1962 PROCEEDINGS

NATIONAL

SHELLFISHERIES

ASSOCIATION

Volume 53

PROCEEDINGS

of the

NATIONAL SHELLFISHERIES ASSOCIATION

Official Publication of the National Shellfisheries

Association; an Annual Journal Devoted to

Shellfishery Biology

Volume 53
August 1962

Published for the National Shellfisheries Association by
Bi-City, Inc., Bryan, Texas

1964

TABLE OF CONTENTS

Research and the Oyster Industry J. L. McHUGH 1

A New Hydraulic Rake for Soft-Shell Clams

I.C. MEDCOF and J. S. McPHAIL 11

Causes of Mortality of the Sea Scallop, Placopecten magellani-

cus J. C. MEDCOF and NEIL BOURNE 33

Mortality of Pacific Oysters, Crassostrea gigas (Thunberg) in

Various Exposure Situations in Washington

WALTER T . PEREYRA 5 1

Oyster Mortality Studies in Virginia. IV. MSX in James River

Public Seed Beds JAY D . ANDREWS 65

Dried Unicellular Algae as Food for Larvae of the Hard Shell

Clam, Mercenaria mercenaria

HERBERT HIDU and RAVENNA UKELES 85

Experimental Farming of Hard Clams, Mercenaria mercenaria

in Florida R. W. MENZEL and H. W. SIMS 103

Seasonal Growth of Northern and Southern Quahogs, Mer-
cenaria mercenaria and M_. campechiensis , and Their
Hybrids in Florida R. WINSTON MENZEL 111

Seasonal Gonadal Changes in Female Soft-Shell Clams, Mya

arenaria , in the Tred Avon River, Maryland

WILLIAM N. SHAW 121

Incidence of Malacobdella in Mercenaria campechiensis Off

Beaufort Inlet, North Carolina HUGH J. PORTER 133

Tests of Internal Tags for Green Crabs (Carcinus maenas) ....

JOHN W. ROPES 147

Studies on Oyster Scavengers and Their Relation to the Fungus

Dermocystidium marinum . . . HINTON DICKSON HOESE 161

A Technique for Separating Small Mollusks from Bottom Sedi-
ments .... GARETH W. COFFIN and WALTER R. WELCH 175

ASSOCIATION AFFAIRS

Annual Convention 181

Officers and Committees 181

Information for Contributors 182

"■

OTHER TECHNICAL PAPERS PRESENTED AT THE
1962 CONVENTION

A laboratory-field approach to soft-shell clam farming

JOHN W. BLAKE
Preparation of frozen sections of oysters for pathological studies

and some applications of the method. .MELBOURNE R. CARRIKER

The gill circulation in the American oyster Crassostrea virginica

ALBERT F. EBLE
Role of salinity in the distribution of the oyster drill.. Urosalpinx

cinerea , in the Rappahannock River, Virginia

GEORGE W. GRIFFITH and JAMES 3. ENGLE
Shell and meat growth of oysters as influenced by various carbo-
hydrate food supplements DEXTER S. HAVEN

Seasonal sexual pattern in the Pacific oyster (Crassostrea gigas

Thunberg) in Washington State

STANLEY C. KATKANSKY and ALBERT K. SPARKS

Evaluation of different materials as spat collectors

WARREN S . LANDERS

Shellfish hatcheries and their future

VICTOR L. LOOSANOFF and HARRY C. DAVIS
Effects of selected quantities of chemicals tried under field con-
ditions on shellfish-killing gastropods

CLYDE L. MACKENZIE, JR. and WILLIAM T. GNEWUCH
Some comments on the nature of the environment from which shell-
fish draw their food NELSON MARSHALL

Neuroendocrine regulation in the oyster, Crassostrea virginica

R. NAGABHUSHANAM
A review of the Dermocvstidlum culture method, with suggested

modifications and precautions SAMMY M. RAY

Notes on the occurrence of Dermocvs'tidium in the Gulf of Mexico

during 1961 and 1962 SAMMY M. RAY

Growth and mortality of tray oysters in Lower Barataria Bay,

Louisiana, during 1959 and 196 0

SAMMY M. RAY and JOHN G. MACKIN

Gametogenesis in Mya arenaria from New England

JOHN W. ROPES and ALDEN P. STICKNEY

An experiment on mass culture of bivalve larvae

CARL N. SHUSTER and THEODORE P. RITCHIE
Normal postmortem changes in the Pacific oyster . . . ALBERT K. SPARKS
Geographic and seasonal patterns of paralytic shellfish toxicity in

Washington A. K. SPARKS,

A. SRIBHIBHADH and K. K.CHEW

RESEARCH AND THE OYSTER INDUSTRY

J . L . McHugh

Chief, Division of Biological Research
U.S. Bureau of Commercial Fisheries

ABSTRACT

In an address to a joint meeting of oyster growers and biologists,
it is pointed out that the oyster industry could be greatly benefited
immediately if the knowledge already available were fully utilized. An
"extension service" is proposed for the shellfish industry.

INTRODUCTION

One of the things I like about these Annual Meetings, which I
have been attending regularly since I first came to Chesapeake Bay in
1951, is the close and friendly association it fosters between scien-
tists and the industry. I think we have learned to know and respect
each other over the years .

But I have suspected that we do not really understand each
other very deeply and that we somehow are failing to help each other
as much as we might do. After all, our shellfish biologists exist only
to serve you, the shellfish industry. If the scientific information we
produce is not fully utilized by industry, nor by State legislation and
management agencies, then your investment in scientific research is
not paying dividends .

I am going to talk about some of our failures— ours and yours
as well. This may give you all some things to think about, and may
start action that can lead the oyster industry out of its present slump.
I am going to use the industry in Chesapeake Bay as an example,
because this is the region I know best. But remember, any specific
problems I mention are merely examples. I am not singling out parti-
cular people, organizations, or agencies for criticism .

OYSTER PROBLEMS

What are some of the principal problems of the Chesapeake
oyster industry today? Uppermost in many minds, I am sure, is the
so-called oyster blight, MSX. But this is by no means the only

serious problem we face. Another problem, Dermocystldium, has
killed large numbers of oysters in the past and will do so again.
Oyster drills are a constant threat to planters in many areas, sea stars
a recurring plague in others. The success of oyster setting varies
widely from year to year, making the future uncertain. Oysters often
do not grow well in places where setting is good . Grounds on which
fattening is best are not necessarily good for setting or growth. These
are some of the principal problems of biological origin.

But these are not all our troubles . Many physical factors
threaten the success of your operations . Hurricanes and other severe
storms may wipe out your crops by burying them in silt and mud.
Unusually severe freshets, especially in summer, may kill without
warning. Man-made troubles are even more disturbing, for they are
becoming more frequent and troublesome as our cities and industries
grow. Pollution from sewage can eliminate your harvest without killing
a single oyster. Dredging, filling, bridge and tunnel building, and
other engineering activities can do great damage. Industrial pollution
can foul the waters so that only noxious pollution-loving forms can
survive. Insecticides and radioactivity are relatively new sources of
trouble, which cannot fail to increase as time goes on. Even such
innocuous-appearing substances as the common household detergents
now so widely used offer a yet unknown threat of injury to oysters and
other estuarine animals .

WHAT HAS SCIENCE ACCOxMPLISHED ?

What have scientists done to help you with these problems?
Much research effort has been devoted to MSX in the past two or three
years . Yet we do not know the identity of the organism we call MSX,
nor have we proven that it is the cause of death. We have learned a
great deal about the geographic distribution, seasonal occurrence,
and biological effects of Dermocystidium . Yet we have no direct
method to control its lethal effects . We have developed several
methods to control oyster drills . We believe that these methods are
much more effective than the tedious and inefficient trapping method.
Yet none of these methods is in general use in the oyster industry.
We understand quite well the geographic and seasonal patterns of
oyster setting. We have fairly extensive historical records of their
annual variations . Yet we do not know how to control oyster setting
in the natural environment. We know that oysters often set best in
certain areas, grow best in other places, and fatten best elsewhere.
Yet we do not understand the reasons for these differences and cannot
control them. Someone has said, quite truthfully, that we probably

know more about oysters than we do about any other marine animal.
Have we made good use of this abundant biological knowledge? I do
not think so .

Where do we stand with respect to the effects of physical
forces, natural or man-made? We understand the biological effects of
storms . Storms kill oysters not only by smothering them in mud or
sand, but also produce side effects that kill by depleting the dissolved
oxygen supply or by releasing toxic hydrogen sulphide. Yet we must
stand helplessly by while these effects do their damage. We are
equally powerless to influence the effects of freshets . We have
achieved a considerable degree of control over the effects of domestic
pollution, and a somewhat lesser, but still significant, control over
industrial wastes. Yet pollution problems are with us continually, and
they involve important issues that conflict often with fishery interests.
Thus, we are achieving far less than we would desire in modifying the
effects of pollution, and probably always will.

By proper planning and study we can mitigate the effects of
engineering projects upon our oyster resources . Yet we cannot always
anticipate these effects before the change is made . Problems related
to proposed deepening of the James River channel, which passes through
the heart of the Virginia seed oyster area, are a good example. Every-
thing we know about the circulation of estuaries and the characteristics
of oyster setting tells us that deepening the channel is apt to affect
the seed-producing characteristics of the James River adversely. Yet,
since we do not understand completely the characteristics of a good
oyster setting area, and probably never will, we cannot be certain.

We are just beginning to develop scientific knowledge of the
effects of radiation and of pesticides upon marine organisms . We have
much to learn. We know that detergents have adverse effects upon
organisms when present in sufficient quantity. We know that these
substances may pass through sewage treatment plants almost unchanged.
We know almost nothing of the effects of wastes of various kinds on
maturation, spawning, setting, feeding, growth, mortality, and other
characteristics of oysters . The complex and difficult problems pre-
sented by all these sources of pollution have led some people to sug-
gest that the oyster culture of the future must be carried out entirely
under artificial conditions, using the most modern scientific techniques.
This certainly seems true in the immediate future for areas close to
large human population concentrations. The Bureau has made con-
siderable strides already in developing such techniques . We have
plans for further development in the near future. But there are still

large areas of virgin or near-virgin estuarine waters where scientific
knowledge, if it is applied intelligently, can improve oyster quality
and yield .

USING OUR SCIENTIFIC KNOWLEDGE

I have purposely painted a gloomy picture of the practical
application of our scientific knowledge of oysters. I did this to
develop the thought that our private viewpoints and public policies
have been at fault. Let us look at our scientific knowledge from
another point of view, from a direction that suggests how this knowl-
edge can be put to practical use.

Previously I dealt with the principal biological problems of
oyster growers from the standpoint of one who would attack them
directly. This was consistent with the popular belief that we should
solve them by frontal assault, as it were. It is obvious that our
extensive scientific knowledge offers no immediate hope of direct
control of epidemics of M3X or Dermocystidium . Industry has been
slow to adopt or even test moire effective methods of pest control .
Biologists have a great deal of information on spawning, setting,
growth, fattening and mortality, among other things, but they have not
shown how these characteristics can be altered favorably by direct
methods .

In other respects, however, there is a great fund of knowledge
that shows how these problems can be attacked indirectly. The Vir-
ginia oyster, for example, is an estuarine animal. Under natural con-
ditions it usually thrives best in waters of intermediate salinity. If it
survives at all in saltier waters, it usually does so most successfully
along the shoreline, where it is periodically exposed by the tides.
What does this signify to the oyster planter? As far as Chesapeake
Bay is concerned, it means that the upper limit of oyster distribution
is determined by its ability to tolerate fresh water. Its seaward limit
is set by the tolerance of its enemies and diseases to waters of
reduced salinity. The oyster's principal advantage over these enemies
and diseases is that it can survive and thrive in salinities that do not
drop below five to seven parts per thousand for any considerable period,
while its predators such as drills and starfish, and diseases like MSX
and Dermocystidium, are not a serious problem in waters below 12 to
15 parts per thousand. Extensive areas in Chesapeake Bay lie within
these limits of salinity (Fig . 1), and this zone includes most of the
naturally producing oyster grounds . Some of this favorable water is
unsatisfactory because the bottom is unsuitable, or for other reasons.

$£;

Fig. 1. Chesapeake Bay, showing areas shallower than 3 0
feet in the Bay and less than 20 feet in the estuaries, where salinities
are favorable for oysters but unfavorable for their principal enemies
and diseases. Where the bottom is suitable and other conditions are
satisfactory, these are the areas most favorable for oyster planting.

But much of it could be utilized for oyster growing if local traditions
and State laws permitted it. It is significant that in Virginia, for the
most part, these favorable zones lie up the estuaries . The James
River seed area and the public bars of the Rappahannock River are
examples. In Maryland, on the contrary, a very extensive part of the
Bay itself lies between these limits. This, then, suggests one way in
which the problem of predators and disease can be solved directly. We
could grow oysters in areas where these problems are absent . Is the
industry taking full advantage of this knowledge? I do not think so.

Your immediate reaction may be— public opinion and policy
are against this in many areas, and we are prevented by law from fol-
lowing your advice. My reply is: laws can be changed if public
opinion favors change . How can we work with you to help you take
advantage of accumulated scientific knowledge?

SEED OYSTERS

What is probably the greatest single seed oyster producing
area in the world is concentrated in a short stretch of the James River.
We do not know all the reasons for the great productivity of this area .
We do know that it has continued to yield a bountiful harvest each year
almost without fail, despite a heavy fishery. Nowhere else in the Bay
is such a steady and heavy set of oyster spat obtained . Yet once set,
these oysters grow quite slowly. The characteristics that favor
spawning and setting obviously do not favor growth.

This very slowness of growth, however, favors this seed for
planting in drill-infested waters in the lower part of Chesapeake Bay.
Slow-growing oysters are thick-shelled oysters, and drills can pene-
trate thick shells only with difficulty. This is why an extensive oyster
planting industry has managed to develop in the salty waters of the
lower Bay, until the recent epidemic of MSX virtually wiped out this
operation. If bottoms more favorable for planting, in terms of absence
of drills and blight, had been open for leasing, these less favorable
bottoms might never have had to be used .

The waste involved in the crude farming techniques employed
in the Virginia parts of Chesapeake Bay is startling. I do not have
accurate figures for the total numbers of oysters in a bushel of seed
from the James River, for these oysters range in size from microscopic
spat just recently set to oysters of market size. Because growth rates
are slow in the seed area, many ages are represented in the seed
harvest. The most abundant of all age groups, naturally, is the group

that set during the year of harvesting . When harvesting begins in
October, these newborn oysters, only about two months old, are very
delicate . Probably none survives the multiple perils of damage during
harvesting, desiccation on board the tonger's boat and the buy-boat
before replanting, and smothering by mud and silt or predation by drills
and other enemies shortly before replanting. This means that at least
half of the seed harvested in the James River is wasted.

There are several ways in which much of this waste could be
avoided. It might be practical to delay the seed harvest until spring,
when these newly born young will have increased to a size more able
to withstand transplantation. Survival could be increased substantially
if bottoms could be leased in waters of lower salinity. Seed would be
used more efficiently if economical methods could be developed to
plant cultch in seed areas and transplant the spat each year. Have
we done everything possible to solve these problems? I do not think.
so.

Even the traditional methods of harvesting and marketing
oysters work against the best interests of the oyster grower. The old
myth of the "R" months has had much to do with this, I suspect.
Oystering traditionally begins in fall, shortly after the end of the
spawning season. At this time the meats are perhaps at their poorest,
thin and watery with poor reserves of glycogen, the animal starch that
gives plump oysters their sweet and delicate flavor. Moreover, sum-
mer is the greatest period of stress for the oyster in othsr ways . All
biological activity is at its peak. Many oyster pests and diseases
are in their most active period. In the saltier waters of the lower
Chesapeake, where most planting acreage is located, biologists know
that summer mortalities of 25 to 50% are not uncommon. This means
that, in many years, from one-fourth to one-half the oysters alive on
these grounds at the end of June will be dead by October. Moreover,
the growth rate of adult oysters slows down in summer. There is very
little gain in size to balance these substantial losses by death.

If you were to harvest your crop in late spring, about four or
five months earlier than usual, you would gain in at least three ways .
Your oysters would not be appreciably smaller. You would harvest
perhaps one and one-third to twice as many bushels . Each bushel
would yield perhaps twice as many pints of meats . These meats
would be of prime quality, at their fattest and tastiest. It would not
be unusual to produce twice to four times the usual yield of shucked
meats by adopting this simple change in harvesting practice. I can
hear some of you saying the price would be lower, because demand is
poor at that season. This is primarily a matter of tradition, I think.

People's habits can be changed, or methods can be developed to pre-
serve these prime meats, for sale when market conditions are at a
maximum. Have you done everything possible to take advantage of
this knowledge? I do not think so.

HAS PUBLIC MANAGEMENT SUCCEEDED?

In most oyster producing states the naturally producing
grounds have been set aside as public grounds, either to produce seed
for planting on barren leased ground, or for harvesting market-size
oysters directly. Many states spend considerable sums to protect
and rehabilitate these grounds . I think we would all agree that these
efforts have not been entirely successful. Combined oyster produc-
tion of Maryland and Virginia has declined from 22 million bushels in
189 0 to less than 6 million in 1961. When oystering began, the entire
crop was harvested from public grounds. Now about three-quarters of
Virginia's production comes from formerly barren grounds leased to
private planters. Many formerly productive public bars in Virginia now
are barren. These could be returned to production under private mange-
ment, but state laws and local traditions are opposed to leasing of
public grounds, even if they produce no oysters .

The decline of the Potomac River oyster industry has been
well documented . Toward the end of the nineteenth century the total
area of productive natural oyster ground in this river reached the
impressive figure of 42 square miles. Today the productive area is
only a quarter this size, and yields per acre almost certainly are
lower. The newly created Potomac River Fisheries Commission has
the opportunity to create oystering history by taking bold steps to
reverse this trend .

THE FUTURE

These matters are far more complicated than I have indicated
here. To introduce all the complications in a short talk would hide
the point I am trying to make. To stimulate thought and discussion,
I am going to make a positive statement that none of you may accept.
It is simply this: I believe that if the favorable characteristics of
Chesapeake Bay and its estuaries for oyster culture had been used to
their maximum, the oyster industry would not be suffering the diffi-
culties it suffers today. Please give this matter some serious thought.
The knowledge necessary to establish a rational system of scientific
management has been available for many, many years. Tradition and
prejudice have prevented maximum use of scientific and practical
knowledge .

We could, even today if we were free to do so, revolutionize
the oyster industry. Almost any good oysterman could write out a plan.
The basic knowledge could be drawn on charts, showing the areas in
which oysters can thrive but where their principal enemies and diseases
cannot. Other charts might show setting areas, growing areas, and
fattening areas . Our knowledge is not perfect by any means; but if all
available information were brought together, its extent might surprise
us all. Armed with such a compendium of knowledge, an atlas of
oystering conditions if you will, we then could draw up a master plan
for utilizing these areas to their maximum. We could even make rough
estimates of the increases in yield and quality that might ensue, and
take a stab at making cost estimates. With such a plan, we could
attack the difficult problem of securing general support from all seg-
ments of the industry, then from state legislatures. We might be
surprised to find that changes are not so difficult to make as we have
suspected .

AN ATLAS OF OYSTER GROUNDS

The blackened areas on Fig. 1 are those areas of suitable
depth with salinities favorable to oysters but unfavorable for their
major enemies and diseases . Most of these areas are not available
to private planters . It is obvious that most planting grounds now under
lease are outside this optimal area. Oystering has been sufficiently
profitable, and planters have been fortunate enough to escape serious
natural disasters, so that it has been possible to continue planting
these marginal grounds . But the recent onslaught of MSX has demon-
strated the danger of such practices . This is the time to put our
knowledge to work. The atlas will be a most useful document. It will
bring together the knowledge of scientists, and the accumulated
experience of oystermen. No one with an open mind can deny the con-
clusions that it is certain to bring.

This atlas of oystering conditions will not resemble the crude
chart represented by Fig. 1 . It will be based on a much larger scale,
which will permit detailed identification of bottom types and all the
other pieces of information we will need to know. It will be developed
in consultation with all the people who have pertinent knowledge . It
will not be assembled overnight, for it will require much painstaking
research and discussion. But it may well become the most important
document that your industry could possibly have.

As scientific knowledge grows, these favorable areas can be
extended. Presently proposed methods of oyster drill control, when
they are proven, will allow a considerable extension. The time may

come when control of oyster blights may be possible, either by direct
methods, or by breeding resistant strains. Whatever the future may
bring, the atlas will provide a useful base.

The atlas might be useful in dealing with man-made problems
also. It might, for example, show areas where industrial and domestic
expansion could take place without serious damage to the oyster indus-
try. It might show where dredging, filling, and other engineering
developments should and should not be undertaken. It might well lead
to a concept of zoning our estuarine waters for maximum development
of all such uses, with minimum damage to fishery resources.

A SHELLFISH EXTENSION SERVICE

For some time we in the Bureau have been considering the
advisability of organizing an extension service for the shellfish indus-
try. We believe that we have failed in our responsibility to communi-
cate our scientific knowledge to you. Only by providing a group of
knowledgeable people, who can work with you and help you to develop
the information you need, can we make progress . This group would
assist in testing the effects of changes in oystering methods and make
sure that these trials were properly designed and adequately tested.
Their first job would be to bring together the information and develop
the plans I have outlined. Their continuing responsibility would be to
improve the basic knowledge on which the plan was based. Then they
would feed this knowledge back into their planning, to the benefit of
industry. We hope to put such an extension service into operation
soon. As we develop our plans, we would like to have the help of
industry, state scientists and state administrators. As a team, I
believe we can accomplish miracles . Shall we begin?

