (Philippines) are 
compared with the morphotype distinguished 
as N. pompilius suluensis by Habe & Okutani 
(1988, figs. 1-4). N. belauensis is known only 
from Palauan waters, about 800 km from the 
range of N. pompilius. 

Methods 

Following the methods described in Tan- 
abe & Tsukahara (1 987), all animals captured 
were weighed, sexed, and measured (see 
Hayasaka et al., 1982, table 10, and the re- 
vised version in Tanabe et al., 1983, table 1 ; 
Tanabe, 1988, table 3; Tanabe & Tsukahara, 
1989, table 2). Some were tagged and re- 



leased near the sampling locations for long- 
term growth analysis, and most of the remain- 
ing animals were dissected, and their fresh 
soft tissues and gonads were weighed by 
means of a dial scale (accuracy ± 10 mg) for 
biometry. In addition, the buccal mass was 
removed from the body of selected speci- 
mens. It was soaked in a 20% KOH solution 
for 20 minutes, and thereafter the mandible 
and radular ribbon were carefully removed. 
The radular and jaw morphologies of each 
specimen were observed under the optical 
and scanning electron microscopes. 

We further analyzed the ontogenetic shell 
growth patterns in several specimens se- 
lected from each sample. For this purpose, 
radius vector (Я), breadth (ß), height (H) and 
flank length (F) of a whorl, and half length of 
umbilicus (C), disregarding secondary umbil- 
ical deposits (callus), which were measured in 
each dorso-ventrally sectioned shell at inter- 
vals of 180° using a profile projector (NIKON, 
VI 6), attached to a digital micrometer (accu- 
racy ± 1 |xm) (magnification x 20; see Tan- 
abe & Tsukahara, 1987, figure 1, for mea- 
surements). Based on these measurements, 
four geometric parameters; i.e. whorl expan- 
sion rate [(FtJRn-Af'^ " >1тт], flank position 
(F/D), whorl inflation (B/hf) and involution {C/ 
R) at different growth stages were calculated 
for each specimen. 



SHELL MORPHOLOGY 

Gross Morphology and Coloration 

The shells of the Palau, Philippine (Tañon) 
and Fiji Nautilus essentially resemble one an- 
other in overall morphology and shell colora- 
tion. Their whorls are tightly coiled with a nar- 
row umbilicus, mostly filled with a callus in the 
middle-late growth stages. The shell colora- 
tion consists of two elements, i.e. irregular 
reddish brown to brown serrate radial stripes 
extending from the inner flank to venter and 
branching across the mid-flank, and a longi- 
tudinal stripe of the same color around the 
umbilical area (Fig. 1). In mature and almost 
mature shells, the former element tends to 
disappear toward the aperture, retaining only 
its trace on the inner flank. The mode of dis- 
tribution, strength and hue of the shell color- 
ation is fairly variable even in the specimens 
from the same area, but the Fiji sample con- 
sists mostly of the phenotype with relatively 
short and broad radial stripes (Fig. 1). 



COMPARATIVE MORPHOLOGY OF NAUTILUS 



299 







/ 




E --4 





FIG 1 Mature shells of Nautilus belauensis (A-B) and Nautilus pompllius (C-G), showing the similarity in 
overall morphology and coloration. A-B. Male (A: T3-2; UMUT RM 18708-3) and female (B: T9-3; UMUT RM 
18708-9) from Palau. C-D. Male (C: B21 ; UMUT RM 18705-3) and female (B30; UMUT RM 18705-7) from 
Tañon Strait, the Philippines. E-F. Male (E: SV6-1; UMUT RM 18707-2) and female (F: SV5-3; UMUT RM 
18707-1) from off Suva, Viti Lebu Island, Fiji. G. Sex-unknown specimen (SQ3; UMUT RM 18706-2) from 
Bohol Strait near Siquijor Island, the Philippines. Scale bar represents 5 cm. 



300 



TANABE, TSUKAHARA & HAYASAKA 




FIG. 2. Optical micrographs of the ventral shell surface of Nautilus belauensis (A) and Nautilus pompilius 
(B-D), showing longitudinally crenulated sculpture. A. UMUT RM 18708-2 (T2-4; female) from Palau. B. 
UMUT RM 18707-7 (SV13-8; female) from Fiji. С UMUT RM 18707-8 (SV13-13; male) from Fiji. D. UMUT 
RM 18705-8 (B31; female) from the Philippines (Tañon Strait). Scale bars indicate 500 fim. 



The whorls of every specimen exhibit 
dense sinuous growth lines. In addition, well- 
marked, longitudinally crenulated ridges 
showing a reticulate pattern are developed in 
every specimen from Palau. This sculpture 
was assigned by Saunders (1981a) as one of 
the diagnoses for distinguishing the Palauan 
N. belauensis from N. pompilius. However, It 
also occurs on the ventral side of many spec- 
imens of N. pompilius from Fiji and the Phil- 
ippines, although it is especially conspicuous 
in the Palau specimens (Fig. 2). 

Ontogenetic Shell Variation 

Biometrie analysis of selected specimens in 
dorso-ventral section reveals that the three 
samples exhibit similar ontogenetic patterns 



of shell geometric parameters, as repre- 
sented by the gradual decrease of whorl in- 
flation [B H) with increase of whorl number, 
sudden decline of flank position [F D) near 
the end of the first whorl, and abrupt increase 
and subsequent decline of distance of the 
whorls to the coiling axis (C R) in the second- 
third whorls (Fig. 3). In every sample, the 
ranges of variation of geometric parameters 
are larger in the early stage than in the later 
stage. The observed ranges of each param- 
eter at a given whorl stage in Fiji and Palau 
specimens mostly overlap each other, except 
for the larger С R ratio in the later stage of the 
Palau specimens. The umbilicus of every 
specimen is initially free from a callus. The 
callus begins to appear during the develop- 
ment of the second whorl, increasing its thick- 



0.7 t- 



0.6 



^05 



0.4 



0.3 



COMPARATIVE MORPHOLOGY OF K^WTILUS 



♦ Palau (N=7) 

oFiji (N = 21) 

* Philippines(N = 4) 



301 



J I L 




05 1.5 2.5 3.5 4.5 5.5 6.5 я^ 2.5 3.5 4.5 5.5 6.5 7t. 




0.5 1.5 2.5 3.5 4.5 5.5 6.5 tl 1.5 2.5 3.5 4.5 5.5 6.5 л^ 

TOTAL ROTATION ANGLE 

FIG. 3. Ontogenetic changes of whorl expansion rate [{Rr/Ftn-^)% frank position (F/D), whorl inflation rate 
(B/H), and whorl involution rate {C'R) versus total rotation angle of spiral for specimens of Nautilus belauen- 
sis from Palau and Nautiluä pompilius from the Philippines (Tañon Strait) and Fiji. Vertical bars indicate the 
range of one standard deviation. 



ness as the shell grows (Fig. 4). A complete ing the formation of the second whorl for the 
seal of the umbilicus by the callus occurs dur- Fiji and Philippine specimens, while it is de- 



302 



TANABE, TSUKAHARA & HAYASAKA 




FIG. 4. Drawings of cross-sectioned specimens of 
Nautilus belauensis from Palau (A-C) and Nautilus 
pompilius from the Philippines (Tañon Strait) (D-F) 
and Fiji (G-H). A. UMUT RM 18708-7 (T9-1, ma- 
ture male), B. UMUT RM 18708-8 (T9-2, mature 
male), С UMUT RM 18708-2 (T2-4, mature 
female), D. UMUT RM 18705-5 (B27, mature 
male), E. UMUT RM 18705-6 (829, mature male), 
F. UMUT RM 18705-4 (B22, submature female), G. 
UMUT RM 18707-3 (SV12-1, submature female), 
H. UMUT RM 18707-5 (SV1 3-1, submature male), 
b.c.: body chamber, с: callus. 



layed after the formation of the second whorl 
for the Palau specimens. This observation 
correlates well with the description of Saun- 
ders (1987, pp. 43-44). 

The scatter plot of B/H ratios of all captured 
animals exhibits wide ranging intra- and inter- 
populational variation of this parameter at 
least for premature and mature specimens 
(Fig. 5). At the same shell size (D = 1 50-1 60 
mm) most Fiji specimens have a more com- 
pressed shell than the Philippine specimens. 
The Palau specimens display remarkably 
wide variation in B/H ratio both in the imma- 
ture and mature stages, and the values of 



immature and submature specimens partly 
overlap those of mature specimens from Fiji 
and the Philippines. 

The ontogenetic pattern of chamber length 
( = distance between contiguous septa) in the 
early to middle stages is fairly alike among 
specimens of the three samples and the one 
Siquijor specimen (Fig. 6). 

Variation of Mature Shells 

As demonstrated by previous authors 
(Haven, 1977; Ward et al., 1977; Saunders & 
Spinosa, 1978; Ward & Martin, 1980; Ha- 
yasaka et al., 1982, 1987; Tanabe et al., 
1983; Tanabe & Tsukahara, 1987), species of 
living Nautilus show distinct sexual dimor- 
phism in the size and weight of animals and 
shell proportions at maturity. Namely, mature 
males are generally larger and possess 
broader shells than mature females. 

By examining the gonad development in 
live-caught animals, Tanabe & Tsukahara 
(1987) discussed the sexual dimorphism in N. 
pompilius from the Philippines (Tañon Strait) 
and Fiji. The difference in shell size at maturity 
among the Palau, Fiji and Philippine (Tañon) 
populations is made clear by summahzing the 
gonad and tissue weight data on live-caught 
animals (Tsukahara, 1985; Tanabe & Tsuka- 
hara, 1987) (Fig. 7). In each sample, abrupt 
increase of gonad weight initiates at the same 
shell size for both sexes. Full development of 
the gonad is well marked in the male speci- 
mens from Palau and Fiji, and this causes the 
relatively larger shell size in males than in fe- 
males at the same gonad index [= gonad 
weight/tissue weight (%)] (Fig. 7). 

Figure 7 also shows the difference in shell 
diameter at maturity among the three sam- 
ples. The average diameters of male and fe- 
male specimens in the Palau sample (ca. 210 
mm and 190 mm respectively) are much 
larger than those in the Fiji sample (ca. 150 
mm and 140 mm, respectively). Those in the 
Philippine (Tañon) sample (ca. 170 mm and 
160 mm; see also Tanabe et al., 1983, table 
3) are intermediate between the Fiji and 
Palau samples. Thus, the above differences 
in mature shell size among the three samples 
are much larger than that between sexes 
within the same sample. 

Recognition of maturity is also shown by 
such charactehstic shell features as approxi- 
mation of the final two or three septa, a thick- 
ened last septum, and blackened and thick- 
ened aperture (e.g. Stenzel, 1964; Collins & 



COMPARATIVE MORPHOLOGY OF NAUTILUS 



303 



1.0 



0.9 



I 
Ü 

Ш 

I 



0.8 



< 
Ш 
CC 

^0.7 



0.6 







■ PALAU (N=94) 






A PHILIPPINES (N = 34) 




à * 


• FIJI (N = 179) 




A A 


■ 


• . Ч ' ^ 


'^^ ' 


A 










• 


Л • . 


■ Ф 


•. • . i 


A 


, * 








, • t 


. ■ . ■ . 




• • - •* . 


■ ■■ S 




.•.•.-:.: . 


■ ■■ ■ 

■ ■■ '■■ 


, 


■ " •'.;>. > 


• ■ ■ . ■ ■ * 




*- •.•'.-■.;; 


. ■' ■ 






Í ■/.■ - . ■ - 






■ 

■ ■ ■ 

■ ■ ■ 

■ 
■ 

■ 



100 



150 200 

SHELL DIAMETER (mm) 



250 



FIG. 5. Scatter plot of shell breadth/height ratio (B hf) versus shell diameter for specimens of Nautilus 
belauensis from Palau and Nautilus pompilius from the Philippines (Tañon Strait) and Fiji. f\/!easurements of 
179 animals captured in 1986 (Tanabe, 1987, table 3) are used for the Fiji sample. 



Ward, 1987), because these features com- 
monly occur in specimens with a large gonad 
index. In accordance with these criteria, the 
two specimens from Bohol Strait near Siquijor 
Island (the Philippines) are regarded as ma- 
ture or submature shells. They are much 
smaller in shell diameter (ca. 130 mm; Fig. 
IG) than the mature specimens from Tañon 
Strait. Total number of septa at maturity ap- 
pears to be different among the three sam- 
ples (33-39, 32-35, and 28-32 septa in the 
Palau, Philippine (Tañon) and Fiji samples re- 
spectively) (Fig. 8). 



RADULAR AND JAW MORPHOLOGIES 

Radula 

The radula of Nautilus is secreted by colum- 
nar epithelial cells, named odontoblasts, in 
the posterior part of the radular sac, and is 



generated anteriorly in a series of rows 
(Tanabe & Fukuda, 1987). Each row consists 
of nine primary teeth (one central rachidian, 
and two pairs of laterals and marginals on 
each side) and two pairs of marginal support 
plates (Thiele, 1893; Vayssière, 1896; Griffin, 
1900; Naef, 1923; Solem & Richardson, 
1975; Lehmann, 1976; Mikami et al., 1980; 
Saunders, 1981a, 1987; Tanabe & Fukuda, 
1987). This arrangement is clearly distin- 
guished from that in modern coleoids, which 
in general have seven primary teeth and a 
pair of marginal plates (Solem & Richardson, 
1975). 

Morphological features of each radular ele- 
ment are essentially identical among the Phil- 
ippine (Tañon), Fiji and Palau Nautilus (Figs. 
9-10). Namely, the central rachidian tooth is 
triangular in shape, being more than two or 
three times as high as the two laterals (Fig. 9). 
The two marginal teeth are much longer than 
the central and laterals; they are gently 



304 



TANABE, TSUKAHARA & HAYASAKA 



20 



E 

E 

Ш 

Ü 
z 
< 

co 

Û 



Ш 
CO 



1 5 



10 



^A T2-1(male), Palau 

@B T5-3(female), Palau 

*C B-5(male), Tañon St., Philippinei 

* D B-3(f emale), Tanon St., Philippin 

■ E SQ-1(sex unknown), Bohol St., PI 
" F SV13-2(male), Suva, Fiji 




1 20 

CHAMBER NUMBER 



30 



40 



FIG. 6. Ontogenetic change of chamber length ( = septal interval) for selected mature specimens of Nautilus 
belauensis from Palau and Nautilus pompilius from the Philippines (Tañon and Bohol Straits) and Fiji. A. 
UMUT RM 18708-1, В. UMUT RM 18708-5, С. UMUT RM 18705-2, D. UMUT RM 18705-1, E. UMUT RM 
18706-1, F. UMUT RM 18707-6. 



curved and acutely projected anteriorly, with 
two strong grooves along their longitudinal 
axis (Fig. 10). In the anterior portion, the teeth 
are subcircular in cross section with a round 
tip, but they become rapidly broaden and 
compressed toward the base. A characteristic 
spatula-like antenor expansion is present at 
the base of the marginal teeth of every spec- 
imen from Palau and Fiji (Fig. 10A-C & E), 
but this feature is not so prominent in many 
specimens from the Philippines (Fig. 10D; 
see also Saunders, 1981a, figure 2). The 
marginal support plates are rectangular in 
outline; the inner one is larger than the outer. 



A marked depression is developed in the an- 
terior portion of the outer plate. 

The shape of each radular element is mark- 
edly variable even in the specimens from the 
same area, and the range of variation of the 
height'width ratio of the central tooth in the 
Palau sample apparently overlaps those in 
the Fiji and the Philippine samples (Fig. 11). 

Jaws 

The jaw apparatus of Nautilus differs from 
those of modern coleoids by the presence of 
conspicuous anterior calcareous coverings 



COMPARATIVE MORPHOLOGY OF NAUTILUS 



305 





8 




1 

■ Male 
D Female 


1 1 
PALAU 


1 I 






А 


1 


I 


1 I 1 1 

■ 






7 


- 




















- 








A Male 
Д Female 


PHILIPPINES 




¿ 


Л 


Л 




■ ■ 


























D 






6 


- 


• Male 
о Female 


FIJI 


• • 




А 










- 


X 

Ш 


5 


- 






• 
• 






А 


А 


А 


D 


- 


Q 










о • 












О П D _ 




z 


4 








о •• 








А 




D 




Q 










•• 












G 

■ 




< 










Оо • 
















Z 








• 


о •*. • 


• 














О 


3 


- 






•• 
• 














- 


о 


2 

1 



S. 


• t- ° 


о 
о 

8 

о 
о 


о 
о 

■ 
• 

1 1 




D 


1_ 


D 

1 


□ 

1 


lili 


- 



100 110 120 130 140 150 160 170 180 190 200 

SHELL DIAMETER (mm) 



210 220 230 



FIG. 7. Scatter plot of gonad index [gonad weight/tissue weight (%)] versus shell diameter for specimens of 
Nautilus belauensis from Palau and Nautilus pompilius from the Philippines (Tañon Strait) and Fiji. 



Fiji (N = 19) 

Philippines (N=14) 



Palau ( N = 8 ) 








27282930313233343536373839 
Number of septa 

FIG. 8. Variation in the total number of septa at 
maturity for Nautilus belauensis from Palau and 
Nautilus pompilius from the Philippines (Tañon 
Strait) and Fiji. 



on the chitinous plates of the upper ancJ lower 
jaws an(j by the shorter inner lamellae of the 
lower jaw (Okutani & MikamI, 1977; Saun(jers 
et al., 1978; Tanabe & Fukucda, 1987). Its 
overall morphology, composition and struc- 
tural relationship with the jaw muscles are the 
same among the species of Nautilus, and are 
well designed for a cutting and shearing func- 



tion (Saunders et al., 1978; Tanabe & 
Fukuda, 1987). 

The lower jaws of the Fiji and Palau spec- 
imens are both characterized by a distinct an- 
terior depression in the antero-dorsal margin 
of the outer lamella, followed by a rather 
straight shoulder (Fig. 12A-B & E-F). In con- 
trast, the lower jaws of the Philippine (Tañon) 
specimens mostly lack such a depression, 
and their outer lamella has gently concave 
antero-dorsal margin and roundly convex 
shoulder (Fig. 12C-D). 



DISCUSSION 
Taxonomic Relationships 

The present study shows that the Philippine 
(Tañon Strait), Fiji and Palau Nautilus popu- 
lations have strong affinities in overall shell 
morphology and radular and jaw structures. 
Furthermore, the large collections from the 
populations display similar ontogenetic pat- 
terns for the shell shape parameters and 
chamber length, and they can be distin- 



306 



TANABE, TSUKAHARA & HAYASAKA 




FIG. 9. Scanning electron micrographs of central rachidian and lateral (in part) radular teeth in Nautilus 
be/auens/s from Palau (A-B) and Nautilus pompiliusUom the Philippines (Tañon Strait) (C-D) and Fiji (E-F). 
A UMUT RM 18708-6 (T5-4, mature female). B. UMUT RM 18708-8 (T9-2, mature male), С UMUT RM 
18705-7 (B30; mature female), D. UMUT RM 18705-5 (B27; mature male), E. UMUT RM 18707-4 (SV12-3; 
immature female), F. UMUT RM 18707-9 (SV13-14; immature female). Scale bars indicate 200 цт. 