10

A NEW HYDRAULIC RAKE FOR SOFT-SHELL CLAMS

J . C . Medcof and J . S . MacPhail

Fisheries Research Board of Canada
Biological Station, St. Andrews, N. B.

ABSTRACT

The Martha's Vineyard hydraulic clam (My a) rake is the prototype
from which the present machine was developed^ The new rake is 2 feet
wide and operates on submerged flats in 8 inches to 3 feet of water. It
requires a steady supply of 130 gallons of water per minute at a pressure
of 25 pounds per square inch. It converts the upper strata of soil into a
soil-water fluid and the clams float to its surface. One man can easily
operate and maneuver the rake because its short nozzles do not pene-
trate the soil. A manual collecting rake is required to gather up the
clams that the hydraulic rake brings to the surface. By combined hydrau-
lic and collection raking a fisherman harvests 8 to 10 square feet of flat
per minute and recovers 95% of the marketable clams from the soil
worked. He damages < 5% of the catch and <5% of the small clams left
behind in the soil. The catch per hour per man is only one-third that of
an escalator harvester but three times that of a man with a clam hack.
A hydraulic rake costs less than $500. Thus there appear to be good
economic and conservation reasons for its industrial use.

INTRODUCTION

The clam hack or clam fork is an ancient harvesting tool. It is
very destructive of clam stocks (Needier and Ingalls, 1944) and it is
tiring to fishermen. We believe the hydraulic rake, which operates on
flats while they are submerged, is a better harvesting tool. It causes
so little damage to clam stocks that it must be ranked high as a savings
gear. And it fishes so fast and it is so easy to use that clam fishermen
need no longer be regarded as underpaid drudges.

ACKNOWLEDGMENTS

We wish to thank Dr. D. B. Quayle, formerly of the Coast
Oyster Company, who first drew our attention to hydraulic raking; Mr.
F. C. Wilbour and Dr. G. C. Matthienssen (formerly of the Massachu-
setts Department of Natural Resources), who in 1959 acquainted us with
Martha's Vineyard fishermen who were using hydraulic rakes; the Indus-
trial Development Service of the Canadian Department of Fisheries
which financed our 196 0 development work; Mr. H. Y. Brownrigg of the

-11-

Biological Station at St. Andrews for his assistance in both 1960 and
1961, and our Director, Dr. J. L. Hart, who fostered this project.

The junior author was in charge of the hydraulic rake develop-
ment project and this paper is largely based on his reports (MacPhail,
MS, 1961 and MacPhail and Medcof, MS, 1962).

PROTOTYPE HYDRAULIC RAKE

The prototype of the hydraulic rake which the junior author saw
in use on Martha's Vineyard in 1959 is illustrated in Fig. 1. Water for
hydraulic fishing is supplied by a pump-motor assembly that is floated
in a dinghy. The water passes through a 2-inch discharge hose
attached to the upper end of the rake handle which is a 2-inch pipe.
From there it passes through the rake head, which is also made of 2-
inch pipe, and escapes through nozzles that project from the head like
the prongs of a garden rake. When the nozzles are pulled through
sandy soil, clams (Mya arenaria) are brought to the surface.

Functional Principle

It is often said that hydraulic diggers wash clams out of the
soil. This may be true for the hydraulic escalator harvester. But
experiments with normal and artifically weighted clams (lead pellets in
mantle cavities) show that in hydraulic raking clams are not washed
out — they float up. The significant results are summarized in the
accompanying diagram of what happens in laboratory aquaria when
clams of different specific gravities are hosed.

Sp. Gr.
of Clams

< 1.6
1.6

> 1.6

-12-

The result was the same regardless of whether the test clams
were buried in the soil or lying on the surface of the soil when hosing
started. Normal clams and lightly weighted clams came to or remained
at the soil surface; weighted clams with specific gravities of about 1.6
were usually found at mid-depths in the soil and weighted clams with
specific gravities appreciably greater than 1.6 were usually found on
the very bottom of the aquarium beneath all the soil.

These experiments and field observations reveal what the hydrau-
lic rake's water jets actually do. They convert the upper stratum of the
clam flat from a solid to a fluid state. The fluid soil-water mixture
produced by the rake has a high specific gravity and in it clams pop up
like corks. We presumed that the diameter, length, shape, force and
spacing of the jets could be varied a great deal and still be effective
so long as they performed this converting function. In other words,
it should be possible to modify the design of the prototype rake over a
wide range without diminishing its effectiveness .

Operational Characteristics

In 196 0 we built a modified Martha's Vineyard rake and carried
out performance tests on intertidal clam flats. In the dinghy we used
a 2-inch x 2-inch self-priming centrifugal force pump powered by a
1-cylinder, 6 horsepower, air-cooled gasoline engine. It delivered
approximately 13 0 gallons of water a minute at 25 pounds pressure per
square inch.

We found it necessary to wear breast-high rubber waders when
operating the rake (Fig. 2) which fished in 8 inches to 3 feet of water.
The clams came up clean with little breakage or scattering. At the
following low tide we examined the rake track (Fig. 2). It resembled
the track of the hydraulic escalator harvester (Medcof, 1961). Its
flat bottom was 2 1/2 inches below the level of the undisturbed flat
and was 24 inches wide. The verge-to-verge width of the track at the
level of the flat was 33 inches. Crumbling of the verges accounts for
this difference in widths. No doubt the tidal currents aided in this
crumbling process.

The verges of the track were built up with sand carried from the
track by the water jets. This build-up was about 1 inch deep at the
verge and tapered off so as to be indistinguishable at a distance of 10
to 12 inches from the verge.

After about 24 hours the track was quite firm but the depression
in the flat seemed likely to last for several months as is the case
following operations with escalator harvesters and clam hacks.

-13-

Hot* Clamp,

Fig. 1. Prototype hydraulic rake used in Massachusetts, U.S.A.
(Drawing by F. Cunningham.)

Fig. 2. Hydraulic rake tracks through test plots on intertidal
clam flats at Clam Harbour, Nova Scotia. Mr. MacPhail is wearing
chest-high rubber waders needed for hydraulic raking.

-14-

The prototype rake had some shortcomings . The nozzles pene-
trated deep into the soil and this made the rake hard to haul during
fishing. It was very heavy to carry around and awkward to position
accurately on areas that were to be fished. Furthermore, the fishing
operation itself was awkward because the fisherman was always in
the middle of the assembly. He faced the rake as he hauled it along
by walking backwards. Tnus the dinghy with its pumping equipment
and the discharge hose were always behind him and often got in his
way .

Collection Rake

If clams are not gathered soon after they are raked they burrow
back into the soil. Thus, for hydraulic harvesting, a collecting rake
is a necessary complement to the hydraulic rake.

Form. The collection rake we used is shown in Fig . 3 . It has
a bowed head to fit the contours of the track made by the hydraulic
rake. The teeth were 5 inches long at the center of the head and 7
inches long at the ends. The teeth were covered almost to their tips
with 1/4-inch mesh wire screening and the bow of the head was backed
with similar screening to form a collecting pocket.

Efficiency. To test the efficiency of the collecting rake we
resorted to a field tagging experiment. First, we selected 100 clams,
1 3/4 to 2 1/4 inches long, and marked their shells with a file for
identification. We deliberately avoided using Volger's ink because it
would have made the tagged animals conspicuous and distorted the
experimental results .

Next we ran the hydraulic rake for a distance of 3 0 feet through
a sandy clam flat that was covered by 2 1/2 feet of water. This brought
many clams to the surface and we scattered the tagged clams evenly
among the native stock. The track and verges were then carefully
raked with the collection rake and the marked animals were culled from
the catch and counted . This experiment was carried out three times
with a different person raking each time. The counts of the recovered,
marked clams were 95, 99 and 100.

From this we decided that the collection rake was highly
efficient and that no serious error would be incurred if we were to
regard it as 100% efficient. To minimize error in assessing results of
performance tests that were carried out subsequently, all the collection
raking was done by one man and all the tests were made on one flat.

-15-

Fig. 3.
rake track .

Collection rake used for gathering clams from hydraulic

DESIGN OF NEW HYDRAULIC RAKE

Mr. MacPhail modified the prototype design of the body of the
rake several times in the course of performance tests carried on in 1960
and 1961 and his present model (Fig. 4) performs well. Its main features
are:

1. "D"-type steel handle (adjustable) which is fastened to the

ends of the head .

2. Attachment of the discharge hose directly to the head on the
side opposite the handle .

3. Wheels on the ends of the head.

4. Short nozzles that barely touch the surface of the soil as
the rake is wheeled along.

-16-

I D. Mild Stttl

Handlt Adjustment

Reducing Coupling 7 " ,0 ■ "
\ " X 2" Nozzl*

2 D Fir* Hot*
Coupling

Fig. 4. The 1961 MacPhail model of hydraulic rake. Main
features: (1) "D" -type adjustable handle, (2) attachment of hose
directly to rake head on the side opposite the handle, (3) wheels on
ends of head, (4) short nozzles . (Drawing by P. W. G . McMullon).

All four of these features facilitate maneuverability of the new
rake. It can be wheeled around like a lawn mower. The fisherman is
no longer "in the middle." As he fishes he faces both the rake and
the dinghy which floats out of his way either down-tide or down-wind
(Fig. 5) as he walks backwards.

The fourth feature (short nozzles) was incorporated after per-
formance tests using nozzles of various kinds . These tests were
designed to show how nozzle length affects maneuverability of the
rake, raking speed, and per cent recovery and per cent breakage of
harvested clams. Only the 1961 performance tests are reported here.

-17-

Fig. 5. The 1961 MacPhail model of hydraulic rake being
fished by man at right in 2 1/2 feet of water. Man at left is collection
raking a previously hydraulically raked area.

EXPERIMENTS WITH NOZZLES

Work Site

The experiments with nozzles were all carried out in July and
August 1961 at Clam Harbour, Nova Scotia, on the same clam flat
where similar tests were carried out with the hydraulic escalator har-
vest in 1958. This area has been previously described (Medcof, 1961) .
It was chosen because of the uniformity of its clean sandy soil;
because it resembles many of the areas on our east coast where we
think the hydraulic rake might be used industrially and because it
afforded the opportunity of making direct comparisons of performance
of the hydraulic rake and the escalator harvester.

-18-

The tidal amplitude at Clam Harbour is approximately 5 feet and
the rise and fall averages about 1 foot per hour.

Test Plots

Twenty-four test plots, 3 ft x 10 ft, were set up with their long
axes paralleling the direction of the tidal currents . Living clams were
planted precisely 9 per square foot following the 19 58 procedure (Med-
cof, 1961). They measured 1 1/4 to 2 1/4 inches long and were marked
with Volger's ink for identification. Fishing tests were conducted 5
days or more after planting which is sufficient time for the clams to
establish themselves in their barrows (Medcof, 1950).

Four additional 3 ft x 10 ft plots were set up. Two were
planted 9 per square foot with marked, living clams 3/4 to 1 3/8 inches
long — too small to be attractive to fishermen . The other two plots were
similarly planted but in this case the stock was killed in formalin.

Procedure and Results

Four sets of eight nozzles 2,3,7 and 9 inches long were pre-
pared . The 2-inch nozzles barely touched the surface of the soil but
ths others penetrated into the soil 1, 5 and 7 inches during fishing.
The 7- and 9-inch nozzles had three 1/4-inch diameter holes drilled
into the leading face of each to help wash away the soil in the direction
the rake moved . And the free ends of nozzles of all lengths were flat-
tened to give a flat, fan-spray effect to the jets of water.

The principal object of the first series of experiments was to
compare the performance of the various types of nozzles. A single
lengthwise strip was fished through each plot and four or more plots
were fished in testing each nozzle length (Table 1).

All subsequent experiments were made with 2-inch nozzles .

One series was to determine the effect of increased water pres-
sure on the per cent recovery of clams (Table 2). To achieve the higher
pressure (3 0 pounds per square inch), 2 of the 8 jets were sealed off.

The angle of the 2-inch nozzles was varied by adjusting the
handle (Fig. 4) to see if this affected raking performance. In some
tests the nozzles were perpendicular to the surface of the flat. In
others, the free ends of the nozzles were canted toward the operator
approximately 6° from the perpendicular (Tables 1 and 2).

•19-

o tO 00 -vl

Ol Ul * u

1— ■ o <-0 COMCJIC/IJiUtOl- "

*» Cn en en

03 O N ID

CXI ID CD O
Cn CO NJ CO

i— ■ >— ■ ts3 to
-v] si o >-■

OlOOtOO^CDCOO

CD \J O *" cnJ^COCOoi— >010

3

3"

lO CO (fl CO si ID M si (fl CO to io toiococoiumtooi
000001—' co i— > tsj en o oi sj u o^jcoooocnoto ^

CO ID o l— • CO l— i K> l- ■ CD i— ■ tSJ CO UIIOIOUO^OW

(O u Ol H t-O O O CO CO CO M Ol i_i^C01-'^Cs5C0Cr>
J-, 'M CO Ol H-" Ji. O ^J C71 s] CO *> COOl-'COCDCnslCO

55

Cn
tsj

to

en

to

ISO

[S3

CO

3^

to
o

^1

to

to
t-o

55

CD
cn

CO

en

5?

3 2

o o

- 2

CD O

3 N

u3 N

3* cd

5* co

n! -> ro

C" QJ CD

■5 c a

ID i—

C J r-h

CD

o )_L

< QJ

3 3

CD cn

a

Cf Q

O OJ

to §

>

73 CD

2 3

cd ua

O- CD

CD 7s

° <

9 CD

CD 0J

CD

00 3

° sr

CD CD

s

o 2r

• 3*

H

QJ

cr

5! ffi

u.

a-

O

73

CD

cr

QJ

~> >
CD <

2L CD

o

7T m

l-i

QJ CD

3

iQ cQ

CD CD

31

CD

3"
CD

CD

73
CD

O

c

r+
tsi

3

n

CD

cn

to 3

3 *r
on

CD 3"

cn QJ

- 3

§ CD

3 £

5" 5"

3 iQ

3

X g

■-i

*-• TV
O CD

■a m
cd S

o- cd"
o w

C cn

8} o

CD

a.

QJ
73
73
*-1
O

CD R
» CD

73 O
O g
cn cn

s ?

CD -i
CL CD
cn
cn

;-■ 73

CD

t — •

k;

oi
o

r*

O

s

QJ

to

C

{? CD
cn

rt O

tsOHf,

^s [S3

~ <-"
CO

O 73

73
QJ

cn

QJ

3

a

co

° s

i-h N

i — »

3" CD

■< cn

a

-1 tsJ

QJ »

C

co

O

-20-

Table 2. Hydraulic harvesting of market-size clams with pressures of
3 0 p.s .i. and 6 two-inch-long nozzles in 3 ft x 10 ft test
plots, planted 9 per ft^ (158 clams in path of hydraulic rake-
width of cut 21 inches). Nozzles sloped approximately 6°
toward operator in plots 21 and 22; perpendicular in plots
23 and 24.

Plot Raking Clams Clams Average Average Average

no. speed recovered broken speed recovery breakage

%

ft/min

%

%

ft/min

i

%

21

13

.6

103.2

1.8

12,

.7

110

.4

22

11

.8

112.7

4.5

23

16

.2

96.8

0.7

15

.8

99

.3

24

15

.4

101.9

0.0

3.1

0.4

In all the above tests fishing was done during ebb tide and the
clams were gathered with the collecting rake as soon as the murkiness
of the water cleared after hydraulic raking.

The last series of experiments was designed to show how small
clams are scattered by the rake and by the tide after the rake passes .
Under commercial fishing conditions small clams would be left on the
flats. So, in our experiments, these were not gathered up with the
collecting rake. The plots were fished through with 2-inch nozzles
and the positions of the clams were examined at the following low tide
when the flat was exposed (Tables 3 and 4). To further simulate com-
mercial conditions, hydraulic raking was begun 10 feet down -tide from
these plots of small clams and continued up-tide through the plots and
10 feet beyond .

-21-

Table 3 . Scatter pattern caused by hydraulic raking with 2-inch noz-
zles through two 3 f t x 1 0 ft test plots of small, tagged,
formalin-killed clams planted 9 per ft2 . Counts made next
low tide. "T' indicates clams found in track; "S" indicates
clams found on undisturbed soil. The raked area extended
from a point 10 feet down-tide from the plot to a point 10
feet up-tide .

Plot #25 Plot #26 Av.

Disposition
of clams

Distance down -tide
from plot

Surface
T S

Surface
T S

ft

no. no,

no ,

no ,

/o

Raked ground
(above plot)

-10 - 0

0 22

Plot

0 22

Raked ground
(below plot)

0-10

88 0 34

Unraked ground

10 - 35

11

—

23

35 - 60

40

-

13

60 - 85

7

-

8

85 - 110

-

-

2

Total number recovered

99 58 78

46

Distribution of clams
recovered (%)

63 37 63

37

No . recovered
No. in path of digger

Tfr87%Tit=69% 78

No. broken
No . recovered

■—■= 1.3% 7T7= °°/° <1
157 124

-22-

Table 4. Scatter pattern caused by hydraulic raking with 2-inch noz-
zles through two 3 ft x 10 ft test plots of small, tagged,
living clams planted 9 per ft^. Counts made next low tide.
"T" indicates clams found in track; "S" indicates clams found
on undisturbed soil. There was no effort to recover clams
buried outside the track. The area raked extended from a
point 10 feet down-tide from the plot to a point 10 feet up-
tide .

Plot #27

Plot #28

Av

Disposition
of clams

Distance

down-tide

from plot

Surface
T S

Buried
T

Surface Buried
T S T

ft

no .

no .

no .

no .

no .

no. %

Raked ground
(above plot)

-10 - 0

4

0

0

0

0

0

Plot

0

2

0

3

r
3

0

0

Raked ground
(below plot)

0-10

4

0

36

7

0

30

Unraked ground

10 - 35

-

1

-

-

1

-

Total number recovered

10

36 12

1 30

No. dead (included in
totals)

0 0

Distribution of clarm
recovered (%)

21

77 2!

3 69

No . recovered
No. in path of digger

_47
180

--T - 26%

43
180

= 24% 25

No . living clams on surface 7

No . in path of digger 180

4%

12
180

= 7% 5.5

No. broken

0

-23-

HARVESTING EFFICIENCY

Hydraulic harvesting efficiency is a composite. It is affected
by the several factors which modify hydraulic raking and collection
raking efficiencies and may be assessed from the data gathered during
the experiments with different lengths of nozzles .

Hydraulic Raking Speed

Commercial hydraulic raking, as seen on Martha's Vineyard, is
always done at maximum speed. The operator exerts a steady pull on
the handle and forces ths rake through the soil as fast as it will go
without jumping out of its track . Our experiments with the longer noz-
zles were carried out in this same way. They showed (Table 1) that
speed varies inversely with nozzle length, averaging 19.2, 9.6 and
5.2 feet per minute with 3-, 7- and 9-inch nozzles.

The results indicate that the simplest way of increasing hydraulic
raking speed is to shorten the nozzles . This reduces soil penetration
and soil impedance and these cease to be limiting factors when nozzle
length is reduced to 2 inches . The fisherman can then wheel his rake
as fast or as slow as he pleases . He can choose the speed that best
suits his purposes, e.g., the speed that gives him the highest catch .
A series of tests was made to determine this optimum speed for 2-incn
nozzles on the sandy Clam Harbour flats .

Recovery Rate

If the combination of hydraulic raking and collection raking
were completely efficient, 180 clams would have been recovered from
the path of the rake through each plot. It was on the basis of this 180
ideal return that we calculated the tabulated percentage recoveries from
the numbers of clams actually caught.

With 3-, 7- and 9-inch nozzles the average recoveries were
high (79-92%) and not remarkably affected by nozzle length or raking
speed (Table 1).

With 2-inch nozzles, however, the recoveries varied greatly
with raking speed. At the highest speed tested (40 feet per minute) the
recovery was only 53% (Table 1). Apparently a speed of 40 feet per
minute does not provide enough time for the water jets to create and
maintain that consistency and depth of the soil-water fluid mixture
which is required to float clams from their burrows. At intermediate
speeds (20-30 feet per minute) recoveries were high. Neglecting the

-24-

results fcr the foul cut through plot 3, they varied from 88 to 97% and
averaged almost 95% (Table 1). At the lowest speed tested (10.5 feet
per minute) the recovery was artificially high— 120%. This is believed
to have been caused by a deepening of the track and an undercutting
and crumbling of its verges . Our method of calculating the recovery
rate does not allow for this increase in the effective width of the track
and results in some values exceeding 100%.

From Table 2 it might be concluded that increasing the water
pressure from 25 to 3 0 pounds per square inch increased the recovery
but the effect is not clear for two reasons . First, the raking speeds
during the 3 0-pound-pressure tests ranged from 11.8 to 16.2 feet per
minute and we made no 25-pound-pressure tests at comparable speeds
with 2-inch nozzles (Table 1). Interpolation to make allowance for the
speed difference suggests that the pressure change had no effect on
recovery rate .

The second reason for uncertainty is that the width of the track
was changed from 24 inches to 21 inches by stopping a nozzle at each
end of the rake head in order to achieve the higher pressure. This
tended to deepen the track and exaggerate border effects. That is to
say, the number of washouts from the steep sides of the deep track
became relatively larger in proportion to the total catch than when the
track was shallower and 24 inches wide. This accounts for the arti-
ficially high estimates of recovery rates. We do not know how to
correct for this distortion of the results so the data in Tables 1 and 2
are not comparable and we cannot say whether increased water pressure
increases the recovery rate .

We believe that in soils that are more compact than those at
Clam Harbour pressure increases might increase recoveries.

The recovery rate was not remarkably affected by altering the
angle of the nozzles. Differences reported in Tables 1 and 2 which
might be attributed to angle differences might be attributed equally well
to differences in raking speeds .

Harvesting Rate

When rated on the basis of catch per minute of fishing, the rake
performed best when equipped with 2-inch nozzles and fished at an
average speed of 25 feet per minute.

Collection raking of 2-foot-wide, 25-foot lengths of track
hydraulically raked at this speed occupied 4 to 5 minutes in addition

-25-

to the one minute of hydraulic raking . Thus the time for complete
hydraulic harvesting is 5 to 6 minutes per 50 square feet and the rate
is 8 to 10 square feet per minute. Presumably this rate would change
with clam population density and bottom character because collection
raking slows down with increases in the number of clams to be gathered,
and with increases in the roiliness of water when collection raking is
carried on over muddy bottoms .

Harvesting Period

The efficiency of hydraulic harvesting depends partly on the
length of time during which the hydraulic rake and collection rake can
be operated . Hydraulic raking can be carried on only when the flats
are covered with water 8 inches to 3 feet deep. Collection raking has
about the same depth limitations . Clam Harbour is typical of many of
the tidal areas where clam fishing is carried out and there the flats can
be worked steadily for about 2 1/2 hours on falling tides . One starts
on the highest beds and works down to the lowest as the tide falls . By
reversing the order, it is possible to work for 2 1/2 hours on rising
tides . In June and July the days are long enough to permit working two
tides a day if these occur in daylight hours . Generally, however, it is
impossible to rake for more than 5 hours per day.

Breakage

Breakage is an important factor in harvesting efficiency because
it affects the usefulness of the catch (Medcof and MacPhail, 1952).
Repeated observations of clams on the surface following hydraulic
raking with 2-inch nozzles but preceding collection raking showed that
there was virtually no breakage of marketable clams. However, in
gathering the clams under water with the collecting rake, 3 .5% were
broken (Tables 1 and 2). This was caused mostly by the raker walking
on and crushing some of the clams that had been washed up on the
verges of the track . He cannot always see these when the water is
murky from collection raking.

With longer nozzles breakage rates were as high as 11.3%, an
increase of almost 8% beyond the 3 .5% breakage which we attribute to
collection raking. This increase is apparently caused by the forward
pressure of the metal nozzles crushing the clams while they are still in
their burrows .

-26-

EFFECTS ON UNDERSIZE STOCKS

Formalin-killed Stocks

After plots 25 and 26 were hydraulically raked, they were
examined at the following low tide. (There was no collection raking .)
Most of the tagged animals were found on the soft soil of the track
within 10 feet of their plots (Table 3). From this and general observa-
tions during fishing, it appears that the hydraulic rake washes clams
only very short distances and carries very few onto the verges where
the soil is firm. All the clams that were more than 10 feet from the
plot were found down -tide where they had apparently been carried by
the vigorous currents flowing through the rake tracks .

The total recoveries averaged 78%. Most of the 22% loss is
attributable to the fact that many of these dead clams had bubbles of
air within their shells. As soon as they were fished they floated to
the surface of the water and were carried away by the tide .

Living Stocks

After hydraulic raking almost all the living, tagged, undersized
clams dug back into the soil before the flats were exposed by the falling
tide . We watched them burrowing . And low-tide examination of plots
27 and 28 showed that the rake does not scatter clams to any great
extent (Table 4). Almost all that were recovered were found in the
tracks and within 10 feet of the plots . Recoveries of these buried clams
were low because we had no means of fishing them from the track
except by probing with the hands which is most inefficient. Inspection
at subsequent low tides showed siphon holes in the sand indicating that
we had missed large numbers . Nevertheless, the distribution of the
siphon holes was essentially the same as that of the clams that we had
recovered .

A few (5.5%) small living clams were found on the verges of
the track or on the undisturbed soil down-tide from the plots where they
presumably would have died. At first we regarded them as victims of
hydraulic raking and took 5.5% as an estimate of indirect fishing mor-
tality . However, this value seemed high compared with that for large
clams (Tables 1 and 2), especially since there had been no collection
raking of plots 27 and 28. Furthermore, the 5.5% had intact shells and
no signs of mechanical damage. Indeed, none of the recovered clams
were broken (Table 4). We therefore suspected that some of the 5.5%
were sick before they were raked from the plots and incapable of bur-
rowing afterwards. This suspicion was strengthened by finding five of

-27-

our tagged clams, dead but intact and lying on the surface in the track
after fishing (Table 4). We assumed that these had been killed in our
handling, tagging and planting operations.

Eventually we decided that 5.5% was a likely over-estimate of
indirect fishing mortality of young clams and that until a better estimate
is available we should use the same value as for market-size animals,
i.e., <5% (Tables 1 and 2). It seems safe to say therefore (Table 5)
that with careful collection raking, about 95% of small clams left behind
by hydraulic rakes survive . They are able to burrow back into the soil
before the flats are exposed by the falling tide . They are not exposed
to air or sun or to gull predation . They are brought to the surface by
floatation. None are smothered by deep burial.

Table 5 . Comparison of the hydraulic rake equipped with 2-inch noz-
zles, the hydraulic escalator harvester and the clam hack.

Characteristic

Hydraulic
rake

Hydraulic
escalator
harvester

Clam
hack

Harvesting rate
(sq ft per man per minute)

Fishing time
(hours per tide)

25*
5

5
5

Recovery rate
(% of stock recovered from
ground worked)

90-95

95

60

Ratio of catch per man per
minute

3

9

1

Breakage
(% marketable stock)

<5

< 1

10-15

Destruction of undersized
stock left in soil (%)

<5

7

50

Cost $
(exclusive of boat)

<500

5,000

< 5

* The escalator harvester covers 50 ft2 per minute (MacPhail, 1961)
but requires two men to operate it.

-28-

COMPARISON OF HYDRAULIC RAKE WITH ESCALATOR
HARVESTER AND CLAM HACK

The rate of complete hydraulic harvesting, which involves both
hydraulic raking with 2-inch nozzles and collection raking, is 8 to 10
square feet of flat per minute (Table 5). This is about double the rate
for clam hacks (FRB, 1952) and about one-third the rate (per man) for
hydraulic escalator harvesters (MacPhail, 1961).

It was pointed out above that hydraulic raking occupies roughly
20% of the total harvesting time, 80% being required for collection
raking. When a man works alone, the hydraulic rake is therefore idle
most of the time. Obviously, it could be used more effectively by team
work . One man working steadily with a hydraulic rake should be able
to keep 3 or 4 collection rakers busy. Presumably a team of 4 to 5 per-
sons could amortize the cost of a rake (Table 5) very quickly — probably
in a single season— because of the great increase in their harvesting
capacity. Amortizing the much greater costs of an escalator harvester
would be too challenging for most of our clam men even though it also
involves team work (boat crews of 2 or 3). The clam hack, in contrast,
is a strictly one-man tool and so inexpensive that almost anyone can
undertake hack-fishing for clams .

The number of hours per tide during which the hydraulic rake can
be used is about the same as that for either of the other harvesting
gears (Table 5) but the periods during tidal cycles when fishing can be
done are different in each case. The hack fisherman requires a dry
beach to work on, the escalator harvester requires water 2 to 6 feet
deep (deep enough to float a moderate-sized boat), while the hydraulic
rake fisherman must work in water 8 inches to 3 feet deep.

The escalator harvester and the hydraulic rake recover practi-
cally all the marketable clams from the ground they cover but the hack
fisherman recovers only 6 0% (FRB, 1952). When both harvesting rate
and recovery rate are taken into account the ratios of catches per man
per minute on intertidal clam flats by the escalator harvester, the
hydraulic rake and the clam hack are 9:3:1. By abandoning clam hacks
and adopting hydraulic rakes, teams of fishermen should be able to
triple their earning power .

The breakage rate of the clams harvested is < 1% for the escala-
tor harvester (Dickie and MacPhail, 1957), <5% for the hydraulic rake
and 10-15% for the clam hack (Medcof and MacPhail, 1952). The first
two fishing devices produce clams that have no mud on them. On both
these counts processors would probably prefer these clams to those
harvested with hacks .

-29-

The destruction of both adult and undersized clams left behind
in the fished -over ground is about 5% for the hydraulic rake, 7% for the
escalator harvester (Medcof, 1961) and 50% for the clam hack (Needier
and Irgalls, 1944). In other words indirect fishing mortality could
probably be reduced by 9 0% by substituting hydraulic rakes for hacks
in commercial fishing. Presumably this reduction would effect a sub-
stantial increase in the annual production per acre of our clam grounds .

In practice, the reduction in indirect fishing mortality would
likely be even greater than 9 0% because hack fishermen harvest only
60% of the marketable stock from the ground they turn. And the sur-
vivors of the 40% of the stock they leave behind frequently attract them
back to redig the same ground soon after. At each digging there is a
50% destruction of the stock left behind or a total destruction of approxi-
mately 75% after two diggings of the same ground . When hydraulic rakes
are used the harvesting is so complete that there should be no inducement
to refish any ground until the young stock has grown to commercial size.

The high recoveries of the marketable stock from grounds that are
fished might also make it worthwhile for hydraulic rakers to harvest flats
that are too sparsely stocked with clams to be attractive to hack fisher-
men. If this be true, we might assume that the use of hydraulic rakes in
place of hacks should insure fuller use of our clam resources because
many of our flats are poorly stocked .

CONSERVATION ASPECTS

Primarily we regard the rake as a savings gear because of the
slight damage it does to undersized stocks . Our experiments suggest
that in this and in other ways it is a close approach to being the ideal
harvesting device for clam stocks on sandy, intertidal beaches. But it
can be destructive. For instance, if it is left upright and stationary
for any length of time with the jets running, a hydraulic rake will dig
a "well" in a clam flat that will trap and smother small clams. And
careless collection rakers may not always return all undersized clams
to the soft rake track where they can dig in quickly. Some of these
small clams may then be killed. However, even the most flagrant
abuse of this sort could scarcely effect the havoc among small clams
that is daily caused by innocent-looking clam hacks . And it should be
possible to avoid these risks by simple precautions. Full advantage of
the rake's special features could then be realized.

If conservationists were convinced of the preferability of hydrau-
lic rakes to clam hacks it should not be hard to persuade clam fishermen
io adopt them in areas where they can be used . Their costs are not

-3 0-

great (Table 5) and could be quickly amortized; they would triple
fishermen's earning powers and they would take the drudgery out of
clam fishing .

REFERENCES

Dickie, L . M . , and J . S. MacPhail. 1957. An experimental mechani-
cal shellfish-digger. Fish. Res. Bd . Canada, Atlantic Prog.
Rept. No. 66:3-9 .

FRB, 1952. Annual Report of the Fisheries Research Board of Canada
for the year 1951:19-21, Ottawa.

MacPhail, J. S. 1961. A hydraulic escalator harvester. Bull. Fish.
Res. Bd. Canada, No. 128. 24 p.

MacPhail, J. S. 1961b. 1961. Building and testing a hydraulic clam
rake. Fish. Res. Bd . Canada, MS Rept. (Biol.), No. 711.
10 p.

MacPhail, J. S. MS, 1961. Building and testing a hydraulic clam
rake . Fish . Res . Bd . Canada, MS Rept. (Biol .), No. 711,
10 p.

Medcof, J. C. 1950. Burrowing habits and movements of soft-shelled
clams. Fish. Res. Bd . Canada, Atlantic Prog. Rept. No. 50:
17-22.

Medcof, J. C. 1961. Effect of hydraulic escalator harvester on under-
size soft-shell clams . 1959 Proc . Nat'l Shellfish Assoc .
50: 151-161.

Medcof , J . C ., and J . S . MacPhail . 1952. Breakage— the bug bear in
clam handling. Fish. Res. Bd . Canada, Atlantic Prog. Rept.
No. 54: 19-25.

Needier, A. W. H., and R. A.Ingalls. 1944. Experiments in the
production of soft-shelled clams (Mya). Fish. Res. Bd .
Canada, Atlantic Prog. Rept. No. 35:3-8.

-31-

CAUSES OF MORTALITY OF THE SEA SCALLOP,
PLACOPECTEN MAGELLANICUS

J. C. Medcof and Neil Bourne

Fisheries Research Board of Canada
Biological Station, St. Andrews, N. B.

ABSTRACT

Causes of natural mortality include summer water temperatures too
low for spawning or for larval development, flushing of basins by "tropic
tides," lethal saltations in summer water temperature, predators, and
shell pests. Mass mortalities due to pathogenic micro-organisms are not
known. Causes of fishing mortality include bottom damage by dragging,
damage by turbulence in drags, dumping on deck, culling, shovelling,
air exposure, and shucking. Fouling of beds by discarded rims, and
pressure changes, probably are not causes of mortality. Natural mortality
has been estimated as 10% for adult scallops but there is no reliable
figure. Present methods destroy 10% of discards (scallops returned to
bottom) off Digby, Nova Scotia, and 2 to 20% on Georges Bank due to
practices resulting in long air exposure and much mechanical damage.
Dickie (1955) estimated that 20% of the scallops off Digby, N. S., were
removed each year by fishing (dragging). There is no satisfactory esti-
mate of direct fishing mortality for Georges Bank.

INTRODUCTION

Since 1952 the sea scallop has been eastern Canada's most
valuable commercial mollusc. Landings of shucked meats (adductor
muscles only) have increased from 0.8 to 10.8 million lb (10.1 million
from Georges Bank) in 1961. In spite of the g-eat and growing impor-
tance of this fishery there is little knowledge of how scallop stocks
are reacting to it or how fishing practices might be modified to insure
wisest use of the stocks . What is known is widely scattered in pub-
lished literature or unpublished, and there are many misconceptions
about the causes of mortality.

We wish to thank L. R. Day, of this station, for critical reviews
and many helpful suggestions in the composition of this paper. It sum-
marizes the information on mortality for use by those who are coping
with management problems . It also reports a study of damage to scal-
lops returned to the beds (discards). This is an important source of
fishing mortality .

-33-

NATURAL MORTALITY

Natural mortality involves all ages and sizes of scallops and
affects both exploited and unexploited stocks . So far there are no good
methods of sampling scallops that are less than three years old or of
measuring their natural mortality rates. Nevertheless, there is evidence
of at least seven causes of natural mortality involving both the youngest
and oldest scallops. When researchers learn more about the larvae and
early post-settlement stages of the sea scallop as Yamamoto (1960) has
about Patinopecten yessoensis Jay, it should be possible to identify
other causes and judge their importance.

Causes of Natural Mortality

1 . Low summer water temperature that fails to reach the spawn-
ing threshold . This probably results in destruction of gametes as
Yamamoto (1950) points out for Patinopecten yessoensis and as has
been established for Crassostrea virginica (Loosanoff, 1942). Com-
plete failure of a year-class can result, as evidently happens in some
parts of the Gulf of St. Lawrence (Dickie and Maclnnes, MS, 19 58).

2. Low summer water temperature that delays larval development
After spawning, low temperature may delay development of sea scallop
larvae as it does oyster larvae (Medcof, 1939). This would prolong the
period during which larvae are exposed to predators, increase larval
mortality and decrease recruitment. There is evidence that this happens
in the Bay of Fundy (Dickie, 195 5).

3. Flushing of basins . Periodic heavy flushing of basins
affected by "tropic tides" (Sverdrup et al., 1942) sweeps oyster larvae
out to sea (Medcof, MS, 194 0) where they are lost and, in extreme
cases, virtually eliminates recruitment of whole year-classes. This
seems most likely to happen when larval development is protracted by
low summer water temperatures . And there is evidence that this does
indeed affect recruitment of Bay of Fundy stocks of sea scallops
(Dickie, 19 55).

4. Lethal saltations in summer water temperature. These may
produce mass mortalities among sea scallops and eliminate whole fish-
eries in the Gulf of St. Lawrence (Dickie, 19 58; Dickie and Medcof,

in press).

5. Changes in water temperature. These may debilitate scal-
lops (Dickie, 19 58) and leave them unusually susceptible to heavy
predation. This apparently happens rather frequently in the Gulf of St.
Lawrence (Dickie and Medcof, in press).

-34-

6 . Predators . Sea scallops are found in the stomachs of cod
(Gadus callarias), American plaice (Hippoglossoides platessoides) and
wolf fish (Anarhichas lupus) . Apparently these fish take a steady toll
of the stocks. Chiasson (MS, 1951) found temporary concentrations of
a starfish (Asterias vulgaris) responsible for spectacular rises in mor-
tality of scallops of market size (> 80 mm) in the Gulf of St . Lawrence .
In January 1962, observations in the 3ay of Fundy showed that smaller
starfish (Crossaster papposus) , 6 to 13 cm in diameter, ingest whole
young scallops, 14 to 18 mm in length.

7. Parasites and shell pests. Boring sponges (Cliona vastifica)
are damaging to sea scallops in the Bay of Fundy (Medcof, 1949 ; War-
burton, 1958), and shell worms such as Polydora (Kinoshita, 1939) and
Ceratonerus (Wells and Wells, 1962) may kill or weaken other species
of scallops and leave them susceptible to predators . Similar worms
affect sea scallops and are believed to be damaging although they have
not been studied. A hydroid (Hydractinia echinata) frequently grows
on the upper valve of the sea scallop (Merrill, personal communication).
Sometimes it overgrows the lip and interferes with the mantle, causing
shell distortion and even death of the scallop.

There is no evidence of mass mortality from micro-parasites as in
oysters (Logie, 1956). But the flagellate, Hexamita , has been found
in dying scallops held in aquaria (Medcof, 1961). Dr. M. Laird, for-
merly of the Institute of Parasitology, McGill University, is currently
describing ciliate commensals collected from recently-fished scallops
and suggests (personal communication) that under adverse environmental
circumstances these (e.g., Trichodina) may harm their hosts.

FISHING MORTALITY

Fishing mortality includes the shucking of scallops for market
(direct fishing mortality) and other killings that take place incidental
to capture (indirect fishing mortality). Fishing mortality occurs only
in exploited stocks but involves animals of all sizes and affects
recruitment .

In this paper seven significant causes of fishing mortality are
identified and two putative causes discounted .

Causes of Fishing Mortality

1. Bottom damage by drags . Yamamoto (1960) showed that
drags churn up soft bottoms in Mutsu Bay and help create anaerobic

-35-

and turbid conditions that kill whatever scallops (Patinopecten yessoensis)
are present. He showed that dragging reduced the settlement of scallop
spat on soft bottoms for the same reasons . Both effects cut down the
yields from scallop beds .

In 1950, one of our staff, Mr. MacPhail, reported finding many
dead and dying sea scallops on soft bottoms in Maces Bay, New Bruns-
wick. These were packed full of mud. Still others were shorn in two.
The only reasonable explanation was that these had been forced into the
mud by drags which had previously passed over them. Similarly Merrill
(1960) found weakened sea scallops on soft bottom on Georges Bank.
They had mud and sand packed between their valves or between their
valves and mantles . He believes this is caused by drags .

On hard bottom many scallops must be mechanically injured by
drags passing over them but this has not been demonstrated .

2. Damage in drags. Underwater television has shown that the
contents of Digby-type scallop drags (MacPhail, 1954) are in such con-
stant and violent turmoil during the haul that scallops must be mechanically
damaged or smothered . Many of the damaged animals in Digby catches
may be injured in this way and many of the small ones (discards) probably
die when they are returned to the beds .

Tremendous mechanical pressure is produced on scallops in the
drag, especially when it is hoisted through the air and no longer buoyed
up by water. This pressure apparently crushes scallops and produces
"bullet holes" (see comparison of damage to discards in Digby and
Georges Bank fisheries). When the drag slaps against the side of the
boat still more are crushed.

3. Boarding and dumping. After hoisting, the full drags (Georges
Bank drags with their catches weigh up to 3 tons) are suddenly dropped
to the deck from a height of 4 to 6 feet. This, plus the subsequent
dumping of the drag contents, including rocks, onto the deck by hoisting
the free end of the bag to 6 or 8 feet, damages many scallops. Scallops
lying on deck from previous hauls are also damaged in this process .

4. Culling . The rough handling and trampling involved in cul-
ling breaks some scallops.

5 . Shovelling . In the Georges Bank fishery discards are shovelled
overboard . This may cause mechanical damage but it is probably slight
compared with that from the sources mentioned above. In the Digby fish-
ery there is no damage from shovelling because the drags are emptied

-36-

onto a "dumping board," a false deck that is hinged to the gunwale.
When the catch has been culled, the dumping board is tilted and the
trash and discards slide overboard .

In winter fishing, scallops often freeze on deck and the rough
handling involved in shovelling may then be lethal even when it is not
violent enough to cause mechanical damage (Marshall, 1960; Friedman,
1933).

6 . Air exposure . Air exposure on deck may have little adverse
effect in cool weather but in hot weather even 2 or 3 hours can be
lethal to scallops as our tagging experiments have shown (Dickie, 1955).

7. Shucking. The foregoing six causes contribute to indirect
fishing mortality and affect recruitment . They involve scallops of all
sizes including some that are not caught in the drags and some that are
caught but discarded . In contrast, shucking is the only cause of direct
fishing mortality. It involves only the marketable scallops (>95 mm)
that are caught. The Digby and Georges Bank fisheries are so intense
that the numbers of scallops shucked have an important effect on abun-
dance and therefore affect yields to the fishery.

Supposed Causes of Fishing Mortality

Two putative causes of indirect fishing mortality are frequently
claimed by fishermen and conservationists and should be mentioned
here .

1 . Cluttering and fouling of beds . Many people believe that
dumping empty scallop shells and "rims" (waste body parts) damages
scallop beds. In Canada this practice was forbidden by law until as
late as 1950 (Anon, 1950) and still is forbidden in Australia (A. M.
Olsen, personal communication). Many believe that, for survival,
scallops require clean, uncrowded bottom with few shells and that
waste meats "sour" the bottom and kill or sicken healthy scallops.