COMPARATIVE MORPHOLOGY OF NAUTILUS 




FIG. 10. Scanning electron micrographs of overall radula (A) and its marginal element (B-E) for specimens 
of Nautilus pompilius from Fiji (A-B) and the Philippines (Tañon Strait) (C-D), and Nautilus belauensis from 
Palau (E). A-B. UMUT RM 18707-4 (SV12-3), С. UMUT RM 18705-5 (B27), D. UMUT RM 18705-7 (B30), 
E. UMUT RM 18708-6 (T5-4). Scale bars indicate 500 ^JLm. Anatomy, c: central rachidian tooth, L, and Lg: 
inner and outer lateral tooth, M, and Mg: inner and outer marginal tooth, MP^ and MPg; inner and outer 
marginal support plates. 



guished mainly by the dimensions of adult an- 
imals, such as the total live weight, shell size, 
and total number of septa. These observa- 
tions may offer serious problems in recogniz- 
ing the Palauan population as a separate spe- 
cies. 
The Palauan Nautilus was identified by 



Dugdale & Faulkner (1976) as Nautilus sp. It 
was subsequently identified as N. pompilius 
(Faulkner, 1976; Saunders & Ward, 1979; 
Carlson, 1979) or N. cf. pompilius (Saunders 
et a!., 1978; Saunders & Spinosa, 1978, 
1979). Later, Saunders (1981a) proposed a 
new species, N. belauensis, on the basis of 



308 



TANABE, TSUKAHARA & HAYASAKA 



Г5 


" 










1.4 


■ 


PALAU (Í ) 




о H/W= 20 




a 




(* ) 




/ 




A 


PHILIPPINES (.Í) 




/ 


1.3 


'^ 




{*) 




/ 




• 



FIJ 1 






/ ■ 
/ ■ ' 


12 


' 








/ ■ H/W- 


1.1 


- 






/ 


D / 




- 




a 
a 

m / 


• 


/ ° 


xO.9 






/ A 


A 


/ 


Ш 0.8 

I 






1 • 
* • / • / 


/ 


A 

H/W = 10 


07 






•/ / 
/ ° / • 


• 
• 


/ 


0.6 


" 




/ * / 




/ 


05 


- 


/ 


/ * 


-^ 




0.4 
m 


/ 


У 


/ 


1 





0.1 0.2 0.3 04 05 06 07 08 

BASAL WIDTH (mm) 

FIG. 11. Scatter plot of central rachidian tooth 
height (H) and basal width (W) for specimens of 
Nautilus belauensis from Palau and Nautilus pom- 
pilius from the Philippines (Tañon Strait) and Fiji. 



examination of more than 1,000 live caught 
animals. According to Saunders (1981a, b, 
1987), N. belauensis is distinguished from N. 
pompilius by its larger mature size and wider 
central rachidian radular tooth, and by the 
presence of longitudinally crenulated growth 
lines on the shell. The present work, however, 
confirms the presence of crenulated shell or- 
namentation in many specimens of N. pom- 
pilius from Fiji and the Philippines. Further- 
more, the width/height ratios of radular 
elements are highly variable even in the spec- 
imens from the same area, suggesting that 
the shape of radular teeth appears to be of 
little significance at least for the species-level 
systematics of living Nautilus. The remaining 
diagnosis of the Palau population, unusually 
large mature shell size, stnould not be relied 
on for distinguishing species for the following 
reasons. Indeed, the widespread species, N. 
pompilius displays well-marked morphologi- 
cal differentiation regarding overall weight 
and size at maturity, proportion and coloration 



of shells, and the trends of the allometric re- 
lationships of several characters of the shells 
and soft tissues, not only among the geo- 
graphically separated populations (Ward et 
al., 1977; Hayasaka et al., 1982, 1987; Tan- 
abe & Tsukahara, 1987; Saunders, 1987; К. 
Tanabe's observations on specimens from 
various localities housed in the U.S. National 
Museum of Natural History), but also among 
neighbohng populations (Hayasaka et al., 
1982; Saunders, 1987; Swan & Saunders, 
1987; Habe & Okutani, 1988). The minor dif- 
ference in the lower jaw morphology between 
the Philippine and Fiji specimens can proba- 
bly be attributed to conspecific variation. 

In addition to the above results at morpho- 
logical level, recent examinations of large col- 
lections using electropheretic techniques pro- 
vided interesting data relevant to taxonomic 
relationships of Nautilus populations from a 
genetic viewpoint (Masuda & Shinomiya, 
1983; Woodruff et al., 1983, 1987). These 
works have made clear that Nautilus exhibits 
normal or slightly high levels of genetic vari- 
ation and that the isolated populations are 
well differentiated genetically. Relying upon 
Neis (1978) genetic distance coefficients, 
Woodruff et al. (1987) suggested that the 
Palau population (Л/. belauensis) and possibly 
the Fiji population (Л/. pompilius) are closely 
related to, but well differentiated at a species 
level from the populations of N. pompilius in 
the waters around New Guinea and Queen- 
sland. The genetic distance coefficients be- 
tween the samples of N. belauensis from 
Palau and N. pompilius from eight localities in 
the southwestern Pacific excluding the Philip- 
pines (< 0.2) are, however, much smaller 
than those between paired samples of N. 
scrobiculatus. N. macromphalus and N. pom- 
pilius (> 0.5) (see Woodruff et al., 1987, table 
IV & fig. 2). As Woodruff et al. (1987) docu- 
mented, there is no simple basis to translate a 
genetic distance into a taxonomic decision, 
because the processes of speciation are not 
closely coupled to the changes of structural 
genes. To sum up the above-mentioned mor- 
phological and genetic data, two different in- 
terpretations can be considered for the taxo- 
nomic relationship among the three 
populations. The one is that the populations in 
the Philippine, Fiji and Palauan waters are 
summarized into the amphimictically out- 
breeding species, N. pompilius, with high lev- 
els of genetic and morphological differentia- 
tion, and the other is that N. belauensis is a 
distinct species reproductively isolated from 



COMPARATIVE MORPHOLOGY OF NAUTILUS 



В 



309 




^^^ ÍW 




FIG. 12. Drawings of upper (right side) and lower (left side) jaws for specimens of Nautilus pompilius from 
Fiji (A-B) and tfie Ptiilippines (Tañon Strait) (C-D), and Nautilus belauensis from Palau (E-F) (lateral views). 
A. UMUT RM 18707-9 (SV13-14; mature female), B. UN/IUT RM 18707-10 (SV28-4-2; mature male), С 
UMUT RM 18705-7 (B30; mature female), D. UMUT RM 18705-2 (B5, mature male), E. UMUT RM 18708-5 
(T5-3; mature female), F. UMUT RM 18708-4 (T5-1; mature male). Scale bars indicate 1 cm. 



the populations of N. pompilius. In this paper, 
we refrain from choosing between the two be- 
cause of the insufficient data for the genetic 
variation of N. pompilius throughout its wide 
geographic range, especially of the popula- 
tions in the Philippine waters. 

Interpretation on Mature Shell Size Variation 

In his discussion of Nautilus systematlcs, 
Saunders (1987) suggested that the differ- 
ence in mature shell size between N. belauen- 
sis and N. pompilius does not result from the 
difference in the period of growth, on the basis 
of counting of septal number and the stage of 
the umbilical callus appearance. Although the 
absolute growth and longevity of Nautilus in 
their natural habitats are not fully understood, 
previous direct and indirect growth rate mea- 
surements by release-recapture experiments 
of tagged specimens, radiographic observa- 
tion of aquarium specimens, and radiometric 
dating of septa have shown that the period of 
chamber (septal) formation increases expo- 
nentially with increasing chamber number 
(Cochran et al., 1981 ; Saunders, 1983, 1984; 



Ward, 1985; Cochran & Landman, 1984; see 
compilation in Landman & Cochran, 1987, fig- 
ure 4, table V). The marked difference in the 
total number of septa among the mature spec- 
imens from Fiji, the Philippines and Palau can 
be, therefore, interpreted as reflecting the dif- 
ference in the pre-reproductive age among 
them. This interpretation is in accord with 
Landman & Cochran's (1987) age estimate 
from septal growth equations (10.9 y and 5.9 
y for N. belauensis and N. pompilius respec- 
tively). The Palau population may attain sexual 
maturity at slower rates than the Fiji and Phil- 
ippine populations, although its rate of septal 
formation in earlier stages may not differ 
greatly from those in the Philippine and Fiji 
populations. The rate of shell growth and the 
time required to attain sexual maturity may be 
controlled by both ecology (food supply, tem- 
perature, water depth etc.) and genetic fac- 
tors, and the degrees of dependence of these 
factors on growth apparently differ among in- 
dividual populations. Based on the data from 
genetic analysis. Woodruff et al. (1987) sug- 
gested that the Palau and Fiji populations have 
distinctly diverged from the ancestral form of 



310 



TANABE, TSUKAHARA & HAYASAKA 



N. pompilius by peripheral isolation for about 
1 million years. We have no available data on 
the fossil record of Nautilus to verify this hy- 
pothesis, but if it is correct, the adult size in- 
crease or decrease in relation to the length of 
pre-reproductive age in the history of N. pom- 
pilius stock can be expressed by hypermor- 
phosis and progenesis in terms of McNama- 
ra's (1986) definition of heterochrony. 

Conclusion 

The Nautilus populations in Palau, the Phil- 
ippines and Fiji are essentially similar in over- 
all shell morphology, ontogenetic shell varia- 
tion, and jaw and radular structures. They are 
distinguished mainly by the dimensions of 
adult animals. From these morphological ev- 
idence and the available genetic data, the 
Palau and the other two populations are re- 
garded as either summarizing into the wide- 
spread species, Nautilus pompilius. or be- 
longing to the closely related sibling species, 
N. belauensis and N. pompilius respectively. 
The size difference among the adult animals 
from the three populations probably results 
from the difference in their pre-reproductive 
ages. 

ACKNOWLEDGMENTS 

We acknowledge Yoshiko Kakinuma for her 
facilities and encouragement, both in the field 
and laboratory. Our thanks to Angel С Alcalá, 
Uday Raj, and David K. Idip for providing fa- 
cilities to operate our field research, other 
members of the project in the Philippines, Fiji 
and Palau for their assistance in collecting 
live Nautilus and for helpful discussions, and 
Clyde F. Roper for allowing one of us (K. T.) to 
observe the collections of Nautilus at the U. S. 
National Museum of Natural History in his 
care. W. Bruce Saunders and Neil H. Land- 
man read the manuscript critically and gave 
useful comments for improvement of this pa- 
per. Supported by the Scientific Research 
Fund from the Japanese Ministry of Educa- 
tion, Science and Culture (Mombusho) (No. 
57043059 for 1982; No. 58041055 for 1983; 
No. 59043050 for 1984; No. 62043064 for 
1986; No. 63041 103 for 1987; No. 63540622 
for 1988-89; No. 01460062 for 1989). 

LITERATURE CITED 

ABBOTT, R. T. & S. P. DANCE, 1983, Compen- 
dium of seashells. E. P. Dutton, New York, 411 
PP- 



CARLSON, B. A., 1979, Chambered nautilus: A 
new challenge for aquarists. Freshwater & l\Aa- 
rine Aquarium. 2: 48-51 , 63. 

COCHRAN, J. K. & N. H. LANDMAN, 1984, Radi- 
ometric determination of the growth rate of Nau- 
tilus in nature. Nature (London). 308: 725-727. 