Underwater photography and tagging (Dickie, 19 5 5; Posgay,
1953) indicate that empty shells and live scallops are too widely
scattered on the bottom to cause crowding. There is no evidence that
scallop rims accumulate on the bottom. They are found in catches but
those that are found appear to be freshly cut. Bottom long-line fisher-
men in the Bay of Fundy rate scallop rims as first-class bait for haddock
and state that the stomachs of fish taken in areas where scallop fishing
is going on are filled with rims. Furthermore, fish trawler captains,

•37-

both in the Bay of Fundy and on Georges Bank, regularly sweep grounds
that scallop boats have just dragged over. They say the best fish
catches are to be had there. These observations bear out the sugges-
tion of Choat (196 0) that groundfish keep the bottom clean by feeding on
rims almost as fast as they are shucked. There is no sound evidence
for the claim that dumping the refuse of shucking damages scallop beds
in any way .

2- Pressure changes. Some scallop fishermen argue that pres-
sure changes which occur when scallops are brought to the surface and
returned to the bottom either kill scallops or weaken them to such an
extent that they are easy prey to their natural enemies . Qasim and
Knight -Jones (1957) have shown that pressure changes affect some
aquatic invertebrates . However, high returns of tagged scallops in
vigoroas condition (Dickie, 19 55) suggest that such changes do not
cause scallop mortalities .

MEASUREMENTS OF MORTALITY

Natural Mortality

Dickie (1955) describes a method of measuring the annual rate
of natural mortality of market-size scallops from the abundance of
"cluckers" (empty shells still attached at the hinge). His estimate of
the rate was 10% for stocks of adult scallops off Digby, Nova Scotia,
and it seems to average about the same for Georges Bank. However,
Dickie's method has shortcomings and there is no method of measuring
natural mortality rates among small scallops .

Fishing Mortality

Direct fishing mortality. From returns of tagged scallops
released on beds off Digby, Nova Scotia, Dickie (1955) reports a 20%
direct fishing mortality, i.e., the proportion of the usable stock that is
caught and shucked annually. So far there is no satisfactory estimate
of direct fishing mortality for Georges Bank .

Indirect fishing mortality. No one has yet estimated indirect
fishing mortality. In 19 52, however, we did estimate certain of its
components in the Digby, Nova Scotia, fishery.

-38-

DAMAGE TO DISCARDS IN THE DIGBY FISHERY

Damage to discards is an important source of fishing mortality
in fisheries for many species. Conservationists and designers of
savings gear try to reduce this damage and thus increase the number of
young that reach "optimum" size. Data on the frequency of lethal dam-
age to discards were gathered during our studies of the Digby fishery
(Medcof, 1952; MacFhail, 1954). These are presented here because
both industry and the International Commission for the Northwest Atlan-
tic Fisheries are now interested in a savings gear for the Georges Bank
scallop fishery .

Basis for Study

We examined living scallops to judge the significance of dam-
age to discards. Many were "self-repaired." Those that had lost
marginal fragments, about 5 mm wide, were common (Fig. 1A) . In others,
part of an ear had been broken away and regenerated (Fig. IB). Still
others had lost wider fragments but none extended centrad beyond the
pallial line (Fig. 1C). Less common types of repaired breakage included
shell punctures in the region marginal to the pallial line (Fig. 2A) and

Fig. 1. Repairs to slight marginal damage to arched, left (upper)
valves (outer faces showing). A. Ventral fragment was broken away.
B. Anterior ear of hinge was broken off. C. A ribbon of shell from ventral
and anterior margins was broken off.

■39-

damage to the shell margin and probably the mantle too. The latter is
thought to be responsible for subsequent distorted growth (Figs. 2B and
2C). Most of this damage is attributed to the fishery and if sea scal-
lops react like bay scallops, these types of damage are of little con-
sequence (Belding, 1931). Merrill (1960) believes that repair of damage
like that shown in Figs . 1 and 2 is carried out in a short time, possibly
in a few days, just as oysters quickly repair similar damage without
appreciable effects on growth rate or general well being (Loosanoff and
Nomejko, 1955).

Fig. 2. Repairs to more severe marginal damage (outer faces of
valves showing). A. Puncture of posterior -ventral margin of left valve.
B. Large fragment from posterior -ventral margin of left valve. C. Large
fragment from ventral and from posterior margin of flat, right (lower)
valve .

Some scallops had repaired more extensive breaks, e.g., cracks
reaching beyond the pallial line in the posterior, ventral and anterior
parts of shells. In all cases, however, the broken shell segments had
remained in place and were neatly joined by inner layers of lime which
bridged the cracks (Figs. 3A, 3B and 3C). All these self-repaired scal-
lops seemed vigorous. Accordingly, in the study of discards, all

-40-

Fig. 3. Repairs to extensive damage possible only when most
broken fragments stay in place (outer faces of shells showing). A.
Whole posterior -dorsal section of right valve crushed, including ear of
hinge. B. Deep anterior-ventral fracture of a left valve. C. Loss of
a ventral fragment and deep anterior -ventral and posterior-ventral frac-
tures of right valve .

scallops showing breakage of the kinds described in Figs. 1-3 were
classed as "undamaged" along with those whose shells were intact.
However, we found no evidence that scallops can repair:

1. loss of wide marginal segments of shell (extending centrad
beyond the pallial line) (Fig. 4, lower row);

2. shell cracks extending all the way across the valve in the
region between the hinge and the muscle scar (Fig. 4,
upper row center);

3 . damage to the hinge in the region of the hinge ligament;
4. obvious damage to internal organs (Fig. 4, middle row and
upper left) .

-41-

Fig. 4. Lethally-damaged discards showing various types of
injury to shells and soft parts (see text).

Procedure

All the discards were collected from the dumping board just
before it was to be tilted after each haul. They were then counted and
classified, according to the condition of their shells, as undamaged
or as lethally damaged. In some cases they were measured (Table 1).
Scallops with intact shells but showing evidence of serious internal
injuries were also classified as lethally damaged .

Records were also made of the amount and sizes of rocks in the

hauls

-42-

Table 1

Frequency of lethal damage to discards left by cullers on the
dumping board of a Digby boat fishing various beds, April
and May 19 52. In some hauls all seven of the drags in the
gang had the same size mesh, either standard or savings.
In other hauls the mesh varied from drag to drag .

Rocks in

rotal

Lethally damaged

Bed

dredge d

iscards

discards

No.

No.

%

Hauls with standard mesh

Gullivers 2nd Ridge

Many

605

138

23

Gullivers 2nd Ridge

Few

201

32

16

Gullivers Inner Ridge

Few

438

73

17

Gullivers Inner Ridge

Few

497

68

14

Hour Ground

Few & small

560

65

12

Hour Ground3

Few & small

78

33

42

Broad Cove

Rare

624

70

11

Hauls with mixed gear— standard and savings mesh

Gullivers Inner Ridge

Few & small

228

37

16

Hour Ground

Few & small

712

102

14

Hauls with savings mesh

Hour Ground

Few

150

22

15

Gullivers 2nd Ridge

Many &

large

54

14

26

Discards were trampled by crew while gear was changed

Results

The number of discards per haul varied considerably depending
on the mesh size of the drags but averaged more than 50% of the total
catch. Approximately 95% of these discards ranged from 65 to 80 mm
in height. About 5% ranged from 8 0 to 100 mm.

The frequency of lethal damage among the discards varied from
11% to 42% and averaged about 15% (Table 1). It was assumed that the
"undamaged" 85% survived when returned to the beds.

-43-

Discussion of Digby Results

The Digby study measured the pooled effects of three of the
seven causes of fishing mortality (causes 2 to 4 in above list) as they
operate on discards. Its main weaknesses are that the criteria used to
distinguish lethally and non-lethally damaged scallops have not been
tested experimentally and that a two-category classification is too
gross for high precision. Even though slightly damaged scallops
(counted here as undamaged) can recover, some are almost certain to
be killed by predators soon after being returned to their beds . And
Chiasson (personal communication) has observed that damaged scallops
are more readily killed by starfish than undamaged. Furthermore, lethal
internal damage is sometimes hard to identify. Thus, our study probably
provides minimal estimates of mortality of discards due to mechanical
damage in the Digby fishery.

This study dealt only with mechanical damage as a source of
mortality and neglected air exposure (fishing mortality cause 6 in above
list). This is justifiable because, in the Digby fishery, discards from
one tow are dumped before the next is boarded and the tows are seldom
longer than 25 minutes. The intact animals are still lively after this
period of exposure .

The 15% mortality estimate for Digby discards cannot be analyzed
because it appears to be the pooled effect of three causes . However,
the relative importance of these causes can be inferred. The table indi-
cates that the percentage of lethally damaged discards varies directly
with the quantity and size of rocks in the drag and with the amount of
trampling . The effect of extra trampling was clearly demonstrated fax-
one Hour Ground haul. The lethal damage was tripled— 42% compared
with 12% to 15% for samples from the same ground which received only
the normal amount of trampling that is involved in culling .

No relationship was observed between the percentage of dis-
cards damaged and the size of the catch. Any correlation which might
have existed was probably masked by other causes of mortality, pri-
marily by the amount and size of rock in the catch.

■44-

COMPARISON OF DAMAGE TO DISCARDS IN DIGBY
AND GEORGES BANK FISHERIES

Mechanical Damage

In studies of mechanical damage to discards, the "bullet holes,"
a type of shell damage not seen at Digby, was frequently encountered in
the Georges Bank fishery. It is a single small hole about 3 to 4 mm in
diameter piercing the upper shell only where it is most arched and at
about the center of the adductor muscle scar. The margin of the hole is
quite even on the outside (Figs . 5A and 5B) but flaringly splintered on
the inside (Fig. 5C) like a bullet hole through a pane of glass. Fine
shell fragments are often found in the hole, some of them still attached
to the ends of muscle fibers . Sometimes there is gravel there too.

Fig. 5. "Bullet holes," a kind of lethal damage found in left
valves of Georges Bank scallops . The rim of the hole is clean and
even on the outer face of the shell (A and B) but flaring and splintered
on the inner face (C).

-45-

Facsimiles of bullet holes have been produced with a drill press
by applying pressure on gravel pellets laid on the outside of intact
scallops. It is believed that most bullet holes are produced while scal-
lops are in the drags, during hauling, hoisting or dumping, from extreme
pressure applied over a small area — perhaps by drag links or pieces of
coarse gravel. Because most scallops with this type of damage are dead
or dying when caught, it has been tentatively classified as a cause of
indirect fishing mortality of discards (fishing mortality, cause 2 or 3).
In studies of Georges Bank discards, those with bullet holes were
classed as lethally damaged .

Data are lacking for a good comparison of the lethal mechanical
damage to discards on Georges Bank and the average of 15% obtained
at Digby in 1952. However, two 1959 observations on Canadian boats
fishing Georges Bank and involving approximately 1,000 discards gave
values of only 2% to 5%. At these times the proportion of discards in
the catches was approximately the same in both fisheries . However,
there were few rocks in the Georges Bank catches and this probably
accounts for the lower rates of damage .

Recently, Canadian fishing practices have changed. Until the
latter part of 19 59, crews were able to shuck scallops as fast as they
were fished. Thus, the deck was cleared of scallops taken in one haul
before the next was dumped, as is the case in the Digby fishery.

Late in 19 59, catches suddenly increased, shuckers were no
longer able to clear the decks between hauls and the practice of "deck
loading" began. In this practice haul after haul is dumped for 2 or 3
hours or until this becomes awkward or dangerous to continue . By that
stage the catch is often 4 feet deep on deck . The boats then anchor and
the crew shucks for as long as 6 hours until the catch is worked through.
By the time the last discards from a deck load are shovelled overboard
they have been out of water for 8 or 9 hours . This practice of deck
loading still continues, although not to the extent it did in 1960 since
most boats have taken on extra crew to increase their shucking power.

General observations, and the few data gathered to date, indi-
cate that deck loading has increased indirect fishing mortality that
arises from mechanical damage. In May 1961, four samples of discards
from two different parts of Georges Bank showed lethal damage varying
about 10% which is two to five times that observed in 19 59. This is
understandable. During deck loading, the heavy drag lands time after
time on the previous catch on deck, crushes and cracks many scallops
and the crews continually walk back and forth over the catch while they
handle the gear (fishing mortality, causes 3 and 4).

■46-

Damage from Air Exposure

In the Digby fishery there is probably little damage to discards
from air exposure as already mentioned and before the practice of deck
loading began there was probably little in the Georges Bank fishery.
However, general observations suggest that there has been an important
increase since then. This increase is not easy to measure but it appears
to be proportionately much greater than the corresponding increase due to
mechanical damage. During our 1960 summer observations on the Georges
Bank fishery air temperatures were high and the long hours of air exposure
apparently killed many discards. Large numbers (roughly estimated at
10%) were gaping and badly desiccated when they were shovelled through
the scuppers. Records of tagging mortality at Digby suggest that the
Georges Bank discards, whose shells were intact, were already dying.
The total lethal damage to Georges Bank discards (mechanical damage
and air exposure combined) was estimated to reach 20% in the summer of
1960, when deck loading was practiced. This is even higher than that
for Digby in 1952 where deck loading was not practiced and when mechani-
cal damage accounted for all the lethal damage .

In winter fishing, scallops are often frozen so stiff that they
cannot be shucked and presumably many discards are in this condition
when they are shovelled overboard . Preliminary experiments carried
out at the Biological Station, St. Andrews, New Brunswick, show that
sea scallops can withstand up to 3 hours air exposure at temperatures
between 0°and 5°C. They also show that sea scallops can survive
freezing, provided they are not handled roughly. Shovelling, which is
practiced only on Georges Bank, is rough treatment and some winter
discards presumably die from it, perhaps as many as the 10% which may
die from air exposure in summer during deck loading .

The actual numbers of Georges Bank discards killed by mechani-
cal damage or air exposure were probably higher in late 1959 than in
196 0 because the population structure changed in the interval and the
per cent of discards in the catch dropped from 60% to less than 20%.
By May 1961, however, the proportion of discards had risen again and
the number lethally damaged probably increased too.

From these comparisons it is ciear that damage to discards is
an important source of fishing mortality in all scallop fisheries and
should therefore be taken into account in forecasting recruitment and
in designing savings gear.

-47-

CONCLUSIONS

This preliminary survey of causes of mortality in scallop stocks
illustrates the complexity of the subject. It shows that estimates of the
several components of total mortality must be assessed and re-assessed
with the greatest of care as fishing gear and fishing practices change.
Otherwise, they could be misleading when used in mathematical models
for predicting trends in the fishery and effects of regulatory measures on
yield. If there is a remedy for this situation, it will emerge from care-
fully planned, long-term biological studies.

REFERENCES

Anon. 1950. Amendment, Special Fishery Regulations for the Province
of New Brunswick . Canada Gazette (Part II) of Wednesday,
March 8, 1950, P. C. 808.

Belding, D. L. 1931. The scallop fishery of Massachusetts . Common-
wealth of Massachusetts, Dept . Conservation, Div. Fish and
Game, Marine Fish. Series No. 3. 51 p.

Chiasson, L. P. MS, 1951. Scallop investigations and explorations
in the southern Gulf of St . Lawrence — 1950 . Fish . Res . Bd .
Canada, MS Rept. Biol. Sta . No. 423. 43 p.

Choat, J. H. 1960. Scallop investigation, Tasman Bay 1959-60. New
Zealand Marine Dept., Fisheries Tech. Rept. No. 2. 51 p.

Dickie, L . M . 1955 . Fluctuations in abundance of the giant scallop,
Placopecten magellanicus (Gmelin), in the Digby area of the
BayofFundy. J . Fish . Res . Bd . Canada 12:797-857.

Dickie, L . M . 1958 . Effects of high temperature on survival of the
giant scallop. J . Fish . Res . Bd . Canada 15:1189-1211.

Dickie, L. M ., and C . D . Maclnnes . MS, 1958. Gulf of St . Lawrence
scallop explorations — 1957 . Fish. Res. Bd . Canada, MS Rept .
(Biol.) No. 650 62 p.

Dickie, L . M . , and J . C . Msdcof . 1963. Causes of mass mortalities
of scallops (Placopecten magellanicus) in the southwestern
Gulf of St. Lawrence J . Fish . Res . Bd . Canada 20:451-482.

-48-

Friedman, M. H. 1933. The freezing and cold storage of live clams and
oysters . Annual Report of Biological Board of Canada for 1932:
23-24.

Kinoshita, T. 1939 . Study on the diseased scallops produced in water
off Shari-machi, Schari-gun. Ten-day Rept. of Hokkaido Fish-
eries Exp. Sta. No. 43 8: 1-9 .

Logie, R. R. 1956. Oyster mortalities, old and new, in the Maritimes .
Fish . Res . Bd . Canada, Atlantic Prog . Rept . No. 65: 3-11 .

Loosanoff, V . L . 1942. Seasonal gonadal changes in the adult oyster,
Ostrea virginica , of Long Island Sound . Biol. Bull. 82:195-206.

Loosanoff, V. L., and C. A. Nomejko . 1955. Growth of oysters with
damaged shell-edges . Biol . Bull . 108:151-159.

MacPhail, J. S. 1954. The inshore scallop fishery of the Maritime

Provinces. Fish. Res. Bd . Canada, Atlantic Biol. Sta. Circular,
General Series, No. 22. 4 p.

Marshall, Nelson. 1960. Studies on the Niantic River, Connecticut,

with special reference to the bay scallop, Aeguipecten irradians .
Limnol . and Oceanogr. 5:86-105.

Medcof , J . C . 1939 . Larval life of the oyster (Ostrea virginica) in
Bideford River. J . Fish. Res . Bd . Canada 4: 287-3 01.

Medcof, J. C. MS, 1940. Oyster investigations . Fish. Res. Bd.
Canada, MS Rept. (Biol.) No. 184. 64 p.

Medcof, J . C . 1949 . Dark meat and the shell disease of scallops .
Fish . Res . Bd . Canada, Atlantic Prog . Rept . No. 45: 3-6 .

Medcof, J. C. 1952. Modification of drags to protect small scallops.
Fish. Res-. Bd . Canada, Atlantic Prog. Rept. No. 52:9-14.

Medcof, J. C. 1961. Trial introduction of European oysters (Ostrea
edulis) to Canadian east coast . 1959 Proc . Nat'l Shellfish .
Assoc. 50: 113-124.

Merrill, Arthur S. 1960. Living inclusions in the shell of the sea
scallop Placopecten magellanicus . Ecology 41 (2): 385-386 .

-49-

Posgay, J. A. 1953. Sea scallop investigations. Commonwealth of

Massachusetts, Dept . Nat. Res., Div. Mar. Fisheries. Sixth
Rept. on Investigations of Shellfisheries of Massachusetts,
pp 8-24.

Qasim, S. Z., and E. W. Knight-Jones . 1957. Further investigation
on the pressure responses of marine animals . Challenger
Soc. Rept. for 19 57, 3(9): 21 .

Sverdrup, H . U . , M . W . Tohnson and R . H . Fleming . 1942 . The
Oceans. Prentice-Hall , Inc., New York , 1087 p.

Warburton, Frederick E. 1958. Boring sponges, Cliona species, of
eastern Canada , with a note on the validity of C . lobata .
Canadian J . Zool . 36: 123-125 .

Wells, Harry W., and Mary J . Wells. 1962. The polychaete

Ceratonerus tridentata as a pest of the scallop Aeguipecten
gibbus. Biol . Bull . 122: 149-159 .

Yamamoto, G. 1950. Ecological note of the spawning cycle of the

scallop, Pecten yessoensis Jay, in Mutsu Bay. Science Repts
Tohoku University, 4th Ser . (Biol.) 18:477-481.

Yamamoto, G. 1960. Mortalities of the scallop during its life cycle .
Bull . Mar. Biol . Sta . , Asamushi, Tohoku Univ. 10: 149-152 .

•50-

MORTALITY OF PACIFIC OYSTERS , CRASSOSTREA GIGAS (THUNBERG),
IN VARIOUS EXPOSURE SITUATIONS IN WASHINGTON 1

Walter T. Pereyra 1

College of Fisheries

University of Washington

Seattle, Washington

ABSTRACT

Pacific oysters were maintained in baskets at various levels of
intertidal exposure and proximity to the bottom at Oyster Bay, Wash-
ington, from January to December, 1960. Mortality was low. occurring
mostly in summer. No increased mortality was caused by exposure to air
up to 40% of the time (the maximum exposure tested) during the test year.
No increased mortality was caused by biweekly mixing and handling of
experimental oysters. Subtidal oysters had significantly greater mortality,
apparently due to siltation. There was a differential mortality with
respect to size, dead oysters being slightly smaller. A starfish (Pycno-
podia helianthoides Brandt), was found extruding its stomach through
the wire mesh of subtidal baskets on several occasions in winter An
MSX-like organism was observed in one "gaper" from the float in July.
A large ciliate, probably a secondary invader, was found in most "gap-
ers."

INTRODUCTION

Although the Pacific oyster industry has occasionally had local-
ized mass mortalities (Woelke, 1961), no epidemic disease has been
reported and predation is relatively low. For float studies on yearling
Pacific oysters maintained in trays at three growing areas in Washington,

This study was conducted in conjunction with an investigation
of the growth of the Pacific oyster in various exposure situations and
was part of a thesis submitted in partial fu'fillment of the requirements
for the degree of Master of Science in Fisheries from the College of
Fisheries, University of Washington .

2
Present address: Bureau of Commercial Fisheries, Exploratory

Fishing and Gear Research Base, 2725 Montlake Boulevard E., Seattle 2,

Washington .

3
In Washington the age of the oyster begins when the seed is

planted on the beds. Hence, seed which is imported from Japan at 9

months actual age is considered age 0 when planted in Washington.

-51-

Sparks and Chew (1961) report a pooled cumulative mortality of 3.5% for
experimental oysters during a 10-month period. Woelke (1961) found the
total mortality of bed oysters in Washington to be 48.6% during the first
year after planting and 21.2% for second-year oysters . He further com-
ments on these mortalities, discussing causative agents and giving mor-
tality rates for each. Quayle (1951) studied a population of 6- to 8-year-
old oysters in trays at the 2-foot tide level in Ladysmith Harbor, British
Columbia, and recorded a 5% mortality over a 16-month period. Additional
information on Pacific oyster mortality and predation is given by Cahn
(1950), Chew (1960), Chew and Eisler (1958), Elsey (1934), Glude (1947),
Galtsoff (1929 and 1932), Kincaid (1951), Thompson (1952), and Woelke
(1955, 1957a and b).

As there is a lack of data on the effects of various levels of
exposure on mortality, this paper is presented to provide comparative
mortality rates for Pacific oysters maintained in experimental baskets
at various levels of exposure and proximity to the bottom.

METHODS

The field experiment was conducted at a commercial oyster bed
in Oyster Bay, Washington from January to December, 1960 in conjunction
with a growth study at different exposure levels (Fig. 1) (Pereyra , 1961).

The experimental oysters were part of a case of broken Pacific
oyster seed, C. gigas , from the Miyagi Prefecture of Japan which had been
used for study on the growth of Pacific oyster seed in Washington waters
(Pereyra, et al . , 19 59). The oysters were in their second year of life
(first year in Washington) at the time this study was initiated.

Twelve lots of from 100 to 106 culled single oysters were randomly
selected from a common stock and placed in experimental baskets . These
baskets, constructed of one-half-inch expanded black iron, measured
35.5 inches by 17.0 inches by 7 . 0 inches, and were coated with a com-
mercial tar-base preparation to prevent corrosion. Three baskets, two
experimental (replicates) and one control, with their complement of oysters,
were placed at each of the following four locations in Oyster Bay: (1) in
a float anchored approximately 100 feet from shore at mean lower low
water, (2) subtidal in the vicinity of the float at a depth of approximately
eight feet at mean lower low water, (3) at the +2 -foot tide level (average
yearly exposure to air approximately 10%), and (4) at the +7-foot tide
level (average yearly exposure to air approximately 4 0%). The control
baskets were included to determine if handling had an effect on the

-52-

Fig . 1 . Map of Puget Sound showing location of experimental
study area .

-53-

experimental animals . In order to keep conditions as uniform as pos-
sible, all subtidal and intertidal baskets were mounted on concrete
building blocks (Fig. 2).

Fig. 2. Experimental basket at the 7-foot level on concrete
building blocks. Note wooden bulkhead below and in front of basket
to minimize wave damage to the experimental oysters .

The experimental oysters were checked for growth and mortality
every two weeks and the controls three times during the study period .
To eliminate bias when sampling oysters for growth examination, all
oysters were mixed prior to checking . On each station check the entire
population in each basket was checked for mortality with all dead oys-
ters ("boxes") being measured and discarded. All gaping or dying
oysters were measured and the meats fixed in Zenker's fluid for later
histopathological examination. Observations were made of the associ-
ated fauna. Dissolved oxygen, pH, salinity, and temperature data
were collected biweekly at the float station.

-54-

DATA ANALYSIS

The monthly and cumulative mortality data are presented in
Table 1. Testing by chi-square showed that no significant differences
existed between the cumulative mortalities of the experimental and
control oysters from baskets in the float, at the 2-foot level, and at
the 7-foot level, but at the subtidal location the difference was highly
significant (X2 = 9.785 [ P< 0.01]).

Since the mortalities between lots of oysters in the float, at the
2-foot level, and at the 7-foot level were not significant, the lots
(experimental and control) at each of these locations were combined
and the differences between these locations tested by chi-square .
These differences were not significant at the 5% probability level.

Oysters from the above three locations were then grouped and
tested against the experimental lot that remained from the subtidal
location. The results of this test show that the difference between the
cumulative mortalities was highly significant (X2 = 27 . 26 [P< 0. 001]) .

RESULTS AND DISCUSSION

The cumulative mortality in this study was low (Fig. 3), agree-
ing well with the results from float studies in Washington (Sparks and
Chew, 1961), but being considerably lower than that reported from bed
studies (Woelke, 1961). Mortality occurred predominantly during the
warmer months .

Under the experimental conditions exposure to air was not found
to increase mortality of oysters . Oysters which were exposed to air 40%
of the time (the maximum degree of exposure studied), and mostly during
the periods of climatic extremes, experienced no increased mortality.

To eliminate bias in the growth phase of the overall study, the
oysters were mixed prior to being examined. Yet, as can be seen from
Fig. 3, biweekly mixing and handling of the experimental oysters did
not cause any increased mortality over that incurred by control oysters.
This is in accord with the results of Sparks and Chew (1961).

At the subtidal location a differential mortality was found be-
tween control oysters and those in one of the experimental baskets .
This difference is thought to have resulted from siltation of the control
group. The control basket was not handled until May, at which time
a 3 0% mortality was observed. During this same period an experi-
mental basket was handled biweekly and experienced a cumulative

-55-

o

I—-

a

c
3

CO
r+

QJ

3
O
CD
co

cr

CD

o

3
Q.

O
O

3

r+

o

•o

CD
<

CD

3

r+

CD
Q-

r+

QJ

»—-
3

3

CD
0)
CO

C
•i
CD

3

CD
3

*
*

*
*

CO
QJ
CO

CD

O
CO

cr co

CD •<

1± qj

3 D-

a

CD

o

CD

3

o
<

CD

3

o

o

r-r
O

cr

CD

CO

CD

•o

t-r

CD

3
cr

CD

•-i

>

c

c

CO

t-t-

<— 1

c
•<*

d
3
CD

2

0)

>

•o

►1

QJ
3"

-n

CD

cr
•i

c

QJ

•—<
0)

3

c

QJ
-i

O
3

r-t-

2-c

cr

cr

■-I

■1

"<

3'

CD

CD

"<

3 co

•-i

■-I

»Q £■

CD
O Q-

o

o

O

<—•

o

o

OO

o

1 — 1

o

o

o

2

O

55

— "O

•

•

•

t-l

££

t— «

o

o

r-t-

CD -g

-n

on

on

On

on

4^

►fc.

4^

1 — '

1 — »

o

o

o

K

O

I

i — •

a 3

t— •

i — •

t— *

i — »

O

o

O

o

o

o

r-t-

M|

_• «-**

•

•

3 rt

< rr

o w

2

mea:
lved

o

O

O

O

i — i

1 — 1

i — i

1 — I

o

o

o

o

o

o

o

o

o

•-s

r-t-

c35

■n

•n

• 3

■

cT

r*

ji.

*>.

►fc>

J^

4^

oo

ro

1 — '

o

o

o

o

2

O

55 £

ro

QJ

3"
QJ

o

o

o

o

O

o

o

o

o

<-h

r-r

3

r*

'

•

3-

CD

3
o

o

1

4i.
oo

1

1

'

I

1

1 — 1

o

1

1

1

o

r-t-

55

•n

rf

0)
I— ■

I

oo

r+

K

Q

■<

Cn

1

on

1

1

1

1

1

I — '

1

1

1

o

v.O C

S

00

00

o

5- g

CO
O

1 — '

s

O

3

C

1 — '

ro

00

0O

*

00

cn

co

o

o

-1

r-t-

55

CO

i-t-

jf

1

CD

a

*

CO
On

ro
4^

ro

i — •

oo

S

Q

1 — •

o

•

•

.

*

*

*

•

o

55§

3

-vl

4^

co

o

r-t-

r-t-

•

•

3"

CD

CO

►*■
N

t— ■

•

00

o

oo

00

o

o

ro
oo

o

4^
On

1 — •
1 — '

1 — I

o

to
o

2

o

•-i

55

C

cr

CD

c-t-

CO

O

1

ro

a.

it-
s'

CD

cn
o

4*.
^1

4^
^1

o

CO

o

CO

t— <
o

CO

CO

cn

co

CD

1 — '

oo
o

ro
o

2

o

r-t-

O

„ c
55 g

QJ

■a

•

o

T3

c

00

2

t— '

CO

1

ro

1

1

1

1

o

1

1

1

1

o

55

r-t-

o

CO

CD

o

r+

CO

i

3

i

oo

QJ

00

OO

00

g

o

r-t-

on

1

CO

1

1

1

1

o

1

1

1

1

o

•-1

.o C

t-r

•

•

•

55 g

3"

00

CD

o

r-f

CD

QJ

cr

CD

r1

o

C)
QJ

o

3

-56-

73
CD

3
C
+■>

C

o

0)

xi

H

■4->

a>

CD

CD

E £* o

•

•

•

3 2

1

1

1

1

CM

1

1

i

i

00

1

■^

CO

■

S5 §
2

1

1

1

1

CD
CM

1

1

i

i

o

•

1—1

1

o

•

1— 1

0)

>

1* a

O

o

o

0>

o 2

o

o

o

o

i — 1

CM

CM

t

^

^r

"■a1

Itf

+->

CM

o

1

•

o

f-

+-»

4-1

1

SS o

o

o

•

• — i

t*»

2

o

o

o

o

■—1

1—1

o

CM

o

o

o

o

Cum.

/o

Mort .

o

o

o

o

O

o

o

o

o

o

o

o

CM

CM

CO

CO

CO

00

00

oo

oo

1 — I

1

4-J

o

o

s* g

o

o

o

CM

O

1 — I

o

O

o

o

o

o

S

c

o

-.1

,

rO

o

1 S5 "g

1

I

1

1

CD

1

1

1

1

CD
i— t

i

CD
1—1

o

00

CM

■M
V? Si

3s- O
2

CD

1

1

1

1

O

1

1

1

1

^S

I

o

0)
>

i* i

o

o

o

o

o

0)

CM

u s

o

o

o

o

O

o

o

CM

CM

CM

CN]

CM

•4-*

1

o

CM

.

o

^ £

o

1

O

o

o

o

o

O

o

o

CM

o

o

o

O

CM

s

3^0

o

o

o

o

O

o

o

O

o

o

o

o

o 2

1— 1

r— t

1— 1

1— 1

CM

co

CO

CO

T

■^

■^

•^

i— <
1

1

CM

o ti

o

O

o

O

a$ o

1 — 1

o

o

o

r-i

1—1

o

o

t— I

o

o

o

2

u
0)
Xl

E

0)

■M

a

CD
CO

Si

Si

c
o

2

>1

Si

<0

c

(0

1 — 1

>1

Si

10

2

X)
0)

x:
o

Si

rO

2

r— 1

Si

a

<

to

2

CO

c

1— >

>1

i — .

4->

ca

3

Si

CO

o

*->

o

o

CD
XI

a

CD

>

o

2

to

XI

£
CD
o

a)

n

-57-

45-

40--

35-

t50--

*-

K

0 25--

520.-

X

u

Q.

15- -

3 I0f
° 5+-

■ EXPERIMENTAL
□ CONTROL

BASKET LOST

8-6-60

\

Li

F-2 F-3

FLOAT

B-l B-2 B-3

SUBTIOAL

1

J_JL

JJJ

2-1 2-2 2-3

2-F00T LEVEL

7-1 7-2 7-3

7-FOOT LEVEL

EXPERIMENTAL LOTS

Fig 3 . Cumulative percentage mortality of oysters from all
baskets during the experimental period from January to December 1960,

mortality of only 8.6%. Difficulty was experienced in raising the con-
trol basket from the bottom, and the oysters in it were covered with a
sticky, black ooze. Even though members of the genus Crassostrea
possess a promyal chamber which provides for more efficient pumping
in silt-laden waters, the accumulation of silt has been shown to
cause mortality in C .virginica (Hsiao, 1950). It seems reasonable to
assume that a similar mortality occurred with the above-mentioned
control oysters in this study.

An experimental basket from the subtidal location became
entangled with the float at the beginning of the experiment; it was
impossible to free it from the bottom for three months . A 21 .8% mor-
tality was observed at the end of this period. Again, from examination
of the oysters, siltation appears to have been the major cause of

-58-

mortality. Subsequent biweekly raising and resetting of all subtidal
baskets, including the control, kept them free of mud and eliminated
any further large mortalities from this cause.

There was a slight differential mortality with respect to size.
While it may seem logical that the smaller individuals are killed more
readily by siltation, the size of dead oysters from locations not heavily
silted was also found to be smaller than the mean size of the population
at that time. It seems reasonable to assume that individuals infected
with a pathogenic organism would show a reduction in growth. If this
were so, then the average size of the dead individuals should be
smaller. Support is provided for this hypothesis in that dead oysters
were observed to have very little new shell growth .

On all checks from December through the middle of April a
voracious species of starfish, Pycnopodia helianthoides (Brandt) was
always observed attached to the underside of the subtidal baskets. On
several occasions this starfish had extruded its stomach through the
wire meshes of the basket and was in contact with the oysters. No
mortality could be directly attributed to this predator.

Twelve "gapers," dying oysters with the meat intact, were
recovered during the study and fixed in Zenker's fluid. These were
sectioned, stained with Iron-alum Hematoxylin and Eosin, and
examined histologically.

One "gaper" taken from the float in July was found to be infec-
ted by a multinucleated MSX-like organism, possibly pathogenic
(Fig. 4). The infection appeared to be general throughout the tissue
with concentrations of mulberry-like aggregations of cells in the
Leydig cell area, especially in the region of the digestive tubules.
A marked infiltration of leucocytes was noticed . The pathogenicity
of this organism is not known. Apparently the same organism has
recently been observed in other gapers from Washington by Dr . A. K.
Sparks; and Dr. J. G. Mackin (Dr. A. K. Sparks, personal communica-
tion) has found what may be the same organism in oysters from Europe
and the Gulf of Mexico and from both the Olympia and Pacific oysters
in Washington waters .

A large ciliate was found associated with most "gapers," but
this organism was probably a secondary invader and is not thought to
be pathogenic .

Although Pereyra (1961) has shown that yearly shell growth in
length is reduced 56% for oysters maintained at the 40% exposure level,

-59-

10 H

Fig. 4. Photomicrograph of a transverse section through palps
region of oyster infected with multinucleated organism.

this study has demonstrated no heightened mortality in these same
oysters as a result of exposure. Yields of oysters in this study thus
becomes essentially a function of growth and fattening, and not sur-
vival. These facts could be used to advantage by the oystermen if the
demand existed for a high-quality, small oyster.

For example, in Washington waters harvest time depends more
on the condition index or "fatness" of the oysters rather than on the
growth. This practice has caused some concern as the oysters are
often larger than wanted at the time they have reached the desired
"fatness ." This problem might be alleviated if some stocks of oysters
were maintained higher on the beach during their first year or two of
growth, and then were relaid at a lower tide level to "fatten" during
their last year. No increase in mortality would be expected to occur
and, early growing having been reduced, the size of the harvestable
oysters would be more desirable. Also, by proper rotation of stocks

-60-

on the bed, more growing area could be utilized. Of course, this
practice would only be advantageous in growing oysters where a smaller
oyster is desired .

ACKNOWLEDGMENTS

I wish to express my gratitude to the faculty of the College of
Fisheries for their guidance and suggestions, and particularly to Dr.
A. K. Sparks for his encouragement, counsel, and helpful criticism
throughout this study and for his interpretation of the prepared tissue
slides. The editorial advice of Messrs. D. L. Alverson and A. T.
Pruter is gratefully acknowledged . I am also grateful to Dr. Douglas
Chapman and Mr. Charles Junge for their valuable suggestions con-
cerning the statistical analysis of the data . Appreciation is expressed
to Dr. Kenneth Chew and Mr. Stanley Katkansky for collection and
analysis of hydrographic data and for their help in this field . Thanks
are also due to Dr. Ronald Eisler and Messrs . Richard Johnson, Martin
Nelson, Dennis Olsen, Juri Peet, Raymond Simons (for microphotography),
Arporna Sribhibhadh, Douglas Weber and Richard White; and Mrs . Daphne
Pereyra and Miss Sherry Haight who aided the completion of this study
by the donation of their time and effort.

LITERATURE CITED

Cahn, A. R. 1950. Oyster culture in Japan . U . S . Fish Wildl . Serv . ,
Fish. Leaf. 383. 80 p.

Chew, K. K. 1960. Study of food preference and rate of feeding of
Japanese oyster drill Ocinebra japonica Dunker . U. S. Fish
Wildl. Serv., Spec. Scient. Rept.— Fish. 365- 27 p.

Chew, K. K. and R. Eisler. 1958. A preliminary study of the feeding
habits of the Japanese oyster drill, Ocinebra japonica . J. Fish.
Res. Bd. Canada 15:529-535.

Elsey, C. R. 1934. The Japanese oyster in Canadian Pacific waters.
Proc . 5th Pan Pac . Sci. Cong . 5:4121-4127.

Galtsoff, Paul S. 1929. Oyster industry of the Pacific coast of the

United States. Bur. Fish. Document 1066, Append. VIII, Rept.
U. S. Comm. Fish. 1929:367-400.

Galtsoff, P. S. 1932. Introduction of Japanese oysters into the
United States . U. S. Bur. Fish., Fish. Circ . 12. 16 p.

-61-

Glude, John B. 1947. Oyster investigations. Ann. Rept . , Wash.
State Dept. of Fish. 1947: 17-20.

Hsiao, S. C. 1950. Effects of silt upon Ostrea virginica. Proc .
Hawaiian Acad. Sci. 25:8-9.

Kincaid, Trevor. 1951. The oyster industry of Willapa Bay, Washing-
ton. Calliostoma Co., Seattle. 45 p.

Pereyra, W. T. 1961. Growth of the Pacific oyster (Crassostrea gigas
Thunberg) in various exposure situations, with consideration of
the experimental and methodological difficulties encountered.
Unpubl.M.S. Thesis, Univ. of Washington, Seattle. 87 p.

Pereyra, W. T., K. K. Chew, and A. K. Sparks. 19 59. Growth of

1959 seed oysters in several areas of Washington. Pac . Coast
Oyst . Grower's Assoc . , 13th Ann. Conv . , Papers and Abstr .
Sclent. Sess. 1959:18-23.

Quayle, D. B. 1951. The seasonal growth of the Pacific oyster
(Ostrea gigas) in Ladysmith Harbour. Ann. Rept., British
Columbia, Dept. Fish, for 1950:L85-L90.

Sparks, A. K. and K. K. Chew. 1961. Preliminary report on growth

and survival of the Pacific oyster in Washington waters . Proc .
Nat'l Shellfish. Assoc . 50: 125-132 .

Thompson, J. M. 1952. The acclimatization and growth of the Pacific

oyster (Gryphaea gigas) in Australia . Austr. Jour. Mar. Freshw
Res. 3:64-73.

Woelke, C. E. 19 55. Introduction of the Kumamoto oyster, Ostrea

(Crassostrea) gigas, to the Pacific Coast. Wash. State Dept.
Fish., Fish. Res. Papers 1:41-50.

Woelke, C. E. 19 57a. The flatworm Psuedostylochus ostreophagus
Hyman, a predator of oysters . Proc . Nat'l Shellfish . Assoc .
47:62-67.

Woelke, C. E. 1957b. The quality of seed oysters from Japan. Wash.
State Dept. Fish., Fish Res. Papers 2:35-43.

-62-

Woelke, C. E. 1961. Pacific oyster Grassostrea gigas mortalities
with notes on common oyster predators in Washington waters
Proc . Nat'l Shellfish . Assoc . 50: 53-66 .

Woelke, C. E. Undated. Preliminary report on Pacific oyster mor-
tality survey data. Unpubl . ms . , Wash. State Dept. Fish.
Shellfish Lab., Quilcene, Wash., 5 p.

-63-

OYSTER MORTALITY STUDIES IN VIRGINIA
IV. MSX IN JAMES RIVER PUBLIC SEED BEDS

Jay D . Andrews
Virginia Institute of Marine Science

ABSTRACT

"MSX," an unnamed pathogen of oysters, caused an epizootic in
Chesapeake Bay which removed from production nearly half of Virginia's
private oyster-planting acreage between 1959 and 1961. The organism
did not appear in James River seed beds until fall of 1960. A tongue-
shaped distribution of MSX was apparently related to influx of salt water
along the channel. In 1960-61 and 1961-62, infections of MSX appeared
at Wreck Shoal in the middle of the seed area in October, and disappeared
the following April coincident with lowest salinities. Infection levels
were approximately 30 to 35% each year in populations adjacent to the
channel. No appreciable cold-season mortality occurred at Wreck Shoal.
MSX was nearly absent from Wreck Shoal oysters during the warm season
in summer salinities of about 15 pot, but at Brown Shoals, with salini-
ties 2 or 3 ppt higher, it persisted through spring freshets and caused
summer deaths. From observations for three rather wet years, it is con-
cluded that persistence of MSX infections in the James Piver seed area
depends upon importation of infective material from the saltier waters
of Lower James River and Hampton Roads. Also, damage to the seed
area will probably be reflected in quality of seed rather than direct mor-
tality. Planting infected seed in high-salinity waters leads to serious
losses.

INTRODUCTION

The 12-mile stretch of James River beginning at the bridge above
Hampton Roads exhibits several important characteristics which make it
the major seed area for Virginia: (1) Natural reproduction occurs regu-
larly with an intensity that produces seed of excellent quality; (2) growth
is slow and fattening is poor, which almost necessitates use of the
oysters as seed; (3) predation is negligible and diseases are restricted
by low salinities .