COCHRAN, J. K., D. M. RYE & N. H. LANDMAN, 
1981, Growth rate and habitat of Nautilus pom- 
pilius inferred from radioactive and stable isotope 
studies. Paleobiology. 7: 469-480. 

COLLINS, D. H. & P. D. WARD, 1987, Adolescent 
growth and maturity in Nautilus. In: SAUNDERS, 
W. B. & N. H, LANDMAN, eds., Nautilus. The 
biology and paleobiology of a living fossil, pp. 
421-432. Plenum Press, New York. 

DOUGDALE, H. K. & D. FAULKNER, 1976, The 
chambered Nautilus, exquisite living fossil. Na- 
tional Geographic (Magazine. 149:38-41. 

FAULKNER, D., 1976. Dwellers in the sea. Read- 
er's Digest Press, New York, 194 pp. 

GRIFFIN, L. E., 1900, The anatomy of Nautilus 
pompilius. Memoirs of the National Academy of 
Sciences. 8: 101-230. 

HABE, T., 1980, Description of living Nautilus. In: 
HAMADA, T., I. OBATA & T. OKUTANI, eds., 
Nautilus macromphalus in captivity, pp. 1-3. To- 
kai University Press, Tokyo. 

HABE, T. & T. OKUTANI, 1988, A new subspecies 
of living Nautilus (Cephalopoda: Nautiloidea) 
from the Sulu Sea. Venus. 47: 91-94. 

HAVEN, N., 1977, The reproductive biology of Nau- 
tilus pompilius in the Philippines. l\Aarine Biology, 
42:177-184. 

HAYASAKA, S., K. OKI, K. TANABE, T. SAISHO & 
A. SHINOMIYA, 1987, On the habitat of Nautilus 
pompilius in Tañon Strait (Philippines) and the Fiji 
Islands. In: SAUNDERS, W. B. & N. H. LAND- 
MAN, eds.. Nautilus. The biology and paleobiol- 
ogy of a living fossil, pp. 179-200. Plenum Press, 
New York. 

HAYASAKA, S., T. SAISHO, Y. KAKINUMA, A. 
SHINOMIYA, K. OKI, T. HAMADA, K. TANABE, 
Y. KANIE, M. HATTORI, F. V. VUSSE, L. AL- 
CALA, P С CORDERO, JR., J. J. CABRERA & 
R. G. GARCIA, 1982, Field study on the habitat 
of Nautilus in the environs of Cebu and Negros 
Islands, the Philippines. Memoirs of the Ka- 
goshima University Research Center for the 
South Pacific. 3: 67-137. 

LANDMAN, N. H. & J. K. COCHRAN, 1987, Growth 
and longevity of Nautilus. In: SAUNDERS W. B. 
& N. H. LANDMAN, eds., Nautilus. The biology 
and paleobiology of a living fossil, pp. 401-420. 
Plenum Press, New York. 

LEHMANN, и., 1976, Ammoniten. Ferdinand Enke 
Verlag, Stuttgart. 171 pp. 

MASUDA, Y. & A. SHINOMIYA, 1983, Genetic van- 
ation in Nautilus pompilius. Kagoshima Univer- 
sity Research Center for the South Pacific, Oc- 
casional Papers, 1 : 22-25. 

McNAMARA, K. J., 1986, A guide to the nomencla- 
ture of heterochrony. Journal of Paleontology. 
60: 4-13. 



COMPARATIVE MORPHOLOGY OF NAUTILUS 



311 



MIKAMI. S., T. OKUTANI, H. HIRANO, Y. KANIE & 
T. HAMADA, 1980, Behavior in captivity. In: MA- 
MADA, T., I. OBATA & T. OKUTANI, eds., Nau- 
tilus macromphalus in captivity, pp. 1 1-22. Tokai 
University Press, Tokyo. 

NAEF, A. 1923, Die Cephalopoden (Systematik). 
Fauna und Flora des Golfes von Naepel und der 
Angrenzenden Meeres-Abschnitte. Monographie 
35; 863 pp. 

NEI, M., 1978, Estimation of average heterzygosity 
and genetic distance from a small number of in- 
dividuals. Genetics. 89: 583-590. 

OKUTANI, T & S. MIKAMI, 1977, Descriptions of 
beaks on Nautilus macromphalus Sowerby. Ve- 
nus. 36: 115-121. 

SAUNDERS, W. В., 1981a, A new species of Nau- 
tilus Uom Palau. The Veliger. 24: 1-7. 

SAUNDERS, W. В., 1981b, The species of living 
Nautilus and their distribution. The Veliger, 24: 
8-17, 

SAUNDERS, W. В., 1983, Natural rates of growth 
and longevity of Nautilus belauensis. Paleobiol- 
ogy. 9: 280-288. 

SAUNDERS, W. В., 1984, Nautilus belauensis 
growth and longevity: Evidence from marked and 
recaptured animals. Science, 224: 990-992. 

SAUNDERS, W. В., 1987, The species of Nautilus. 
In: SAUNDERS, W. B. & N. H. LANDMAN, eds.. 
Nautilus. The biology and paleobiology of a living 
fossil, pp. 35-52. Plenum Press, New York. 

SAUNDERS, W. B. & L. E. DAVIS, 1985, A prelim- 
inary report on Nautilus in Papua New Guinea. 
Science In New Guinea, 1 1 : 60-69. 

SAUNDERS, W. B. & С SPINOSA, 1978, Sexual 
dimorphism in NautilusUovn Palau. Paleobiology, 
4: 349-358. 

SAUNDERS, W. B. & С SPINOSA, 1979, Nautilus 
movement and distribution in Palau, Western 
Caroline Islands. Science. 204: 1199-1201. 

SAUNDERS, W. В., С. SPINOSA, С. TEICHERT & 
R. С BANKS, 1978, The jaw apparatus of recent 
Nautilus and its palaeontological implications. 
Palaeontology, 21: 129-144, 

SAUNDERS, W. B. & P. D. WARD, 1979, Nautiloid 
growth and lunar dynamics. Lethaia, 12: 172 

SOLEM, A. & E. S. RICHARDSON, 1975, Paleo- 
cadmus, a nautiloid cephalopod radula from the 
Pennsylvanian Francis Creek Shale of Illinois. 
The Veliger, 17: 233-242. 

STENZEL, H, В., 1964, Living Nautilus. In: 
MOORE, R. C, ed.. Treatise on invertebrate pa- 
leontology. Part К (Mollusca 3). pp. K59-K93. 
Geological Society of America & University of 
Kansas Press, Lawrence. 

SWAN, A. R. H. & W. B. SAUNDERS, 1987, Mor- 
phological variation in Nautilus from Papua New 
Guinea. In: SAUNDERS, W. B. & N. H. LAND- 
MAN, eds.. Nautilus. The biology and paleobiol- 
ogy of a living fossil, pp. 85-103. Plenum Press, 
New York. 

TANABE, K., 1985, Record of trapping experiment. 
In: HAYASAKA, S., ed.. Marine ecological stud- 
ies on the habitat of Nautilus pompilius in the 



environs of Viti Lebu, Fiji. Kagoshima University 
Research Center for the South Pacific. Occa- 
sional Papers. 4: 10-17. 

TANABE, K., 1988, Record of trapping experiment. 
In: HAYASAKA, S., ed.. Marine ecological stud- 
ies on the habitat of Nautilus pompilius in Fiji 
(The second operation). Kagoshima University 
Research Center for the South Pacific. Occa- 
sional Papers. 15:5-14. 

TANABE, K. & Y. FUKUDA, 1987, Mouth part mor- 
phology and histology. In: SAUNDERS, W. B. & 
N. H. LANDMAN, eds.. Nautilus. The biology and 
paleobiology of a living fossil, pp. 313-322. Ple- 
num Press, New York. 

TANABE, K., S. HAYASAKA, T. SAISHO, A. SHI- 
NOMIYA & K. AOKI, 1983, Morphologic vanation 
of Nautilus pompilius Uom the Philippines and Fiji 
Islands. Kagoshima University Research Center 
for the South Pacific, Occasional Papers. 1:9- 
21. 

TANABE, K., S. HAYASAKA, & J. TSUKAHARA, 
1985, Morphologic analysis of Nautilus pompil- 
ius. Kagoshima University Research Center for 
the South Pacific. Occasional Papers. 4: 38-49. 

TANABE, K. & J. TSUKAHARA, 1987, Biometrie 
analysis of Nautilus pompilius from the Philip- 
pines and the Fiji Islands. In: SAUNDERS, W. B. 
& N. H. LANDMAN, eds.. Nautilus. The biology 
and paleobiology of a living fossil, pp. 105-1 13. 
Plenum Press, New York. 

TANABE, K. & J. TSUKAHARA, 1 989, Notes on the 
trapped Nautilus from Palau. Kagoshima Univer- 
sity Research Center for the South Pacific, Oc- 
casional Papers, (in press). 

THIELE, J., 1983, Beiträge zur Kenntnis der Mol- 
luskaen. II. Über die Molluskenschale. Zeitschrift 
für Wissenschaüliche Zoologie, Abteilung A. 55: 
220-251. 

TSUKAHARA, J., 1985, Histological and his- 
tochemical studies of gonads of Nautilus pompil- 
ius from Fiji. Kagoshima University Research 
Center for the South Pacific. Occasional Papers, 
4: 50-60. 

VAYSSIÈRE, A., 1896, Étude sur l'organisation du 
Nautile (caractères zoologiques, dimorphisme 
sexuel, tentacules et spadice). Annales des Sci- 
ences Naturelles. Zoologie et Paléontologie. 8: 
137-186. 

WARD, P., 1985, Periodicity of chamber formation 
in chambered cephalopods: Evidence from Nau- 
tilus macromphalus and Nautilus pompilius. Pa- 
leobiology, 1 1 : 438-450. 

WARD, P. D. & A. W. MARTIN, 1980, Depth distn- 
bution of Nautilus pompilius in Fiji and Nautilus 
macromphalus in New Caledonia. The Veliger, 
22: 259-264. 

WARD, P. D., R. STONE, G. WESTERMANN & A. 
MARTIN, 1977, Notes on animal weight, cameral 
fluids, swimming speed, and color polymorphism 
of the cephalopod Nautilus pompilius in the Fiji 
Islands. Paleobiology. 3: 377-388. 

WOODRUFF, D. S., M. MULVEY, W. B. SAUN- 
DERS & M. P. CARPENTER, 1983, Genetic van- 



3-12 TANABE, TSUKAHARA & HAYASAKA 

ation ,n the cephalopod Nautilus belauensis. Pro- W. B. & N. H. Landman, eds., Nautilus. Thebi- 

TedngTofîheT^^^^ of Natural Sciences ,n ology and paleobiology of a living fossil, pp. 65- 

Phf^elphia. 1 35: 1 47-1 53. 83. Plenum Press, New York. 
WOODRUFF, D. S., M. P. CARPENTER, W. B. 



MALACOLOGIA, 1990,31(2): 313-326 

CROSS SECTIONAL MORPHOLOGY OF THE GLADIUS IN THE FAMILY 

OMMASTREPHIDAE (CEPHALOPODA: TEUTHOIDEA) AND ITS BEARING ON 

INTRAFAMILIAL SYSTEMATICS 

Ronald B. Toll 
Department of Biology, The University oi the South, Sewanee, Tennessee, U.S.A. 37375 

ABSTRACT 

The cross sectional morphology of the ommastrephid gladius is compared among 15 species 
in 1 1 genera of the three currently recognized subfamilies. The three axial complexes of the free 
rachis are shown to comprise a suite of characters of systematic importance. Intrafamilial rela- 
tionships derived from characters of the gladius generally conform to the traditional classification 
of the family based on a synthesis of traditional characters, with the exception of subfamilial 
organization. The depositional layering of chitin which occurs as part of the accretive growth of 
the gladius is easily seen in cross section using either light or scanning electron microscopy. 
Examination of these layers can provide information on ontogenetic changes in gladius con- 
struction because the early morphology is covered but apparently not altered during subsequent 
depositional events. 

Key words: Teuthoidea, Oegopsida, Ommastrephidae, gladius morphology, phylogeny, shell, 
squid. 



INTRODUCTION 

Squids of the family Ommastrephidae are 
robust, muscular, powerful swimmers. Adults 
range in size from about 8.0 cm {Hyaloteuthis 
pelágica) to over 1 .0 m (Dosidicus gigas) in 
mantle length (Wormuth, 1976; Nesis, 1983). 
The majonty of taxa are open ocean epipe- 
lagic animals while some (e.g. ///ex and To- 
darodes) range over continental shelves. The 
ommastrephids are predaceous carnivores 
that feed primarily on finfishes and other 
squids. They are prey to many predator spe- 
cies including marine mammals and finfishes. 
Many ommastrephids are commercially ex- 
ploited for both human consumption and for 
bait in finfish fisheries and as such form the 
basis of substantial fisheries in many areas of 
the world (Clarke, 1966; Roper, 1983; Rath- 
jen & Voss, 1987). 

Roper (1983) included the Ommastre- 
phidae as one of the four families of cephalo- 
pods (along with the Sepiidae, Octopodidae, 
and Loliginidae) most critically in need of 
comprehensive systematic revision based on 
four criteria. The groups are: (1 ) speciose and 
occur in greatest abundance in shallow wa- 
ters; (2) support major fisheries; (3) support 
biomedical, ecological and other biological re- 
search; (4) poorly known systematically. Prior 
to and as a result of Roper's (1983) listing of 



these four groups, the systematics of the Om- 
mastrephidae represents an area of consid- 
erable recent research (e.g. Roper, et al., 
1969; Adam, 1975; Zuev et al., 1975; Wor- 
muth, 1976; Nesis, 1978, 1983; Lu & Dun- 
ning, 1982; Roeleveld, 1982, 1988; O'Dor, 
1 983; Okutani, 1 983). A variety of morpholog- 
ical and meristic characters have been used 
in the systematic study of ommastrephids. 
Traditionally, the three subfamilies, Ommas- 
trephinae, Todarodinae, and lllicinae, have 
been separated on characters associated 
with the funnel groove, specifically the occur- 
rence of side pockets and foveolae. Below the 
subfamilial level, characters used to delineate 
taxa include fin angle, club sucker arrange- 
ment and dentition, spermatophore morphol- 
ogy, arm protective membrane development, 
condition of the funnel-mantle locking carti- 
lages, type and distribution of light organs, 
hectocotylus morphology, and various mor- 
phometric relationships of the arms, tenta- 
cles, clubs, fins, etc. Details of the gladius or 
pen have been absent from the descnptions 
and systematic analyses of most of the recent 
systematic contributions to the Ommas- 
trephidae. In contrast, in many of the older 
contnbutions to ommastrephid systematics 
(e.g. Pfeffer, 1912; Sasaki, 1929) the gladius 
is described, sometimes in great detail, and 
often illustrated. Collectively, these reports, in 



313 



314 



TOLL 



particular Pfeffers (1912) monumental mono- 
grapfi of the Oegopsida, wfiich contains de- 
tailed illustrations of the gladius of many 
ommastrephid taxa, suggest that the om- 
mastrephid gladius is a highly conservative 
structure, exhibiting little morphologic vana- 
tion across the family, with the exception of 
relatively minor differences in the width of the 
free rachis and the length of the cone field. An 
exception to this perceived homogeneity 
among ommastrephid gladii is the unique, 
layered deposit of chitin within the concavity 
of the cone field of Dosidicus gigas (Steen- 
strup, 1857; Pfeffer, 1912; Toll, 1982). The 
existence of the widely held assumption that 
the ommastrephid gladius is of little value to 
systematic study is probably the cause of the 
relative lack of interest in this structure as re- 
flected by more recent systematic contribu- 
tions to this group. Indeed, the descnptive ac- 
counts of two recently described species lack 
any mention of the gladius whatsoever {Orni- 
thoteuthis antillarum Aäam, 1957; Todaropsis 
filippovae Aóam, 1975). 

The examination of cross sections of the 
ommastrephid gladius is not new. Lesueur 
(1821), Ball (1841), Posselt (1890), Vernll 
(1882), Naef (1923), Sasaki (1929), and Ran- 
curel (1970) included variously detailed men- 
tion of the cross sectional shapes of the gla- 
dius or described the characteristics of the 
axes along the free rachis allowing inferences 
of cross sections to be made. Indeed, Naef 
(1923) used aspects of the cross sectional 
structure of the gladius as part of species di- 
agnoses. Toll (1982) demonstrated that cross 
sections of ommastrephid gladii contained 
hitherto unknown systematic characters that 
could be useful in phylogenetic reconstruc- 
tions. The same characters could be valuable 
in identifying fragmentary ommastrephid re- 
mains encountered in stomach contents of 
predators. This paper presents the results of 
a comparative study of ommastrephid gladius 
cross sections. The results show that while 
the gross shape of the ommastrephid gladius 
is similar throughout the family, there is con- 
sistent variation in cross sectional shape. This 
vanation should be assessed as part of future 
studies regarding ommastrephid systematics. 
Finally, overall shape and structure of the gla- 
dius are two of the few anatomical characters 
of squids that allow direct comparison be- 
tween fossil and Recent teuthoids and could 
prove useful in establishing phylogenetic re- 
lationships to fossil ommastrephid antece- 
dents (see Donovan & Toll, 1988). 



MATERIALS AND METHODS 

Gladii were dissected out of preserved 
specimens by means of a longitudinal incision 
along the ventral midline. The cut edges of 
the mantle wall were reflected back to expose 
the viscera. The left gill was severed from its 
attachment to the inner surface of the mantle 
musculature. Beginning anteriorly and pro- 
gressing posteriorly, the visceral mass was 
freed from the mantle wall along its left side 
and reflected to the right exposing the glad- 
ius, still in the shell sac, below it. Once com- 
pletely exposed, the shell sac was cut open 
ventrally and laterally to allow the gladius to 
be removed from the inner surface of the dor- 
sal mantle musculature. As necessary, the 
nuchal muscles, which extend from the 
nuchal cartilage to the ventral surface of the 
anterior free rachis, were severed from their 
insertion on the shell sac. The narrowest part 
of the gladius, that area at the posterior limit 
of the free rachis, is sometimes completely 
buried in the mantle musculature and further 
dissection is required to free it. Also, in some 
taxa, the musculature of the tail region must 
be opened in order to free the apex of the 
conus. The procedure for extraction of the 
gladius descnbed here is preferable to exci- 
sion of the gladius via dissection through the 
dorsal mantle musculature. When carefully 
executed, the ventral removal method results 
in little substantiative damage to the speci- 
men. Excised gladii were kept in either 40% 
isopropyl or 70% ethyl alcohol and stored in 
separate vials or bottles along with the spec- 
imen from which it was removed. 

Because this study represented the first ex- 
tensive, comparative examination of cross 
sections of ommastrephid gladii, a convention 
needed to be established regarding the 
choice and standardization of levels of sec- 
tion to be examined. Four cross sections were 
selected and named as follows: level A, level 
B, level С and level D. The anterior three 
levels (A, B, and C) were established at dis- 
crete proportional distances from the anterior 
tip of the gladius. These are 0.10 GL (one- 
tenth of the length of the gladius measured 
from its anterior tip), 0.25 GL (one-quarter of 
the length of the gladius measured from its 
anterior tip), and 0.60 GL (six-tenths of the 
length of the gladius measured from its ante- 
rior tip), respectively. Level D coincides with 
the posterior limit of the free rachis where it 
meets the cone field. Cross sections were 
made using a new single edge razor blade 



OMMASTREPHID GLADIUS MORPHOLOGY 



315 



with the gladius held firmly on a hard rubber 
block. The cross sections of different taxa are 
drawn approximately to the same size to fa- 
cilitate direct compansons. All sections are 
oriented with the dorsal surface toward the 
top of the page. In each set, level A is at the 
top and levels B, C, and D are in ordered 
sequence below. Each set of cross sections is 
based on near mature or fully mature individ- 
uals and represents a composite, typical for 
that species. Variation is discussed along with 
the treatment of individual taxa and in the 
General Discussion. 

Abbreviations used in the text are as fol- 
lows: M, male; F, female, ML; dorsal mantle 
length; GL, gladius length; ANSP, Academy 
of Natural Sciences of Philadelphia; BCF, Bu- 
reau of Commercial Fisheries (now National 
Marine Fisheries Service); lATTC, Inter- 
American Tropical Tuna Commission; MCZ, 
Museum of Comparative Zoology, Harvard 
University; UMML, Invertebrate Museum, 
Rosenstiel School of Marine and Atmospheric 
Science, University of Miami; USNM, National 
Museum of Natural History, Smithsonian In- 
stitution; DISC, R/V DISCOVERER; ELT, 
USNS ELTANIN; ORE, M/V OREGON; P, R/ 
V JOHN ELLIOTT PILLSBURY; TC, R/V 
TOWNSEND CROMWELL; ET, Engel Trawl; 
IKMT, Issacs-Kidd midwater trawl; MWT, mid- 
water trawl; ОТ, otter trawl. 