Most of the oyster grounds in the seed area are public "rocks"
from which wild oysters are harvested . The importance of James River
as a source of seed oysters to private planters in Virginia can scarcely
be overestimated . The apparent introduction of a new pathogen of

"Contribution No. 155 from the Virginia Institute of Marine Science

-65-

oysters called for special monitoring of the seed area for possible dam-
age . The undescribed pathogen called "MSX," which in Virginia was
first observed in 1959, has caused far more damage to private oyster
grounds than to public beds . MSX occurs only in areas with summer
salinities of about 15 ppt or higher and this is about the level required
by other diseases (e.g. Dermocystidium) and predators (oyster drills).
Consequently, only planted beds flourish in high-salinity waters and
our remaining public beds, with wild populations of oysters unreplen-
ished by man, are limited to low-salinity areas. MSX was not found
in the seed area until October 196 0 after two summers of heavy losses
in high-salinity areas of Chesapeake Bay (Andrews and Wood, in prep-
aration). Delaware Bay, which also has a seed area in relatively low-
salinity waters above the planting region, experienced a kill of seed
oysters in 19 58, the second and peak year of MSX activity in this
estuary (Haskin, 1960).

The primary objective of monitoring in James River was to keep
informed about distribution and incidence of MSX in order to advise
planters of the location of infected seed. By the fall of 196 0, when
MSX had infested the lower seed area, planting had ceased in the high-
salinity waters of Virginia because losses were intolerable. Most of
the planting that continued was in waters approximately of the salinity
range of the seed area. Fortunately, no serious consequences have
ensued from use of limited amounts of infected seed in these low-salinity
planting areas .

James River has been valuable for studies of the environmental
tolerances of MSX. A dense population of susceptible oysters in the
seed area has been continuously exposed to MSX, being close to inten-
sive plantings in an infested area of high salinity (Hampton Roads).
Some appreciation of the effectiveness of Hampton Roads as a reservoir
of infective material can be obtained from the history of private plantings
and losses. Prior to 1960 extensive new beds of James River seed were
planted each year. The last seed plantings were made in the spring of
196 0. After severe losses most remaining beds were harvested by June
1961. After this date less than 10% of normal oyster populations were
left in Hampton Roads and mortalities have continued in these survivors .
Therefore, in 1962 most old oysters were dead and new susceptibles
were not being planted .

Another advantage of Jemes River for disease studies is the
large drainage area and the high runoff which produce a rather steep
horizontal salinity gradient from the mouth to Jamestown. The river
also has probably the greatest seasonal fluctuation of salinity in the
Chesapeake Bay area. Since the "fringe" of the range of MSX falls in the

-66-

lower seed area, James River provides a place for assessing salinity
tolerances of M3X.

PROGRAM OF MONITORING MSX

Three locations were chosen as major stations for tray observa-
tions and sampling of native populations (Fig. 1). Brown Shoals at the
lower end of the seed area is characterized by rather permanent infesta-
tions of Dermocystidium and oyster drills. Wreck Shoal in the middle
of the seed area represents optimum conditions for seed production and
typically has the greatest spatfall and negligible mortality. Horsehead
Rock is near the upper limit of the seed beds and was the source of con-
trol groups of oysters for trays . No serious predators or diseases have
ever been found at this level of the river. A few supplementary stations
are also shown in Fig. 1. The major stations were deliberately located
close to the channel. Distribution patterns indicate that diseases and
predators tend to follow the deeper, saltier waters of the channel in
moving upstream (Andrews and Hewatt, 1957). The circulation of tidal
waters provides a mechanism for transport of materials upstream (Prit-
chard, 1953).

The management of tray stations on oyster beds has been des-
cribed (Andrews, Wood and Hoese, 1962). For each tray in James
River, native oysters were dredged from the vicinity of the station.
Trays were examined approximately biweekly. Death rates were obtained
from trays and were expressed as number dead per thousand per month
(easily converted to per cent) regardless of the lengths of periods between
visits . Few gapers were recovered for diagnosis of disease because
death rates were low, and to avoid biasing death rates, live oysters
were sampled very infrequently from trays. Incidence of disease was
obtained from frequent samples of live oysters from natural beds at the
three major stations . On each trip, counts of live oysters and boxes
were made from dredge hauls as a check on death rates in trays .

SALINITY AND TEMPERATURE RANGES IN JAMES RIVER

The winter season of minimal temperatures (about 5C) lasts
from mid-December until mid-March . During this period oysters were
inactive. Ciliary motion is very sluggish, no food is processed through
the gut, and little water is pumped. Temperatures rise rapidly in April
and May. From mid-June to 1 October, warm season temperatures of
25 to 3 0C prevail. Spawning and setting of oysters occur during this
period. October and November are periods of rapid cooling, although
mortalities may persist through these months .

■67-

3

r+

o

•

c

r+

=r

o

•

3

I — '

■<

3 2

0) <Q

'-'■ .

o

~l l-i

n- •

►i

^>

0) 3

3 Q>

O-TJ

to O

QJ l-h

3 V

•n w

— 3

?T p

lQ W

CO ^°

r+ £

QJ ^

r+ (t)

!-■ -1

o r„

3 w

O

Bj S

0) 3

co »Q

o W

Q)

o

H 3

3 a

C w

3 cr

cr (D

2! o

to £

£" ■"»■

n cr

r* !-»■

CD CL

&"8

CO
<-+
OJ
3

o

CD

QJ
3

a

r+

3"

cd

c
■a co

CD

2 CD

<• a

i-h CD
•-1 QJ

3 cr

Et o

?T <

<0 CD

-68-

The James River shows marked seasonal changes in salinities ,
Low salinities occur in winter and spring with minimal levels being
reached typically about 1 May. High salinities occur in late summer
and fall. Most oyster beds in James River are in shallow water (less
than 10 ft), hence little vertical gradient is observed. Hydrographic
stations were located in the channel but long experience has shown
that surface salinities in the channel approximate those on adjacent
oyster bottoms. Surface salinities given in this paper were taken in
the channel opposite Wreck Shoal at Nun Buoy "12" (Fig . 2).

isSf

aa.

NATIVE OYSTERS

NUN "12 • CHANNEL SURFACE

NOV JAN MAR MAY JUL SEP NOV
1962

Fig. 2. Salinities of surface waters in channel adjacent to Wreck
Shoal offshore station (No. 2); and seasonal incidence of MSX in native
oysters dredged at station 2.

In assessing salinity as a limiting factor in the penetration of
MSX into the seed area, it is important to know the climatological con-
ditions of the years studied. These are difficult to describe from
weather data alone, because the James River drainage area extends into
several climatological divisions of Virginia . Special attention should
be given to winter and spring weather for this is the period when salinity
conditions apparently become intolerable for MSX. The year 196 0 was

-69-

very wet in Tidewater bat dry in other parts of Virginia. However,
exceptionally low temperatures in March and exceptionally high ones
in April resulted in a heavy runoff from melting snow in the latter month.
Consequently salinities were persistently low until July in the seed area
The sudden change from cold to warm weather about 1 April is reflected
in the salinity curve (Fig. 2). Precipitation in 1961 was from 4 to 6
inches above normal, yet salinities were not as depressed as in 1960.
The winter and spring of 1961-62 was quite wet again and this was
reflected in rather strongly depressed salinities in March and April.

It may be concluded that all three years covered in this study
had above average rainfall and that spring salinities were below average
in the seed area. Both intensity and duration of low salinities, as well
as temperature level, are probably involved in controlling diseases and
predators . The salinity data were obtained at various stages of tide in
irregularly-timed trips. However, Fig. 2 does show seasonal trends
and the levels obtained in spring and summer, when MSX activities are
important. Further data on the range of salinities and the maximums and
minimums found at three levels of the seed area are given in Table 8 of
Andrews and Hewatt (19 57). Additional hydrographic information may be
obtained from Hewatt and Andrews (19 54), Andrews, Haven and Quayle
(19 59), and Chesapeake Bay Institute Data Report No. 7.

MSX INFECTIONS AT WRECK SHOAL

Seasonality

Wreck Shoal appears to be the key station for interpreting the
effects of environment on MSX activities. Seasonal incidence of MSX
is shown in Fig. 2. Each point on the graph represents a sample of 25
oysters from the same locality on Wreck Shoal adjacent to the channel.
No infections were found throughout the warm season of 1960 (March
through September), when the heaviest losses of the MSX epidemic
were being experienced in lower Chesapeake Bay. Beginning in Octo-
ber 196 0 and extending through February, about one-fourth to one-third
of the oysters in all samples showed MSX infections . In March and
April incidence declined without appreciable mortality having occurred
and by 1 May 1961 all infections had disappeared from Wreck Shoal
oysters .

In the summer of 1961 occasional infections were found at
Wreck Shoal in July and August . These may have been new infections
initiated in June, or they may have been old infections persisting from
the previous summer at a low level of intensity, which built up again
with the advent of favorable summer salinities and temperatures .

-70-

Again in October 1961 MSX infections rose to a level of about one-
third of the population. This incidence persisted through the winter
but dropped abruptly the first of May 1962. It was obvious in the late
April sample that MSX Plasmodia were declining in abundance and those
remaining did not stain normally. There was no evidence in stained
sections of leucocytic attack or other activity by oysters to remove
Plasmodia . A rapid improvement in the cytological picture of oyster
tissues accompanied this change. Tissue reactions in oysters, as
indicated by leucocyte concentrations, rapidly disappeared from May
and June samples .

Again in the summer of 196 2 occasional infections of uncertain
origin were found, but in the fall incidence remained low (10% or less).
The failure of MSX to establish a high level of infection at Wreck Shoal
in 1962 is probably related to the relative inactivity of MSX on the
depleted beds of Hampton Roads. Oysters were scarce in Hampton
Roads in 1962 and, as a consequence, there was probably a scarcity of
infective material that could be transported upstream into the seed area .

Mortality

MortaliLy of oysters at Wreck Shoal was almost negligible through-
out the period of this study. This has been true in all the 16 years I have
been using Wreck Shoal as a sampling station. Counts of live oysters
and boxes in material dredged from the bar never showed more than 9%
boxes (Table 1). Slight rises in box counts occurred in late winter each
year and a higher percentage of recent deaths was indicated by fresh
new boxes. Fresh-water kill is not expected at Wreck Shoal (Andrews,
Haven and Quayle, 1959) although it did occur upriver in 1958, 1960
and 1962. James River oysters generally exhibited weakness in 1960
and 1962 which was vaguely attributed to wet years and poor feeding
conditions . Few deaths occurred on native beds but the extra hardships
of transplanting led to some losses of seed oysters . None of these
losses above Wreck Shoal were associated with MSX or other known
diseases .

Two lots of tray oysters were maintained in the Wreck Shoal
area to follow mortality (Fig. 1). The offshore station (Tray J3) was in
the oyster sampling area near the channel whereas Tray Jl was located
far inshore (Jail Island station). Death rates never exceeded 3% per
month, except in the late summer of 1962, and usually were less than
1% per month (Fig. 3). Gaper and live oyster samples indicate that
both Dermocystidium and MSX caused slight losses in Tray J3 in the
fall of 1962. The inshore tray had less mortality than the offshore tray,

-71-

Table 1 . Count

s of live

oysters and t

oxes ,

James

River, Virg

inia ,

1960-

1962.

Date

Live
oysters

Boxes

Total
dead

Per cent

Location

New

Old

dead

Brown Shoal

13 May

60

237

9

13

22

9

27 July

274

2

2

4

1

18 Aug

160

1

12

13

8

7 Sept

318

5

9

14

4

20 Sept

278

11

6

17

6

12 Oct

280

1

28

29

9

31 Oct

3 29

15

36

51

13

21 Nov

175

14

21

35

17

27 Dec

171

15

23

38

18

22 May

61

219

22

19

41

16

1 Aug

280

6

34

40

12

14 Aug

260

21

97

118

31

11 Sept

282

7

35

42

13

31 Oct

432

5

5

10

2

14 Nov

474

19

23

42

8

22 Jan 62

242

4

47

51

17

27 Feb

217

9

51

60

22

25 April

420

22

14

36

8

2 2 May

146

-

52

52

26

9 July

358

5

13

18

5

27 July

210

0

7

7

3

14 Aug

385

3

18

21

5

23 Aug

225

8

30

38

14

7 Sept

198

2

8

10

5

24 Sept

158

8

24

32

17

16 Nov

235

5

14

19

7

Wreck Shoal

13 May

60

715

3

6

9

1

28 July

370

1

1

2

1

18 Aug

314

0

1

1

0

7 Sept

375

0

1

1

0

2 0 Sept

365

1

3

4

1

12 Oct

345

1

2

3

1

31 Oct

432

5

5

10

2

21 Nov

285

4

6

10

3

15 Feb

350

3

6

9

3

28 Mar

350

15

8

23

6

19 April

250

16

5

21

8

-72-

Table 1. (continued)

Live
oysters

Boxes

Total
dead

Per cent

Location Date

New

Old

dead

Wreck Shoal 5 June

310

2

8

10

3

26 June

250

1

9

10

4

1 Aug

385

2

7

9

2

2 2 Aug

313

0

7

7

2

5 Sept

220

0

4

4

2

11 Oct

200

1

3

4

2

6 Nov

358

9

10

19

5

24 Jan 62

375

3

8

11

3

28 Feb

280

4

7

11

4

10 April

315

12

20

32

9

2 5 April

395

20

13

33

8

1 May

300

7

20

27

8

23 May

335

0

15

15

4

7 June

285

4

6

10

3

25 June

430

2

9

11

2

11 July

350

1

6

7

2

27 July

350

1

10

11

3

16 Aug

255

0

26

26

9

18 Sept

360

3

7

10

3

15 Nov

295

2

5

7

2

17 Dec

360

4

11

15

4

Rainbow Rock 22 Aug 61

325

0

6

6

3

28 Feb 62

225

3

6

9

7

3 0 July

375

4

11

15

7

Horsehead 13 July 6 0

443

16

9

25

5

1 Aug

149

0

9

9

6

7 Sept

300

2

4

6

2

31 Oct

285

0

4

4

1

15 Dec

400

0

4

4

1

16 May 61

300

5

12

17

5

5 June

340

2

21

23

6

16 June

330

2

15

17

5

26 June

270

2

6

8

3

1 Aug

340

2

7

9

3

2 2 Aug

325

0

6

6

2

11 Oct

321

1

7

8

2

-73-

Table 1. (continued)

Date

Live
oysters

1
Boxes

Total
dead

Per cen1;

Location

New

Old

dead

Horsehead

24 Jan 62

260

1

3

4

2

28 Feb

285

9

10

19

6

10 April

335

26

9

35

9

1 May

325

21

5

26

7

23 May

275

31

8

39

12

7 June

275

11

15

26

9

11 July

275

0

20

20

7

3 0 July

325

2

20

22

6

18 Sept

200

0

19

19

9

15 Nov

345

0

3

3

1

Deep Water

Shoal

12 July 60

298

6

16

22

7

16 May 61

290

3

3

6

2

20 Mar 62

300

2

6

8

3

10 April

210

4

7

11

5

"Boxes" are pairs of empty valves still attached at the hinge. The "new
boxes" column includes "gapers," dead or dying oysters with some or
all of the meat remaining.

-74-

Jl WRECK SHOAL IINSHORE)
J! BROWN SMO*L

J3 WRECK SHOAL [OFFSHORE!

f

S JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT MOV 0£C

§ I960 1961 1962

Fig. 3. Death rates of native oysters in trays from 1960 to 1962

and no diseases . Slight increases in death rates occurred in late winter
of 196 0 (not related to diseases) and again in early spring of 1962 (prob-
ably related to MSX).

The chief conclusion from mortality data is that no appreciable
loss of oysters was associated with the cold season prevalence of MSX
at Wreck Shoal. Therefore, incidence of MSX was not appreciably
changed by mortalities. During the warm season, incidence was low
and mortality remained low.

MSX AT BROWN SHOAL

The Brown Shoal station at the lower edge of the seed area pro-
vides an interesting environment for oyster diseases, intermediate
between low-salinity seed beds where diseases are absent and high-
salinity planting beds where oysters are now decimated by MSX.
Salinities average about 3 ppt higher at Brown Shoal than at Wreck
Shoal with summer levels of 18 to 2 0 ppt (Andrews and Hewatt, 19 57).
The kill on Brown Shoal has never exceeded 5 0% and a substantial
population of oysters has always been present for disease-producing
organisms to persist in. Some recruitment of new year-classes has
occurred .

Mortalities at Brown Shoal are shown in Fig . 3 and Table 1 .
Deaths from MSX began in the fall of 1960 as indicated by a 13 to 18%
count of boxes through the winter. Oysters dredged from the vicinity
of the tray station were placed in Tray J 2 on 1 April 196 0. One year
later 3 0% were dead. Both Dermocystidium and MSX infections were

-75-

found in gapers and it is difficult to determine the proportion of deaths
caused by each. However, only 4 of 8 gapers were infected with
Dermocystidium in the late summer and fall, when the fungus is known
to cause deaths in Chesapeake Bay. Another 17 gapers were obtained
in the late winter and spring, when Dermocystidium is rarely found in
gapers, but not all of these gapers were diagnosed for MSX.

From April 1961 to April 19 62, box counts from natural beds never
exceeded 31%, yet 58% of the oysters in Tray J2 died. In 12 gapers col-
lected from the tray, two had serious Dermocystidium infections (one of
these also had MSX) and 4 of 6 examined for MSX were infected. From
April 1962 to December 1962, mortality at Brown Shoal in Tray J2 was
3 0% and on the natural bed it was considerably lower (Table 1).

Incidence of MSX in live oysters at Brown Shoal is given in
Table 2. Infections appeared earlier in the late summer and fall of 19 6 0
than at Wreck Shoal and resulted in some deaths from MSX. Deaths
were not numerous enough to depress incidence below the level of prev-
alence found at Wreck Shoal in the fall and winter of 1960-61. Further-
more, there is no evidence that MSX prevalence was changed by low
spring salinities as it was at Wreck Shoal. Death rates from MSX were
high in the early summer of 1961 before any Dermocystidium appeared .

MSX infections continued through the summer of 1961, and in
the fall and winter of 1961-62 about 20 to 25% of natural bed oysters
were infected . This is lower than the level of incidence at Wreck
Shoal, but previous selection plus a few fall deaths could account for
the difference. By 1 April 1962, nearly all infections had disappeared
at Brown Shoal and a substantial increase in box counts suggests that
infected oysters died in late winter. Since 1 April 1962, MSX has been
present at Brown Shoal only in a very small percentage of oysters . Low
incidence and no substantial death rate indicate that very few new
infections of MSX occurred at Brown Shoal in 1962.

Incidence of Dermocystidium in live oysters at James River
stations is given in Table 3 . Only very small losses could be expected
on natural beds at Brown Shoal from weighted incidences of 0.50 or less
(Andrews and Hewatt, 1957), but Tray J2 reached an incidence level
(1.24) in 1961 which could be expected to produce a significant kill.
The excess losses in Tray J2 over the death rates on Brown Shoal beds
can probably be assigned to Dermocystidium .

-76-

0

4->

in
O

LO
(Nl

0
N

w

0

"a.

e

rO

co

CD

>

-f-i

co
0

(0

(0

o
si
co

o

CO

rO

X

CO

CD

u

c

CD

•i— c

o
c

CM

CD

rO
H

c

CD

Q)

+J

Q

u

CD

!h

U-l

0

C

a.

CM
CO

en

CD
■*->
CO

Q

c

CD

0

-M

u

0

J-,

CD

CD

C

CL,

■r-l

CO
05

CD

4->

ro
P

4^

c

CD

CD

■i-i

0

O

i-t

(D

CD

C

IX

-.-I

o

CO

CMC0'3,OOO^3<O'^O'q<'<3,'cT'

CM

•Q

CD

•73 >, CD

CO

c
3

2^^ rj rji a a <->

3 3 3 3 » D U
h, h, < < CO CO O
csiNOiotMCOcnN^nN'^cn

N N CO N N H CM f— I CO (M N

a
<

o

CD
P

o

CO

CD

en

LO

CD

■ — i

-t-J

CD

o

4-1

ro

O

P

ro

CM
CO

O

■^ eD O
CM <— I CM

CM ^r

■— I CM

Q.5

CM
CM

>.

cn a> o.

3 3 0)

< < CO

.— I Tf I— I

> o

o 0

£ P

o o -^ o

■vT CO CO CM CM CO
i— I ■— 1 t-t CO CO

r-l

1^

ro

CD
C

i?

Oi

CL
<

2

3
t — *

"3

3
<

— 1

CO

r>-

CO

CO

1 — 1

CM

CM

1 — 1

Q, Cl +j -t-1 ►> u

0 CD U O O 0

co co O O £ P

C~- O CM i— I i— I P-^

CM r— I CO CM CM

ro

-a

xi
c

si

0

x>

CO

0

cn
0

si

4->
C

0

u

ro
a

w

i-.
0

S

3
C

X!

-a
0

4-"

o

c

ro

4-»

CL

0

O
X

w

-77-

QJ

-^

i-h CD

1=1 <Q
^ 3"

3"

O ^
en CL

<T5 (D
W 3

r1 cd
• S

QJ QJ
3 en

ao
o cr

r+

QJ

5 2-
<

QJ

i — ■
C
CD

en

O
*-(

3"
CD
QJ
<

QJ

CO
CD

to

CO
-1
O

co

3'
O

QJ

id

CD

1 — '

K

CD

cd

CD

QJ

O

•-i

H
*-(

QJ

DO
■-!

o

3

CO

3"
O
QJ

n

CO

3"
O

QJ

QJ

CD
•-S

CO

O

QJ

3

o

CO

3

3"

O

QJ

O

t~<-

-i

3

3

CD

to

to

1 — "

to

1 — •

to

to

to

CO

to

1 — '

CD
iQ

0)

en

C£>

*»

•<]

*»

-g

CD

►fc.

cxi

1 — '

^J

i — <

o

CO

cr
•<

o

o

CO

CO

>

« — i

c

o

2

CO

2

a

O

CO
CD

' — 1

c

Cf

o

o

CD

CD

3

n

o

CD

o

CD

o

<

o

01

QJ

w
en

i— •-

r-T

r+

T3

r+

CI

■<

r+

<

73
r+

<

O

r-+

r-t

•<

CD

CI
<-t
(-•■

o

3
W

0

-! :

3

o

D.
CD

"i

Cu

to

CO

to

[O

to 00

en en ►£>■ to oo

CD

CD

to

O

O

CD

J^

CD

CD

O

CD

~vj

CO

O

O

CD

to

O

O

CD

*»

to

to

O

CD

CD

O

o

o

to

o
o

o
o

to

o

CO

o
oo

o
o

o
.fe.