MORPHOLOGY AND ANATOMICAL 

RELATIONSHIPS OF THE 

OMMASTREPHID GLADIUS 

The free rachis (Fig. 1) is long, broadest 
anteriorly, tapered posteriorly and terminates 
anteriorly in a stiff point. Anteriorly, there are 
three axial complexes, one medial and two 
lateral, each with three primary components; 
a ventrally displaced axis, a dorsal plate, and 
a commissure that joins the axis to the plate 
(Fig. 2). The plates and commissures can 
vary in thickness and width. Laterally, the 
plates are rounded or tapered to a point and 
ventrally recurved. The two lateral plates are 
connected to the central one by a pair of 
broad, thin, lateral fields. The lateral axes can 
be bifurcated anteriorly. The lateral axis, 
plate, and commissure progressively coa- 
lesce posteriorly to level С where the lateral 
axial complex vanes from lobate to hook- 
shaped with an admedial cleft. The three axial 
complexes (hereafter referred to as the me- 
dial complex and lateral complexes) converge 



posteriorly to form a single complex of vari- 
able shape at level D. This single axis ex- 
tends to the posterior tip of the gladius. The 
vanes are reduced to a small, spindle-shaped 
cone field that accounts for 1 0% to 25% of the 
total GL. Fine, radiating striae converge ante- 
horly from the anterolateral portions of the 
cone field toward the rachis. There is a small, 
conical primary conus. 

The gladius is partially embedded in the 
ventral surface of the dorsal mantle muscula- 
ture along the dorsal midline. In some taxa, the 
anterolateral edges of the free rachis are over- 
lain by muscle. In many, the narrow posterior 
portion of the free rachis, including part or all 
of the cone field, is completely buried within 
the mantle musculature. The nuchal muscles 
insert on the shell sac covering the ventral 
surface of the medial rachis fields posterior to 
the widest part of the free rachis. The insertion 
sites are oval. The gladius does not invade the 
posterior tail-like extension of the mantle as 
found in Ornithoteuthis, among others. 



DESCRIPTIVE ACCOUNTS 

Subfamily Ommastrephinae Steenstrup, 1857 

Genus Ommasirep/7es d'Orbigny, 1835 

Ommastrephes bartramii (Lesueur, 1821) 

Material examined.^M, ML = 283-205 
mm, GL = 287-108 mm, R/V VELERO, no 
data (probably eastern Pacific Ocean), 
UMML 31 .1 770. —1 F, ML = 241 mm, GL = 
237 mm, Naples, Italy, ANSP A6474. — 
IF, ML = 142 mm, GL = 149 mm, R/V AT- 
LANTIS, off Bermuda, surface night light, 1 1 
Oct. 1937, MCZ 293702. 

Description. — Cross sections (Fig. 3): 
Level A — The medial axis is subellipsoid, 
wider than deep, and broadly attached to the 
thin, medial plate. The lateral plates are mod- 
erately thick, distally tapered to blunted points 
or rounded tips and ventrally recurved. There 
are paired lateral axes. The proximal axis is 
digitiform and curved in some specimens. 
The distal axis is subovoid and wider than 
deep. Both axes are broadly attached to the 
lateral plate; Level В — The medial complex is 
similar in size and shape to that of level A. 
The lateral complexes are irregularly multi- 
lobed; Level С — The medial complex is small 
and fusiform. The lateral complexes are an 
inflated hook-shape. The admedial cleft is 
quadrangular to triangular and about one- 
third to one-half as wide as the complex. 



316 



TOLL 




Level A- 



Level В 



^ ^^ 



^f" 




Medial Axial Complex 
Lateral Axial Complex 



- Level С 



Level D 







Conus 



FIG. 1. Diagrammatic ventral view of ommastrephid gladius. 

FIG. 2. Composite cross sections of gladius with level of sections corresponding to Fig. 1 [a-lateral field; 
b-medial plate; c-medial commissure; d-medial axis; e-lateral plate; f-lateral axis (single); f'-proximal lateral 
axis; f "-distal lateral axis; g-lateral commissure; h-accessory process; i-admedial cleft; j-ventromedial pro- 
cess; k-lateral process; l-dorsal carina; m-body; n-ventral keel]. All sections oriented with dorsal surface 
toward top of page. 



There is a small ventromedial process; Level 
D — The body is subthangular with a pair of 
large quadrangular, ventral keels and stout, 
dorsally upswept, lateral processes that taper 
to blunt points. The dorsal carina is stout and 
slightly inflated apically. 

Discussion. — The gladii show variation in 
the shape of the lateral thickenings at all lev- 
els and the overall shape at Level D. This 
variation was seen as varying degrees of 
thickness and shape of the axis and plate 
components and is greater than in any other 
ommastrephid species examined. The six 
Ommastrephes bartramii examined here 
were from distant localities — the Mediterra- 
nean Sea, western Atlantic Ocean and Pacific 



Ocean. Prevailing taxonomic uncertainties as 
well as geographic variation probably account 
for at least part of this variation. 

Genus Sthenoteuthis Verrill, 1880 
Sthenoteuthis pteropus (Steenstrup, 1855) 

Material examined. — IF, ML = 328 mm, 
GL = 351mm, DISC 425E, Atex Drift, 
12' 27'N, 41 '21 'W, dip net with night light, 12- 
13 Feb. 1969, UMML — 2F, ML = 288 mm, 
GL = 304-286 mm, DISC 425E, Atex Drift, 
10"51 'N, 42' 21 'W, dip net with night light, 14- 
15 Feb. 1969, UMML. — 3F, ML = 305—178 
mm, GL - 317—175 mm, DISC 425E, Atex 
Drift, 12 57'N, 40 09'W, dip net with night 
light, 9-10 Feb. 1969, UMML. — 2F, ML = 



OMMASTREPHID GLADIUS MORPHOLOGY 



317 










FIGS. 3-9. Stylized composite cross sections of gladius levels A-D (see text for explanations of arrows): Fig. 
3. Ommastrephes bartramii; Fig. 4. Sthenoteuthis pteropus; Fig. 5. S. oualaniensis; Fig. 6. Dosidicus gigas; 
Fig. 7. Eucleoteuthis luminosa; Fig. 8. Ornithoteuthis antillarum; Fig. 9. Hyaloteuthis pelágica. 



318 



TOLL 







13 



./^\. 



é 






17 



tÁé 



FIGS. 10-17. Stylized composite cross sections of gladius levels A-D (see text for explanations of arrows): 
Fig. 10. ///ex coindetti: Fig. 11./, oxygonius: Fig. 12. Todaropsis eblanae: Fig. 13. Todarodes sagittatus; Fig. 
14. T. pacificus: Fig. 15. Nototodarus sloani: Fig. 16. N. hawaiiensis: Fig. 17. Martialia hyadesi. 



OMMASTREPHID GLADIUS MORPHOLOGY 



319 



265—134 mm, GL = 256-132 mm, DISC 
425E, Atex Drift, 12°45'N, 40^38' W, dip net 
with night light, 10-11 Feb. 1969, UMML. — 
IF, ML = 211 mm. GL = 207 mm, DISC 
425E, Atex Dnft, 13°43'N, 38°58'W, dip net 
with night light, 6-7 Feb. 1969, UMML. 

Description. — Cross sections (Fig. 4); 
Level A — The medial axis is subellipsoid and 
wider than deep. The medial commissure is 
only slightly constricted. The lateral plates are 
broad, thin, and distally tapered to narrow 
points and ventrally recurved. The two paired 
lateral axes are spindly. The distal one is bifid. 
The proximal lateral axis is digitiform, curved 
medially and swollen apically in some speci- 
mens. The lateral commissures are con- 
stricted; Level В — The medial complex is 
similar in shape and slightly smaller than that 
of level A. The proximal lateral axis is broader 
than that of level A. The distal lateral axis is 
subquadrangular, wider than deep, broadly 
joined to the lateral plate, and slightly incised 
ventrally in some specimens. The accessory 
process is most commonly "U"-shaped, con- 
cave dorsally, but is lobate in one specimen; 
Level С — The medial complex is spindle- 
shaped. The lateral complexes are hook- 
shaped. The deep, medially facing subtrian- 
gular to sickle-shaped admedial cleft is about 
two-thirds to three-quarters of the width of the 
complex. There is a small ventromedial pro- 
cess; Level D — The body is subtriangular, 
broadest ventrally, with a pair of subquadran- 
gular to lobate, ventral keels and a pair of 
dorsally curved lateral processes that taper to 
acute points. The dorsal carina is inflated api- 
cally. 

Sthenoteuthis oualaniensis (Lesson, 1830) 

Material examined. — IF, ML = 317 mm, 
GL = 341 mm, MOE Cr. 4В, 273В, 20°50'N, 
59^1 0'E, 4 Dec. 1963, night light and hand- 
line, UMML 31.1812. —IF, ML = 170 mm, 
GL = 178 mm, 10°N, 92°30'E, night light and 
dipnet, Nov. -Dec. 1961, USNM 656967. 
— 1F, ML = 138 mm, GL = 143mm, Moorea 
Island, Society Islands, 15 Apr. 1937, ANSP 
A6364.— IM, ML = 123 mm, GL = 128 mm, 
Moorea Island, Society Islands, 15 Apr. 1937, 
ANSP A6357. —IF, ML = 115 mm, GL = 
121 mm, Moorea Island, Society Islands, 15 
Apr. 1937, ANSP A6347. — Isex?, ML = 42 
mm, GL = 45 mm, SHOYO MARU Sta. 12, 
23°25.5'S, 104°36.8'W, 21 Jan. 1963, in 
stomach of Alepisaurus, UMML 31.1360. 