00

en
to

o
o

a

QJ

r+

CD

w

cr
<

CD

o

CD

3

r-t-

QJ
iQ
CD

en

CD
O
t-r

o

3

,_. CD

cd «a

3 =r

QJ

cr

CD
CO

to t?
en 3

a

o
o
■<

w 2

r-t 3

CD O

• o

o

n

a

<

CD

O
•<
en

QJ
3

CD

en
73

t— '-

<

CD

-t

cQ

3
QJ

CO
QJ

"2-
CD

N
CD

QJ

en

-78-

MSX ACTIVITY IN LOW- AND HIGH - SALINITY AREAS

Two populations of seed oysters which became infected with
MSX in James River in early summer of 1961 provide an interesting com-
parison of the effects of salinity level. One was the population on
Wreck Shoal which has been described . The other population consists
of lower James River seed moved to Plot 14-16 in Mobjack Bay in Aug-
ust 1961. Summer salinities are about 25 ppt in Mobjack Bay, which
is some 10 parts higher than at Wreck Shoal. Both groups were of
similar age and history.

Incidence of MSX in these two populations from August 1961
through 1962 is shown in Fig. 4. Although infections appeared some-
what earlier in Mobjack Bay than at Wreck Shoal, the level of infection
through the late fall and winter was similar. In April and May, inci-
dence in Mobjack oysters increased to well over 50% from late summer
infections but MSX almost disappeared from Wreck Shoal oysters. By
the first of May 1962 over 6 0% of Mobjack tray oysters were dead
(Andrews, in preparation), whereas less than 10% of Wreck Shoal
oysters had died. Low incidence of MSX at Wreck Shoal during the
warm season from May to October 1962 resulted in less than 10% total
annual losses whereas the Mobjack oysters suffered 80% mortality in
one year from the time the first deaths from MSX were observed in
August 1961 .

80

-

70

-

6U

50

40

30

20

10

1

0

i

1

SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

1961 1962

Fig. 4. Comparison of seasonal incidence of MSX in low- and
high - salinity areas .

-79-

Since both lots were James River seed stock, infected while
still in the seed area, the striking differences in mortalities and per-
sistence of infections appear to be related to salinity levels. It is
probable, however, that the planting in Mobjack Bay was exposed to
increased dosages of infective particles in the fall of 1961, which may
have accelerated and increased mortalities .

DISCUSSION

The epidemiology and theory of disease of MSX has been dis-
cussed (Andrews, in preparation) but further elaboration is possible
with the James River data . Nothing is known of dosage of infective
material and the source is only indirectly suggested by field studies .
My theory of disease rests upon the presumption that MSX is trans-
mitted directly from oyster to oyster via water currents .

The timing of infections is important in understanding MSX
activities in James River. It was shown (Andrews, in preparation)
that disease-free oysters imported to high-salinity areas after 1 August
probably became infected in late summer but did not show infections
until the following spring . If infections do not develop in high salini-
ties it is assumed that they are not likely to develop in lower salinities .
Hence, the Wreck Shoal infections which appeared in October must have
been initiated before 1 August. Furthermore, oysters imported to high-
salinity areas from Wreck Shoal in August 196 0 promptly showed MSX
infections — over a month earlier than they appeared in the population on
Wreck Shoal. This means that Wreck Shoal oysters had undetectable
infections when moved. It appears, therefore, that early summer infec-
tions occurred at Wreck Shoal but that they did not become evident until
October. Furthermore, the usual late summer kill was omitted at Wreck
Shoal. It is presumed that the prevailing summer salinities did not pre-
vent infections but inhibited development of MSX. Late appearance and
high incidence in late fall was also observed in Pocomoke Sound which
appears to be a fringe area for MSX. Late fall and winter incidences
were relatively high at Wreck Shoal (3 0%) and often exceeded the levels
in high-salinity areas where fall mortalities had occurred .

The source of infective material is presumed to be Hampton Roads
and Brown Shoals in lower James River. Oysters in the vicinity of Wreck
Shoal could not have provided a source since sick and dying oysters were
absent in summer when infection occurred. More important is the proba-
bility that distribution of MSX in James River is related to circulation.
MSX is present on deeper beds near the channel but absent from inshore
shallow areas at the same level of the river. This tongue-shaped pattern
of distribution has been observed previously for Dermocystidium in James

-60-

River (Andrews and Hewatt, 1957). The data supporting this apparent
distribution of MSX are scattered by season and location, making pre-
sentation difficult and interpretation somewhat subjective. At the level
in James River just cbove the bridge, MSX has been found in inshore
shallow waters on both sides of the channel but less often and with
lower incidence than at the Brown Shoal station near the channel. MSX
has never been found at the inshore tray station opposite Wreck Shoal.

The apparent decline of MSX in the seed area in 1962 easily fits
into my theory of the source of infective particles. Since no new sus-
ceptibles had been planted in Hampton Roads, a very small scattered
population of oysters, mostly on public beds, remained in 1962. Even
Brown Shoals, which never had more than 5 0% kill, showed relatively
few deaths in 1962 . In the absence of information on changes in viru-
lence of the pathogen, I suggest that infective particles have become
scarce . If MSX were being supported at epizootic levels by infective
particles originating from another host or from the immediate locality
of Wreck Shoal, there is no evident reason why a decline in activity
should have occurred in 1962 . An abundant supply of thickly-populated
susceptibles still exists in the seed area . There is evidence of con-
tinued virulence in York River (Andrews, in preparation).

The disappearance of MSX from oysters at Wreck Shoal in April,
during the lowest salinities of the year, seems to be quite strong cir-
cumstantial evidence that salinity is the primary controlling agent. No
clear case of reduced incidence in high-salinity waters has occurred
in Virginia without accompanying mortality. In high salinity, death of
infected oysters reduces their number so that it is frequently difficult
to reconcile excessive mortality rates with moderate levels of incidence
found in live oysters . The impression is gained that few if any oysters
recover in high-salinity areas once MSX infections are patent.

If oysters contribute to the expulsion of MSX on Wreck Shoal,
it is peculiar that it can only happen in April and that all oysters are
capable of disposing of infections during the same short period. Wreck
Shoal oysters moved to Gloucester Point in early April retained their
infections and exhibited a substantial death rate in spring and summer.
Hence MSX remained viable all winter and infections were well estab-
lished physiologically in early April, yet all were gone by 1 May.

It is a little dismaying to inspect Fig . 2 for the level of salinity
during the period (June and July) when initiation of infections was
believed to have occurred. In 1960 and 1961, salinities apparently
never rose above 10 ppt during June, an important month for new infec-
tions, and only in late July were summer levels reached at Wreck Shoal.

II-

Summer salinities of about 15 ppt apparently delayed the development
of infections at Wreck Shoal, whereas about 20 ppt at Gloucester Point
permitted deaths from early summer infections to begin in August. There
is some evidence that first-summer deaths from MSX begin even earlier
in the high-salinity waters of Mobjack Bay. It is possible that dosage
of infective particles is the important factor in delaying plasmodial
infections, but again close timing and group response suggest that
physical factors are involved, and the quick appearance in oysters
transplanted to high-salinity waters implies that infections are inhibited
at Wreck Shoal. Sprague (1961) reported that MSX persisted and killed
oysters at a salinity range of 14 to 16 ppt in aquaria .

The distribution of many estuarine organisms is apparently regu-
lated by salinity, hence it is not surprising to find that MSX has limits
too. It may well be that April purging of MSX requires a salinity as low
as 5 ppt, for in both 1961 and 1962 salinity reached this level. It is
interesting to note in comparison that some Dermocystidium survived
one winter and spring at the upper end of the seed area and exposure to
fresh water during that sojourn. Virginia winter temperatures do not
appear to affect the survival of MSX.

The ecological technique of searching along the gradient of a
factor for its effect on distribution and tolerance of an organism has been
applied in regard to salinity and MSX. A rather steep salinity gradient
exists from Hampton Bar in lower James River to Horsehead in the upper
seed area. At the low-salinity station at Horsehead, no MSX has ever
been found and no disease-caused mortalities have been established.
At Wreck Shoal, infections of MSX occur, apparently from downriver
imports of infective particles, but development is slow with no late sum-
mer and fall deaths . In April infections disappear, apparently from
exposure to low salinities, and no appreciable death rate occurs because
Wreck Shoal oysters are essentially free of evident infections during the
warm season. At Brown Shoals also, oysters apparently acquire infec-
tions from downriver imports of infective material. MSX did not persist
in a substantial population of Brown Shoal oysters after the source of
infective particles was removed by decimation of Hampton Roads beds .
At Brown Shoals infections developed earlier and a few deaths occurred
in the fall and winter following early summer infections. Infections were
not removed by spring low salinities. Mortalities were less than 50% per
year with both MSX and Dermocystidium as agents of death . The deci-
mation of oyster beds in fully epizootic areas, such as Hampton Bar, is
described in another paper (Andrews, in preparation).

MSX poses no great threat to the James River seed area in terms
of mortality unless exceptionally dry seasons are coordinated with

-82-

intensive epizootics in Hampton Roads as a source of infective material.
Virginia is perhaps fortunate that the current epizootic occurred during
rather wet years. Whenever infections are present in James River, the
guality of seed oysters is seriously compromised for planting in high-
salinity waters . This is fully exemplified by the plot 14-16 oysters in
Mobjack Bay (Andrews, in preparation) in which oysters began dying
within 3 month after transplanting and nearly 8 0% were dead within a
year. It is concluded that the presence of MSX in the James River seed
area depends upon repeated invasions from saltier waters. Oyster
setting may also be dependent upon the tidal transport system to carry
larvae upstream from Hampton Roads . The decimation of Hampton Roads
oyster beds was followed by a nearly complete failure of spatfall in the
seed area of James River in 1961 and 1962 .

ACKNOWLEDGMENTS

I wish to express my gratitude to Dr. John L. Wood for much
advice and help in disease studies . He has been a constant associate
in all phases of the MSX epizootic . My appreciation is also extended
to the Microtechnique Section which under Dr. Wood's direction has
prepared all the slides for disease diagnosis .

LITERATURE CITED

Andrews, J. D. Oyster mortality studies in Virginia . V. Epizootiology of
MSX, a protistan pathogen of oysters. In preparation.

Andrews, J. D., Dexter Haven, and D. B. Quayle. 19 59. Fresh-water
kill of oysters (Crassostrea virginica) in James River, Virginia,
1958. Proc . Nat'l Shellfish. Assoc. 49: 29-49.

Andrews, J. D. and W. G. Hewatt . 1957. Oyster mortality studies in
Virginia . II . The fungus disease caused by Dermocystidium
marinum in oysters of Chesapeake Bay. Ecol . Monogr . 27: 1-25 .

Andrews, J. D. and J. L. Wood. Oyster mortality studies in Virginia.
VI. History and distribution of the pathogen MSX. (In prepara-
tion).

Andrews, J. D. , John L. Wood, and H. Dickson Hoese. 1962. Oyster
mortality studies in Virginia . III. Epizootiology of a disease
caused by Haplosporidium costale Wood and Andrews . J . In-
sect Pathol. 4:327-343 .

-83-

Haskin, H. H. 1960. Delaware Bay oyster mortality project annual

report to the U.S. Fish and Wildlife Service. Rutgers Univer-
sity . Mimeo .

Hewatt, W. G . and J . D . Andrews . 19 54. Oyster mortality studies in
Virginia. I. Mortalities of oysters in trays at Gloucester Point,
York River. Texas J . Sci . 6 : 12 1-133 .

Pritchard, D. W. 19 53. Distribution of oyster larvae in relation to

hydrographic conditions . Proc . Gulf and Caribbean Fish. Inst.
5 (1952): 123-132.

Sprague, Victor. 1961. Apparent decrease in incidence of MSX under
certain laboratory conditions. Univ. Maryland Natural Resour-
ces Institute. Mimeo.

-84-

DRIED UNICELLULAR ALGAE AS FOOD FOR LARVAE OF THE
HARD SHELL CLAM, MERCENARIA MERCENARIA

Herbert Hidu and Ravenna Ukeles

U.S. Bureau of Commercial Fisheries
Biological Laboratory, Milford, Connecticut

ABSTRACT

Three species of unicellular algae preserved by drying have
been shown to be utilizable as food by larvae of the hard shell clam,
Mercenaria mercenaria (Venus mercenaria). The chlorophyte, Dunaliella
euchlora, and the chrysophyte, Isochrysis galbana, when freeze-dried
gave clam growth and survival comparable to that obtained when these
species were fed at similar rates as live algae. Heat-dried samples of
the green algae, Scenedesmus obliquus, mass-produced in Japan, also
gave good growth and survival of clam larvae. Sulmet (sodium sulfa-
methazine) increased growth and reduced mortality of clams when used
with the dried foods. Agitation of clam cultures proved advantageous
when used with dried S^ obliquus but gave variable results with the
other two species of dried algae. Dried unicellular algae under suitable
culture conditions possess desirable physical and nutritive properties
for an ideal non-living food for mollusks. If adequate culture techniques
are developed, dried algae may find wide application, as in the rearing
of juvenile clams and larval oysters.

INTRODUCTION

Larvae of the hard shell clam, Mercenaria mercenaria,
can be reared to metamorphosis on a variety of living unicellular algae,
but the most rapid growth occurs when these larvae are fed naked
flagellates, particularly the chrysophytes , Monochrysis lutheri and
Isochrysis galbana (see Davis and Guillard, 1958). The species of
algae that are the best foods for larvae are usually difficult to main-
tain and a considerable investment of time, space, and equipment
would be required to provide a sufficient quantity of such foods in live
condition for a large hatchery. The desirability of developing a non-
living food, that could be stored for long periods, for the hatchery
production of shellfish is self-evident. An inexpensive, non-living
food available in quantity would not only be extremely practical for
routine use in hatcheries, but would also assure a uniform and reliable
source of food for many experimental studies on bivalve larvae.

Several attempts have been made to develop foods other than
living algal cells for molluscan larvae. Loosanoff (personal communi-
cation), as early as 1944, attempted to rear larvae on yeast cells but

-85-

obtained unsatisfactory results. Carriker (1956) reared larvae on com-
mercial cereal pablum flakes with sporadic success.

Chanley and Normandin (1960) tested the value to clam larvae
of a wide variety of non-living organic materials, most of which were
of little value or had other undesirable characteristics. However,
when L galbana , that had been concentrated and frozen, was resus-
pended in larval cultures, growth and survival of clam larvae were
comparable to those of larvae receiving a live culture of I_. galbana .
Also, fresh ground preparations of sea lettuce, Ulva lactuca, were
found to be an acceptable food for clam larvae. These foods, how-
ever, were not entirely satisfactory, in that they produced substantial
debris in clam cultures.

Recently our attention has been directed towards use of dried
unicellular algae as a molluscan food . It was hypothesized that algal
species suitable for molluscan food as living algae may also be suit-
able when resuspended from a dried preserved preparation. These
species might then be mass-produced commercially and dried to produce
a uniform preserved molluscan food .

Mass production and drying of unicellular algae have been given
considerable attention in recent years (Tamiya, 1957). Production of
various species of green algae, Chlorophyceae, for use as a dietary
supplement in food for humans and livestock, has thus far received most
emphasis. Production of this dried material is now in an advanced stage
of development with cost estimates at 17 to 26 cents per pound dry weight
(Tamiya, 19 57). Little developmental work has been done in producing
and drying, in quantity, species of other algal groups, such as the Chryso-
phyceae. If the chrysophytes are still to be the best foods after freeze-
drying, it may be possible to modify existing mass culture techniques to
produce a product designed specifically for use in the culture of mollusks .

In this paper we report our results using dried unicellular algae as
food for clam larvae, M_. mercenaria . We have tested preparations of three
algal species, specifically the chlorophytes , Dunaliella euchlora and
Scenedesmus obllquus , and the chrysophyte, J_. galbana . In this work we
have compared growth and survival of clam larvae, when fed the dried
algae, with that of clam larvae receiving no supplemental feeding. Also,
in the case of D. euchlora and'_I. galbana, live counterparts of the dried
algae were used as controls .

We have also included in this report a description of the methods
of culture of larvae that have been found to be of value when using dried
algae as food .

-86-

METHODS

Except as noted, standard methods for culturing bivalve larvae
were used throughout the present series of experiments . The methods
of conditioning and spawning adult clams and rearing bivalve larvae in
the laboratory have been described in detail by Loosanoff and Davis
(1950).

Adult clams were spawned, and fertilized eggs were cultured
for 48 hours at concentrations of 3 0 per ml of sea water. The sea water
used was first passed through a Ful-flo orlon filter and then treated
with ultraviolet light. After 48 hours the veliger larvae (Fig. 1) were
placed in the test culture beakers generally at concentrations of 1 0 per
ml at a temperature of 24 + 1 C. The sea water in all cultures was

Fig. 1. Clam larvae 2 days after fertilization at initiation of
experiments .

changed every second day by retaining larvae on a 250-mesh-per-inch
screen. The experimental foods were added daily. After 10 days of
feeding, i.e., 12 days after the eggs were fertilized, all cultures were
sampled quantitatively. Growth was determined by measuring the
length, parallel to the hinge line, of 5 0 larvae from each sample. Sur-
vival was determined by counting the total number of living larvae in
the sample and comparing this with the number present at the initiation
of the experiment. A uniform culture time of 12 days was practiced so
that data from experiment to experiment could be compared directly.

-87-

Dried S_. obliquus was obtained from the Micro Algae Research
Institute of Japan where it was prepared by spray-drying with the use
of heat. The dried algal powder was prepared as a suspension in sea
water before addition to clam cultures . Two methods of preparation of
the algal suspension were tested. One method was by agitating a small
amount of the dried material and sea water in a Waring Blendor. The
metal parts of the blendor that came in contact with the food suspensions
were coated with epoxy resin. Blending for one minute produced a sus-
pension consisting primarily of single intact cells and quadrads of cells.
Single cells were 10 to 15 microns long, a size that clam larvae can
ingest .

The second method of resuspending the dried S_. obliquus pro-
duced broken algal cells. It consisted of grinding the dry powder in a
ceramic jar mill with Burundum grinding cylinders. Two hours of
grinding were sufficient to break almost all of the cells . The resultant
particles were mostly 1 to 5 microns in size, although some were as
large as 5 0 microns . The ground powder was then washed through a
325-mesh-per-inch screen with sea water to form the suspension used
in feeding. With either method, a fresh suspension was prepared daily
for feeding .

The living cultures of D. euchlora and |_. galbana were obtained
from our mass culture apparatus (Davis and Ukeles, 1961). The dried
powders were prepared by concentrating the algal cultures by centri-
fugation and then lyophilizing the concentrate in a Virtis manifold
freeze-drying apparatus. The dried algae was readily resuspended for
feeding by adding sea water to the glass lyophilizing tubes and shaking
vigorously. A Vibro-Mixer, which mixes by means of a vibrating glass
plunger, was also useful in breaking up clumped algal cells that remained
in the food suspensions .

To standardize rates of feeding, equal packed cell volumes of
each food were used (Davis and Guillard, 19 58). Densities were deter-
mined by centrifuging each food suspension in a Hopkins tube. After
drying, the cells were resuspended in the same volume of water as was
present before drying, so that the original packed cell volume could be
used in calculating comparable feeding rates of dried and living _I. gal-
bana and D . euchlora .

In addition to our usual method of culturing larvae in standing
cultures several methods of agitating the cultures were used to deter-
mine whether agitation was necessary to keep the dried algae from
settling. These included (1) an aeration device in which compressed
air was finely dispersed through a "stone" and emitted at the bottom of

the culture container; (2) the turbidity apparatus, described by Davis
(I960), in which closed containers of larvae were rotated on a wheel at
a rate of 8-10 rpm; (3) an Eberbach bacteriological shaker on which
closed containers of larvae were shaken at a rate of 4 0 to 6 0 cycles per
minute; and (4) a machine devised to produce simultaneously a hori-
zontal padding action in some cultures and a vertical plunging action in
others (Fig. 2). The paddles and plungers of this machine were made
of nontoxic Plexiglas and were operated at a rate of approximately 7
cycles per minute .

Fig. 2. Agitation machine with which the water in cultures of
bivalve larvae is kept in motion by slow action of paddles or plungers
Culture vessels are partly immersed in a constant temperature bath.

-89-

Sulmet (sodium sulfamethazine) obtained from the American
Cyanamid Company was used, except where noted, as a bacteriostatic
agent. Dosage rates were between 50 and 100 parts per million (ppm).

RESULTS

Scenedesmus obliquus

In the first series of experiments stationary cultures were used
to compare the rate of growth of clam larvae receiving resuspended
intact cells of dried _S . obliquus with that of larvae receiving dried
S_. obliquus with cells broken by grinding (Table 1). Cultures of larvae
receiving a mixture of living flagellates and cultures receiving no sup-
plementary food served as controls . In these experiments no bacterio-
static agent was used.

Table 1. Mean lengths at 12 days of clam larvae in non-agitated cul-
tures receiving dried_S. obliquus , a mixture of live flagellates
and no supplementary food. Mean lengths are averages of
duplicate cultures. Feeding rate was 0.01 ml of packed cellu-
lar material per liter of larval culture per day.