Description. — Cross sections (Fig. 5): 
Level A — The medial axis is ellipsoid, wider 
than deep, and attached to the medial plate 
by a commissure that is about one-half to two- 
thirds of the axis width. The lateral plates are 
broad, thin, tapered distally to a narrow point 
and ventrally recurved. There are paired lat- 
eral axes. The proximal lateral axis is digiti- 
form and slightly curved medially. The distal 
one is narrow, deep, and slightly constricted 
basally. In two specimens, the proximal lat- 
eral axes are absent. In another specimen, 
the distal lateral axes are slightly bifurcated 
ventrally. There is a small, domelike protuber- 
ance on the ventral surface of the lateral 
plates distal to the distal lateral axis (arrow); 
Level В — The medial axis is similar in shape 
and size to that in level A but is more broadly 
attached to the medial plate. The lateral plate 
and distal lateral axis are fused into a single, 
quadrangular axis. A digitiform proximal axis 
is present in those specimens with a proximal 
lateral axis in level A. The lateral process is 
bulbous and connected to the distal surface of 
the lateral complex by a stalklike commissure 
(arrow); Level С — The medial complex is 
fusiform. The lateral complexes are hook- 
shaped. The deep, broadly excavated adme- 
dial cleft if subtriangular to irregularly polygo- 
nal and equal in depth to about one-half of the 
width of the complex. A small, ventromedial 
process is present; Level D — The body is 
roughly triangular, broadest ventrally, with a 
pair of stout, angular, ventral keels and a sin- 
gle dorsal carina that is slightly expanded api- 
cally. The stout lateral processes are dorsally 
curved and taper to rounded tips. 

ñemar/cs.— The absence of proximal lateral 
axes in two specimens is peculiar in compar- 
ison with the levels of intra-specific variability 
exhibited in all other taxa examined. I suspect 
that the difference is the result of prevailing 
systematic problems that could have con- 
fused proper species-level identification. 

Genus Dosidicus Steenstrup, 1 857 
Dosidicus gigas (d'Orbigny, 1 835) 

Material examined. — 1 sex?, ML = 360 
mm, GL = 412 mm, Chinchua Norte, Peru, 
1 6 Oct. 1 941 , MCZ 293699. —1 M, 1 F, ML = 
349—298 mm, GL = 315—299 mm, R/V 
ALASKA Cr. 74A6, Sta. 59, "coastal waters" 
off La Jolla, California, USNM 729467. — 2F, 
ML = 224—174 mm, GL = 254-204 mm, 
lATTC (28), off Manta, Ecuador, Apr. 1962, 
UMML 31.1769. 



320 



TOLL 



Description. — Cross sections (Fig. 6): 
Level A — The ellipsoid medial axis is rela- 
tively shallow and narrow, wider than deep 
and attached to its plate by a commissure that 
is about one-half of the axis width. The lateral 
plates are relatively thick, distally attenuate 
and strongly ventrally recurved. There are 
paired lateral axes. The proximal lateral axis 
is digitiform. The distal lateral axis is subellip- 
soid to subtriangular, wider than deep, and 
attached to the lateral plate by a constricted 
commissure that is about one-half of the width 
of the axis; Level В — The medial complex is 
similar in shape and slightly smaller than that 
of level A. Both lateral axes are broadly fused 
dorsally to the lateral plate. The lobate acces- 
sory process is joined by a narrow, stalk-like 
commissure; Level С — The medial complex 
is fusiform. The lateral complexes are a 
deeply excavated C-shape. The ventromedial 
process is small; Level D — The body is sub- 
triangular, broadest ventrally, with a pair of 
stout, quadrangular, ventral keels. There is a 
tall, medially constricted, dorsal carina. The 
dorsally curved lateral processes are rela- 
tively long and taper to rounded tips. 

Genus Eucleoteuthis Berry, 1916 
Eucleoteuthis luminosa (Sasaki, 1915) 

Material examined. — IF, ML = 177 mm, 
GL = 170 mm, WH-455-II-71, 13°12'S, 
8°58'W, 6 Apr. 1971, USNM 730198. —IM, 
ML = 151 mm, GL = 155 mm, ANTON 
BRUUN Cr. 17, 29°22'S, 79°57'W, dip net 
with night light, July 1966, MCZ 278110. — 
Isex?, ML = 98 mm, GL = 99 mm, SHOYO 
MARU Sta. 16, 16°25.0'S, 96°58.3'W, 13 
Jan. 1963, in stomach of Alepisaurus, UMML 
31.1558. — 4sex?, ML = 96-89 mm, GL = 
101-89 mm, SHOYO MARU Sta. 16, 16° 
25.0'S, 96°58.3'W, 13 Jan. 1963, in stomach 
of Thunnus obesus, UMML 31.1557. 

Description. — Cross sections (Fig. 7); Level 
A — The medial axis is ovoid to subellipsoid, 
wider than deep, and is broadly attached to the 
medial plate. The lateral plates are broad, di- 
stally tapered to a narrow rounded tip and ven- 
trally recurved. The single lateral axis is sub- 
ovoid, wider than deep, and broadly attached 
to the plate; Level В — The medial complex is 
similar in shape and about one-half to one- 
third of the size of that in level A. The lateral 
axis and plate are broadly fused. The lobate 
accessory process is separated from the plate 
by an offset pair of proximal, ventral and distal, 
dorsal sulci (arrows); Level С — The medial 



complex is subcylindhcal and minute. The lat- 
eral complexes are bullet-shaped to lobate, 
rounded laterally with a pair of small, ventro- 
and dorsomedial processes (arrows). An ad- 
medial cleft is absent; Level D — The body is 
subcylindrical with a shallow, ventral sulcus 
(arrow) between a pair of low, dome-like ven- 
tral keels. The lobate, slightly dorsally curved 
lateral processes are broadly attached to the 
body. The short, relatively broad, dorsal carina 
is slightly inflated apically. 

Genus Ornithoteuthis Okada, 1927 
Ornithoteuthis antillarum Adam, 1957 

Material examined. — IM, ML = 179 mm, 
GL = 175mm, ORE II Cr. 43, Sta. 123, 
12°54'N, 70 31'W, 0-732m, 24 Feb. 1973, 
trawl, UMML31.1726. — 4F, ML = 164—123 
mm, GL = 147—125 mm, ORE 3250, 
29'14'N, 87°40'W, 0-732 m, 60' MWT, 26 
Apr. 1961, UMML 31.397. —IM, IF, ML = 
157—129 mm, GL = 147—116 mm, ORE 
3670, 20°00.5'N, 88°22'W, 732 m, 40' flat 
trawl, 30 July 1962, UMML 31.438. —IM ML 
= 155 mm, GL = 133mm, ORE 2944, 
2740' N, 90°50'W, 60' MWT, 183-229 m, 24 
Aug. 1960, UMML 31.253. —IM, 1F, ML = 
153—126 mm, GL = 136—115 mm, ORE 
3254, 29°00'N, 88°02'W, 247 m, 60' MWT, 27 
Apr. 1 961 , UMML 31 .476. — 2F, ML = 1 49— 
104 mm, GL = 129—91 mm, ORE 5449, 
19°55'N, 68°57'W, night light, 1 June 1971, 
UMML 31.1618. —IM, ML = 101 mm, GL = 
95 mm, CI-264, 23°53.4'N, 77°08.9'W to 
23^54. 7'N, 77°11.7'W, 1301-1329 m, 41' 
Standard Blake Trawl, 3 Nov. 1974, UMML 
31.1670. —IM, ML = 75 mm, GL = 70 mm, 
ORE 1959, 26°55'N, 89°10'W, 2269 m, 23 
Sept. 1957, UMML 31.213. 

Description. — Cross sections (Fig. 8): 
Level A — The medial axis is massive, subcy- 
lindrical, broader than deep, and broadly at- 
tached to the medial plate. The lateral plates 
are thick, slightly recurved ventrally and 
broadly rounded laterally. The single lateral 
axis is massive, ovoid, broader than deep, 
with a broad commissure; Level B— The me- 
dial complex is similar in shape and about 
one-third smaller than that of level A. The lat- 
eral complexes are massive, subovoid, with a 
broad, shallow, ventrolateral sulcus; Level 
С — The lateral fields are narrow, thick, and 
extend from the ventrolateral borders of the 
massive subovoid medial complex. The lat- 
eral complexes are subovoid and slightly con- 
cave laterally. An admedial cleft is absent; 



OMMASTREPHID GLADIUS MORPHOLOGY 



321 



Level D — The body is rectangular, wider than 
deep, with shallow, broad, ventral and lateral 
sulci (arrow). The carina is relatively tall and 
broadly inflated dorsally. 

Remarks. — The narrow, thick lateral fields 
reflect the greater tapering of the posterior 
portion of the free rachis in this species as 
compared to all other ommastrephids. 

Genus Hyalotuthis Gray, 1849 
Hyaloteuthis pelágica (Bosc, 1802) 

Material examined. — 3M, IF, ML = 73 — 
50 mm, GL = 79 - 56mm, DEL II, Acre 12- 
81 -N, 32°09'N, 64 07'W, 0-150 m, 1400 
mesh ET, 24 Aug. 1971, USNM 728882. — 
IM, 2F. ML = 72—68 mm, GL = 76—75 
mm, DEL II, Acre 12-79-N, 32°08'N, 64^^09'W, 
0-450 m, 1400 mesh ET, 23 Aug. 1971, 
USNM 728881 . —1 F, ML = 64 mm, GL = 71 
mm, SHOYO MARU Or. 12, Fish Sta. 20, 
23°25.5'S, 104°36.8'W, 21 Jan. 1963, in 
stomach of Alepisaurus, UMML 31.1561. — 
2F, ML = 61—57 mm, GL = 65— 63mm, 
SHOYO MARU Or. 17, Fish Sta. 19, 
19°37.7'W, 108°27.7'W, 19 Jan. 1963, in 
stomach of Alepisaurus, UMML 31.1560. 



Description. — Cross sections (Fig. 9): 
Level A — The medial axis is small, subovoid 
to subcylindhcal and broadly attached to the 
narrow medial plate. The lateral plates are 
ventrally recurved distally and taper to 
rounded points. The single lateral axis is 
large, three to four times as wide as the me- 
dial axis, subovoid, broader than deep, and 
broadly attached to the plates; Level В — The 
medial axis is broadly joined to the medial 
plate. The lateral complexes are unequally bi- 
lobed with a dorsal and ventral pair of oppos- 
ing, shallow, broad sulci (arrows); Level С — 
The medial axis is minute and subcylindrical. 
The medial plate is highly reduced. The lat- 
eral complexes are subovoid with a "U"- 
shaped admedial cleft equal in depth to about 
one-quarter of the complex width. The ventro- 
medial process is small and dome-like; Level 
D — The body is broadly contiguous with the 
lobate, slightly dorsally curved lateral pro- 
cesses. The ventral sulcus is broad and shal- 
low. The carina is slightly inflated apically. 

Subfamily llllcinae Posselt, 1890 
Genus ///ex Steenstrup, 1880 
///ex coindetti (Verany, 1 837) 

Material examined. — IF, ML = 189 mm, 
GL = 222 mm, Cette, France, June 1861, 



MCZ 2304. — 2M, 2F, ML - 185—122 mm, 
GL = 180—128 mm, P-82, 4 57'N, 9 30'W 
to 4 58'N, 9 32'W, 144m, 5 June 1964, 
UMML 31.1335. —IF, ML = 171 mm, GL = 
178 mm, Naples, Italy, ANSP A8008. 

Description. — Cross sections (Fig. 10): 
Level A — The medial axis is subellipsoid, 
wider than deep, with a narrow commissure. 
The lateral plates are broad and thick, ven- 
trally recurved distally, and tapered to 
rounded tips. There are paired lateral axes. 
The proximal one is subtriangular and 
rounded apically. The distal lateral axis is 
subovoid, wider than broad, and attached to 
the plate via a slightly constricted commis- 
sure; Level В — The medial complex is similar 
in shape and slightly smaller than that in level 
A. The lateral complexes are broad, irregu- 
larly lobate with two, shallow, ventral sulci 
(arrows); Level С — The medial axis is subcy- 
lindrical and broadly fused to the reduced 
plate. The lateral complexes are large and 
subovoid and with a weakly scalloped ventral 
outline. The ventromedial process and adme- 
dial cleft are highly reduced to absent; Level 
D — The body and weakly dorsally curved lat- 
eral processes are broadly contiguous and 
collectively bilobed with a shallow ventral sul- 
cus. The dorsal carina is inflated apically. 

///ex oxygonius Roper, Lu, & Mangold, 1969 

Material examined. — IM, IF, ML = 205 — 
184 mm, GL = 209-193mm, ORE 5784, 
24 28'N, 83 39'W, 567m, UMML 31.899. — 
2F, IM, ML = 148—108 mm, GL = 162— 
1 18 mm, TRITON, south of Palm Beach Inlet, 
Florida, 165 m, 26 May 1953, ANSP A8079. 



Description. — Cross sections (Fig. 11): 
Level A — The medial axis is ovoid to ellipsoid, 
twice as wide than deep, and broadly attached 
to the medial plate. The lateral plates are thick, 
ventrally curved distally and taper to broadly 
rounded tips. There are paired lateral axes. 
The proximal lateral axis is broad, low and 
apically rounded. The distal lateral axis is ir- 
regularly lobate to subrectangular with 
rounded edges and broadly joined to the lat- 
eral plate; Level В — The medial complex is 
similar in shape and one-third to one-half the 
size of that of level A. The lateral complexes 
are large, laterally rounded, with two or three 
ventral sulci; Level С — The medial complex is 
similar in shape and about one-third smaller 
than that of level B. The lateral complexes are 
massive and subovoid. The ventromedial pro- 



322 



TOLL 



cess is dome-like. The admedial cleft is highly 
reduced to absent; Level D — The body and 
weakly dorsally curved lateral processes are 
broadly contiguous and collectively bilobed 
with a shallow ventral sulcus. The dorsal ca- 
rina is inflated apically. 

Genus Todarops/s Girard, 1890 
Todaropsis eblanae (Ball, 1841) 

Material examined. — 1M, IF, ML = 131 — 
1 14mm, GL = 133—120 mm, Atlantique Sud 
Sta. 154. 0°15'S, 847'E, 239 m, 15 Mar. 
1949, UMML31.1351.— 2M, IF, ML = 89— 
78 mm, GL = 90— 76mm, Gerónimo Sta. 2- 
236, 4°03'S, 10°22'E, 0-201 m, bottom trawl, 
8 Sept. 1963, USNM 730204. — 2M, 3F, ML 
= 68—46 mm, GL = 70—46 mm, P-254, 
3°50'N,7=08'Eto3=5rN, 7°12'E, 172-148 m, 
14 May 1965, 40' ОТ, USNM 727408. 

Description. — Cross sections (Fig. 12): 
Level A — The medial axis is massive, subel- 
lipsoid and twice as wide than deep. The me- 
dial commissure is about one-third as wide as 
its axis. The lateral plates are relatively nar- 
row, ventrally recurved distally and taper to a 
narrow rounded tip. The single lateral axis is 
subovoid to subellipsoid, wider than deep, 
about one-half as wide as the medial axis, 
and broadly joined to the lateral plate; Level 
В — The medial complex is similar in shape 
and size to that in level A. The lateral com- 
plexes are irregularly lobate and subequal in 
width to the medial complex; Level С — The 
medial complex is large and fusiform. The lat- 
eral complexes are irregular with a shallow, 
admedial cleft and small ventromedial pro- 
cess; Level D — The body and lateral pro- 
cesses are relatively small and collectively 
subrectangular. The dorsal carina is stout and 
slightly inflated dorsally and forms a right an- 
gle with respect to the body to give an overall 
appearance of a "l" shape. 

Subfamily Todarodinae Adam, 1960 

Genus Todarodes Steenstrup, 1880 

Todarodes sagittatus (Lamarck, 1 799) 

Material examined.— ^M, ML = 186 mm, 
GL - 178 mm, R/V TRIDENT, 36°55'N, 
01°04'W, 143—150 m, 10'IKMT, 21-22 Aug. 
1970, USNM 727741. 

Description. — Cross sections (Fig. 13): 
Level A — The medial axis is subovoid, wider 
than deep and broadly attached to the medial 
plate. The lateral plates are relatively thin, 
slightly curved ventrally and taper to rounded 



tips. The single lateral axis is relatively small, 
subovoid and wider than deep; Level В — The 
medial complex is similar in shape and slightly 
smaller than that in level A. The lateral com- 
plexes are roughly bilobed with slightly offset 
dorsal, distal and proximal, ventral sulci; Level 
С — The medial complex is similar in shape 
and about one-half of the size of that of level 
B. The lateral complexes are hook-shaped. 
Theadmedial cleft is roughly "V"-shaped with a 
depth of about one-third of the complex width. 
The ventromedial process is small and dome- 
like, occasionally reduced; Level D — The 
body is broadly attached to the large, dorsally 
curved lateral processes and has a small ven- 
tral sulcus. The lateral processes terminate in 
rounded lobes. The carina is tall, relatively nar- 
row, and inflated apically. 

Todarodes pacificus Steenstrup, 1880 

Material examined.— 2F, ML = 302—287 
mm, GL = 301—292 mm, 40°16'N, 
133°14.5'E, night angling, 7 Sept. 1967, 
USNM 730206. 

Description. — Cross sections (Fig. 14): 
Level A — The medial axis is irregularly ovoid, 
slightly wider than deep, and broadly attached 
to its plate. The lateral plates are thick, ven- 
trally curved and digitiform distally. The single 
lateral axis is ovoid to subovoid and wider 
than deep. The lateral commissure is about 
two-thirds of the width of the axis; Level В — 
The medial complex is similar in shape and 
slightly smaller than that in level A. The lateral 
axis is subellipsoid, broader than deep. The 
relatively large accessory process is sepa- 
rated from the lateral plate by an offset pair of 
narrow, deep, proximal ventral, and relatively 
broad, shallow, distal dorsal sulci. In some 
specimens the accessory process is pedun- 
culate; Level С — The medial complex is sub- 
fusiform. The lateral complexes are large and 
hook-shaped with an angular, admedial cleft 
equal in depth to about one-quarter of the 
complex width. The ventromedial process is 
small and dome-like; Level D — The body is 
broadly attached to the large, dorsally curved 
lobate, lateral processes. The carina is sub- 
trapazoidal, broad basally and tapers apically 
to a broadly rounded tip. 

Genus Nototodarus Pfeffer, 1912 
Nototodarus sloanii {Gray, 1849) 

Material examined. — 1F, ML = 194 mm, 
GL = 214 mm, Tasmania, Jan. 1875, USNM 
576996. — 2M, ML = 170—160 mm, GL = 



OMMASTREPHID GLADIUS MORPHOLOGY 



323 



183 172 mm, Auckland, New Zealand, 

Jan. 1953, USNM 575461. 



Description. — Cross sections (Fig. 15); 
Level A — The medial axis is subovoid and 
broader than deep. The medial commissure is 
about one-half as wide as the axis. The lateral 
plates are relatively thick, ventrally recurved 
and bluntly pointed laterally. The single lateral 
axis is subovoid, wider than deep; Level В — 
The medial axis is ovoid, subequal in width to 
that of level A. The commissure is only slightly 
constricted. The lateral axis and plate, and in 
some specimens the accessory process, are 
broadly fused into a broad, roughly bilobed 
lateral complex with a pair of opposing shal- 
low, broad, dorsal and ventral sulci. In some 
specimens the accessory process is irregu- 
larly lobate and more distinctly set off from the 
plate by a deep, narrow, ventral sulcus; Level 
С — The medial complex is small and subfusi- 
form. The lateral fields are narrow. The lateral 
complexes are subovoid with a small adme- 
dial cleft equal in depth to about one-fifth of 
the complex width. The ventromedial process 
is small and dome-like; Level D — The body is 
broadly fused to the large, lobate, dorsally 
curved lateral processes. The ventral sulcus 
is shallow and broad. The canna is inflated 
apically. 

Remarl<s. — In a single specimen (USNM 
576996), the cross sections at levels В and D 
were grossly asymmetrical and anomalous. 
The cause of this condition is unknown. The 
animal was normal in all other respects and 
was apparently unaffected by this condition. 

Nototodarus hawaiiensis (Berry, 1912) 

Material examined.— ^f, ML = 104 mm, 
GL = 122 mm, TC-36, Sta. 24, 2r09.7'N, 
157°29.3'W to 21°09.8'N, 157 24.6'W, 183 
m, 4-5 May 1968, USNM 730203. — 2F, ML 
= 114— 93 mm, GL = 114— 99 mm, TC-35, 
Sta. 15, 21°05'N, 156°26'W to 2r05'N, 
156°32'W, 361