Culture treatment

Mean length at 12 days (microns)

Experiment 1

Experiment 2

S_. obliquus (intact cells)
S_. obliquus (ground)
Flagellate mixture (live)
No supplementary food

134.20 + 2.48

(1)

161 .50 + 2.14

214.10+1.16

119.70 + 1.20

156.67 + 1.73

191.05 + 1.93

220.27 + 2 .05

142.47 + 2.19

(1)

95% confidence limits (+ 1.98 SE )

— m

Clam larvae ingested the resuspended intact cells of S_. obliquus
and showed somewhat better growth than the unfed controls . Larvae
receiving the broken cells of ground S_. obliquus , however, grew much
more rapidly than those receiving intact cells, although not as rapidly
as larvae receiving the mixture of live flagellates (Table 1). The

-9 0-

relative value of the different foods was the same in two separate
experiments, although in the second experiment all cultures, including
the unfed controls, grew more rapidly. Additional experiments in which
various methods of resuspension of the intact cells were employed gave
similar results; in every case the larvae receiving broken cells grew
more rapidly than those receiving intact cells. In subsequent experi-
ments, therefore, only ground S_. obliquus was used.

In a second series of experiments we tested simultaneously the
value of agitation and the use of the bacteriostatic agent, Sulmet, on
growth of clam larvae fed ground _S. obliquus (Table 2). In two experi-
ments growth of larvae receiving ground _S . obliquus in non-agitated
cultures was compared to growth of larvae in cultures on a wheel
designed to keep materials in suspension. One pair of cultures on the
wheel and one pair of non-agitated cultures received 66 ppm of Sulmet
with each change of sea water. Similar pairs of agitated and non-
agitated cultures were used to determine the rate of growth of larvae
receiving ground _S_. obliquus but no Sulmet. Cultures receiving living
algae as food and cultures receiving no supplemental food served as
controls .

Of the cultures receiving ground S_. obliquus as a food, larvae
in the agitated cultures that were Sulmet-treated consistently grew the
fastest (Table 2). Although either Sulmet alone or agitation alone did
improve growth in S. obliquus-fed clam cultures, growth in these cul-
tures never equalled that in cultures receiving the combined treatments.
At 12 days the average size of larvae receiving both the Sulmet treat-
ment and agitation was 179 microns in the first experiment and 187
microns in the second experiment. Although they were significantly
smaller than larvae receiving the mixture of live flagellates as food,
the larvae did reach setting size and metamorphosed between the 12th
and 16th days after fertilization with negligible mortality. The effect
of the retarded growth was merely to prolong the larval stage a few days

In yet another series of experiments the effects of several addi-
tional methods of agitation were tested. In addition to our original
method using the turbidity wheel, cultures were agitated by means of
aeration devices, an Eberbach bacteriological shaker, and paddles or
plungers on the machine described in the Methods section (Fig. 2).
Although these tests were not extensive enough to detect small dif-
ferences in efficiency, all but the aeration gave satisfactory results
when used with ground S_. obliquus and Sulmet. Aeration resulted in
variable growth and, in many cases, high mortality. When a timing
device was incorporated so that the paddle machine (Fig. 2) stirred the
cultures intermittently for three minutes every half hour, growth of the

-91-

to >->

CO

to

3

55

**•

o

*■*-

o

O

QJ

a

3

CD

3

3

o

O

r-l-

CD

DJ

i — •

i — '

*--

3

1

w

o

<

CD

1+

"I

1— ■

tO

•

o

CD

<£

oo

co

M

2
O

OJ

CO

•

l-h

2

<o

2

o

O

o

CD

o

or

o

I—*

t— -

a

CO

Cu

CO

1

c

r-f

c

c

3
O

1— '

3

CD

CD

3

3

CD

c
cn

CO

i-i-

$?

c

•i

I— ■

^— »^

o

3

I— •
i-1-

c

CD

<

c+

CD

vi.

ISO

cn

VI
VI

1 +
O

CO

to

CD

CO

Cn

1 +

o

to

Cn

to

O
1 +
O

vi

CD

CD

O
O

1 +

o

CO

00

CD

oo

1 +
CO

CD
to

CD
V]

to

OO

1 +

i>o
oo

CT>

cn

CO

1 +

(SO

oo

CO

C

CO

CO

CO

1 +

cn

CO

to

cn

00

In-
to

Cn
CO

cn

CO

VI

CO

1 +
to

|CO
O

rr

i — ■

c
c
cn

-i
o

c

3

CL

VI

CO

>

CO

iQ

00

1 +

I — '

CD

.

cl

CO

CD

1 — '

2

en

o

CO

3

Cn

t

DJ

co

«a

+

r-f

1 — •

DJ

#

c-t-

VI

CD

a-

00
VI

>

iQ

CO

c-l-

1 — ■

OJ

l-t-

l^-

CD

ISO

CL

1— '

i — •

s:

Cn

cn

o

3

VI

V)

£>.

O

1

cu

O

O

+

1 +

t-t-
OJ

to

CO

r+

•

•

CD

Cn

to

a.

CD

VI

to

o

c

I — '

c

"I

CD

f?

CD
QJ
r+

3

CD

3

X

■a

CD

—,
t-<-

3

CD
3

CD
DJ
3

CD
3

QJ

to

DJ
CO

o

-1

o

3
[fl

H

QJ

CT

CD

to

1 — '
DJ

2 s

■-i

CD

CD

3

QJ

QJ

3

QJ
3

t"t

i — ■

I—

CD

CD

CD

M

3

3

DJ

iQ

iQ

3'

r-f

3*

TJ

CO

cn

fD

~! 1

DJ

OJ

>-|

<-!■

I — '

CD

1— 1

r-t-
0)

QJ

to

"I

<

CD

a

o

•t

CD

l-h

OJ

^

I — ■

uQ

cn

DJ

CD

3

W

O

l-h

OJ

o

o

1 — '

l-h

n

a

ST

c

c

3

i — '

Xi

c

1 — '
h-1-

oT

CD

o

QJ

V

c-l-

OJ

■a

CD

CD

CD

-i

O

h--

a

QJ

c

1 — '

3
O

^

c

C

-1

H— *

CD
03

c

CD
cn

-a

-i

CD

CD

CD

O

CL

CD

3

I-*-
<

ua

i-*-

i-l

OJ

3
id

r+

CD

=>

l-h
l-h

OJ

w

CD
CD

o

3
it-

o

1— »

l-h
O

o

3

a

cn

o

OJ

l-h

3

n

a

OJ

r*

o

-i

7T

CD

a

CD
OJ

3

o

CD

CD

3

h-1

r-r

H—

in

c
1

-92-

S_. obliquus-fed larvae was accelerated and there was considerably
less coagulation of the suspended food than with any method of con-
stant agitation.

In a single experiment to determine the optimal feeding rate of
ground S_. obliquus, cultures were agitated and 66 ppm of Sulmet was
given with each change of sea water. One pair of cultures was used
to test each of the following volumes of packed food per liter of larval
culture per day: 0.005 ml, 0.01ml, 0 . 02 ml, and 0. 03 ml . In
general, 0.01 and 0.02 ml of food per day provided the best conditions
for growth of clam larvae. Those receiving only 0.005 ml of food per
day grew at a somewhat slower rate, and concentrations above 0.02
ml per liter per day produced excessive debris. The larvae given 0.01
ml of packed food per day grew as rapidly as those receiving 0.02 ml
of food, until about the 8th day of feeding. From the 8th day to meta-
morphosis, however, 0.02 ml of food per day was required for optimal
growth. A graduated feeding rate in which the concentration of food is
gradually increased from 0.01 to 0.02 ml of food per day has since
produced uniformly good results.

Dunaliella euchlora

In two experiments the food value to clam larvae of D . euchlora ,
both as living cells and as a lyophilized product, was compared
(Table 3). In these experiments 100 ppm of Sulmet was used through-
out and the paddle apparatus (Fig. 2) was used to agitate cultures.

In both trials larvae ingested the dried algae and grew and sur-
vived at least as well as larvae fed the living D . euchlora . The mean
lengths that larvae attained at 12 days of age when fed the dried D.
euchlora (161 to 170 microns) were comparable to growth attained
routinely by larvae receiving D. euchlora (Figs. 3 and 4). This growth
rate is well below that normally attained by clams when fed other,
more desirable species of algae, such as the naked chrysophytes , but,
nevertheless, demonstrates that the process of lyophilization, in the
case of D. euchlora, did not act to lessen its food value to clam larvae.
Also, the lyophilized D. euchlora was mechanically satisfactory in
that it was readily resuspended from the dried form as intact single
cells and did not produce debris in clam cultures.

In the second experiment (Table 3) clam larvae receiving live
D. euchlora failed to grow. This, almost certainly, was the result of
the D. euchlora culture becoming heavily bacterized and, thus, toxic
to larvae. Larvae receiving live algae were significantly smaller than
those receiving no food (Fig. 5), whereas larvae receiving dried D.

-93-

Table 3. Mean lengths at 12 days of clam larvae in cultures receiving
live and freeze-dried D . euchlora and no supplementary
feeding. Mean lengths are averages of duplicate cultures at
each treatment .

Mean length at 12 days (microns)

Culture treatment

Agitated

Non-agitated

Experiment 1

Dried D . euchlora
0.01 ml/day

Live D . euchlora
0.01 ml/day

No food

Experiment 2

Dried D . euchlora
0.01 ml/day

Dried D. euchlora
0.02 ml/day

Dried D . euchlora
0.03 ml/day

Live D . euchlora
0.01 ml/day

No food

163. 60+3. 88( 161.85+ 2.09

150.85+3.05 159.15+3.41

139.90+1.78 132.20+ 1.04

169 .65+ 1.71

170.40 + 3.13 168.55+3.94

169.75+ 2.12

118.80 + 1.21 123.65+ 2.11

137.15 + 1.92 133.20+ 1.18

(1)

95% confidence limit (+ 1 .98 SE )

— m

-94-

ifQx0**

Fig. 3. Larvae receiving lyophilized D . euchlora, 12 days of age

v^y

\xsgS

Fig. 4. Larvae receiving live D . euchlora, 12 days of age.

Fig. 5. Unfed larvae, 12 days of age

-95-

euchlora grew well. Since the algae used in preparing the dried material
were harvested before the experiment began, whereas the living algae
were harvested throughout the experiment, it seems probable that the cul-
ture became contaminated only after the algae for the dried preparation
were removed. It is possible, however, that the algal culture was toxic
at the time the algal sample was taken for drying and that the lyophili-
zation process in some way acted to detoxify the food .

Agitation of cultures did little to increase growth of clam larvae
when used with dried D. euchlora . The mean lengths of 163 and 170
microns attained in agitated cultures in the two experiments were not
significantly different than the 161 and 168 microns mean lengths attained
in comparable standing water cultures. On the other hand, agitation of
cultures was actually somewhat detrimental when used with live D .
euchlora . In the second experiment dosage rates of dried algae between
0.01 ml/day and 0.03 ml/day produced no difference in the rate of growth
of clam larvae .

Isochrysis galbana

In two experiments larvae receiving the dried _I. galbana grew
somewhat slower than those receiving live I_. galbana but, nevertheless,
grew well enough to produce uniform setting with negligible mortality at
the end of the 10-day culture period (Table 4). Several cultures of clams
receiving the dried algae achieved mean lengths of over 200 microns after
10 days of feeding . This is the rate normally achieved with the best live
foods under optimal culture conditions . Larvae ingested the dried parti-
culate material well and took on the brown coloration of _I_. galbana
indicating a direct usage of the dried food. This may be seen as the dark
areas of the digestive diverticula of the larvae in Fig . 6 .

Agitation of clam cultures fed dried I_. galbana produced variable
results . In the first experiment cultures receiving dried J_. galbana plus
agitation reached a mean length of 184 microns, which was significantly
larger than the 172 microns mean length of larvae in non-agitated cultures.
In the second experiment, however, similar agitation was actually detri-
mental to larvae receiving the dried food. Larvae reached only 165 microns
in these cultures, whereas larvae in non-agitated cultures reached a size
of 204 microns. Again, agitation of cultures receiving live food lessened
larval growth somewhat in both experiments (Fig. 7). Mortality was
negligible in all cultures.

In the second experiment we tested three rates of feeding of the
dried I_. galbana, from 0.01 to 0.03 ml of packed food volume per day.
Clam larvae receiving 0.02 ml per liter of culture per day showed the

-96-

Table 4. Mean lengths at 12 days of clam larvae in cultures receiving
live and freeze-dried_I. galbana and no supplementary feed-
ing . Mean lengths are averages of duplicate cultures at each
treatment .

Mean length at 12 days (microns)

Culture treatment

Agitated

Non-agitated

Experiment 1

Dried ]_. galbana
0.02 ml/day

184.00 + 1.68 172.40+2.38

Live I_. galbana
0.01 ml/day

185.75 + 2.10

191.40+ 2.31

No food

139.90 + 1.78 132.20+1.04

Experiment 2

Dried I_. galbana
0.01 ml/day

177.25 + 2.29

Dried I_. galbana
0.02 ml/day

165.50 + 2.36 204.00 + 2.07

Dried I_. galbana
0.03 ml/day

193.25+3.85

Live I_. galbana
0.01 ml/day

205.05 + 2.25 217.85 + 2.42

No food

137.15 + 1.92 133.20+1.18

(1)

95% confidence limits (+ 1.98 SE )

— m

-97-

Fig. 6. Larvae receiving lyophilized I_. galbana, 12 days of age

Fig. 7. Larvae receiving live I_. galbana, 12 days of age

most rapid growth (2 04 microns mean length at 12 days). Growth of the
larvae was slower at feeding rates of 0.01 and 0.03 ml per liter per day
This agrees with the results of Davis and Guillard (1958), who tested
various concentrations of live I. galbana .

-98-

DISCUSSION

It has been shown in this work that dried preparations of the
chlorophytes, S. obliquus and D. euchlora , and the chrysophyte,
I_. galbana , under the proper culture conditions may be used to rear
clam larvae to metamorphosis with little difficulty and with negligible
mortality. From this type of study it may be argued that the larvae
are not feeding on the dried food directly but, rather, are attaining
growth by feeding on heavy bacterial populations resulting from the
presence of large amounts of organic material. Although the use of
bacteria by clam larvae in these studies can not be excluded, at least
two lines of evidence suggest that the larvae utilize the dried algae
directly. First, the larvae receiving the dried food were observed to
ingest the material and to take on the coloration of the food in the
digestive diverticulum and digestive gland. This has always been a
characteristic typical of growing bivalves fed living algae. Opaque
areas can be seen within larvae receiving the dried and live foods
(Figs. 3,4,6,7). The second indication of direct utilization is the
fact that the rate of growth of clam larvae receiving lyophilized D.
euchlora and I_. galbana was comparable to that of larvae receiving
their live counterparts. D. euchlora, a generally mediocre food for
larvae as a live product, was also mediocre as a dried product, while ( '
I. galbana, which is a good food, was also good as a dried product.
If there had been any large utilization of bacteria, this relationship
probably would not have been maintained.

Dried unicellular algae possess desirable nutritive and physical
properties for foods for bivalve larvae and may, with suitable develop-
mental work, play an important part in the future of commercial shell-
fish hatcheries. All the unicellular algae tested, when resuspended
as discrete or unbroken cells, tended to remain so and did not clump to
form unwanted debris. Although in previous studies (Chanley and Nor-
mandin, 196 0) many types of organic materials were nutritious to clams,
nearly all included a great deal of unusable material which accumulated
as debris in the cultures . This hindered the rearing of larvae and abso-
lutely prevented the use of such materials as food for juvenile clams.
The accumulation of debris that takes place during the free-swimming
larval period can not be completely separated from the larvae by screen-
ing . Furthermore, the bottom-dwelling juvenile stages present a particular
problem since they are in close contact with the debris and normal functions
are impaired. Apparently the largely cellular nature of the dried algae acts
to reduce clumping. This was observed in experiments with S_. obliquus
where ground cells produced considerable debris but unbroken cells pro-
duced practically no debris . Clumping and settling of organic material
also encourage microbial growth that may be detrimental to growth and

-99-

survival of clams. The use of Sulmet has been helpful in minimizing
the effect of bacteria, but additional work is needed to develop other,
more effective agents .

Certain precautions must be taken in the preparation of a dried
unicellular alga if it is to be used successfully as a food for larval
mollusks . Many mass-produced and dried algae are either reared or
dried in contact with various metals . This may make these foods
unsuitable since metallic ions have repeatedly been shown to be highly
toxic to larvae in this laboratory. There is also the possibility of intro-
ducing toxic metals by various methods of resuspending the dried material
for feeding. The Waring Blendor, which very nicely resuspended S_.
obliquus as single cells, introduced toxic ions when the metal impeller
blade and housing were not first coated with plastic. The use of a Vibro-
Mixer with glass parts in contact with the food suspension avoided this
difficulty by resuspending the dried algae without danger of introducing
harmful metallic ions .

The development of a non-living food for the highly sensitive
larvae of the American oyster, Crassostrea virginica, presents a major
problem at this time. In preliminary experiments lyophilized preparations
of J_. galbana , which is an excellent food in the live state for larval
oysters, gave little or no growth of oyster larvae. At very low feeding
rates some growth of oyster larvae has been attained, but as the feeding
rate is increased, growth is correspondingly decreased. This indicates
the existence of conditions adverse to oyster larvae rather than any
inability on their part to ingest and utilize the dried algae.

In summary, dried cells of three species of unicellular algae
appear to be directly nutritive to clam larvae and, under suitable con-
ditions, possess the physical properties desirable in artificial foods for
the culture of shellfish . Problems to be overcome include the develop-
ment of methods of culture of juvenile clams and larval oysters with
these foods .

ACKNOWLEDGMENTS

The authors wish to express appreciation to Dr. Victor L.
Loosanoff for obtaining the dried _S . obliquus from the Micro Algae
Research Institute, Kunitachii, Tokyo, Japan and also to Dr. Hiroshi
Nakamura for preparing the material. Appreciation is also extended to
Mr. Joseph F. Lucash for construction of the apparatus in Fig. 2, to
Mr. Manton L. Botsford for photographs, to Mr. Russell E. Clark for
laboratory assistance in the initial phases of this work, and to Mr.
Harry C. Davis for general assistance.

-100-

LITERATURE CITED

Carriker, M. R. 1956. Biology and propagation of young hard clams,
Mercenaria mercenaria . J . Elisha Mitchell Soc . 72:57-6 0.

Chanley, P. E. and R. F. Normandin. 1960. Preliminary studies of
preserved foods for larvae of bivalve mollusks. (Unpublished
address given at N .S .A. Convention, August 1960).

Davis, H. C. 1960. Effects of turbidity-producing materials in sea
water on eggs and larvae of the clam (Venus (Mercenaria) mer-
cenaria). Biol. Bull. 118:48-54.

Davis , H . C . and R . R . Guillard . 19 58. Relative value of ten genera
of microorganisms as foods for oyster and clam larvae. U.S.
Fish Wildl. Sen/., Fish. Bull. 136 . 58: 293-304.

Davis, H. C. and R. Ukeles . 1.961. Mass culture of phytoplankton
as foods for metazoans . Science 134:562-564.

Loosanoff, V. L. and H. C. Davis. 19 50. Conditioning V. mercenaria
for spawning in winter and breeding its larvae in the laboratory.
Biol. Bull. 98:60-65.

Tamiya, Hiroshi . 19 57. Mass culture of algae. Studies from Tokugawa
Inst. VII: 3 09-334. (Reprinted from Ann . Rev., Plant Physiol.,
Vol .. 8 . )

-101-

EXPERIMENTAL FARMING OF HARD CLAMS, MERCENARIA
MERCENARIA, IN FLORIDA1

R. W. Menzel and H. W. Sims
Oceanographic Institute, Florida State University

ABSTRACT

In 1961 and 1962 clams or quahogs, Mercenaria mercenaria,
were planted in Alligator Harbor, Florida, to test the feasibility ot
commercial clam farming. Plantings were in concentrations of 10 to 75
clams 33 to 44 mm long per square foot, in plots of 100 to 625 square
feet, enclosed in fences to exclude predators. At concentrations of 75
per square foot clams increased in length only 4 to 5 mm in 7 months
(0.6-0.7 mm/mo) but at concentrations of 10 to 50 per square foot
growth was uniformly good, 1.4 to 1.7 mm/mo. Mortality was from less
than 5% to about 18% in fenced plots, while 100% of planted clams on
unfenced plots were killed. Over 90% of the shells of dead clams were
cracked by blue crabs. Smaller clams (less than 10 mm long) had 100%
mortality in 1 month even on fenced plots, attributed to effects of ship-
ment from Connecticut. About 80% of the 37-44 mm clams reached market
size of 50 mm in 8 to 10 months, average length being 52-56 mm at
harvest.

INTRODUCTION

In recent years, the potentialities of farming the sea, called
"mariculture" here, have received increasing emphasis. In this
country the sea is being farmed to a limited extent in the oyster indus-
try, but most fisheries are engaged in garnering, not in producing,
seafood. Mangelsdorf (1961) discussed six possible solutions to the
problem of lack of food in the world. His primary solution was "improving
agriculture" and next in order of importance was realizing more food from
the sea. Recently Menzel (1961a) discussed shellfish mariculture.

This paper covers one phase of farming of the sea, the rearing
of quahogs or hard clams, Mercenaria mercenaria . The techniques
developed by Dr. V. L. Loosanoff and his colleagues for laboratory-
rearing of mollusks (Loosanoff, 1954) have made it possible to assure
a supply of "seed" clams . Clams are used extensively as a luxury
food, and the demand would seem to encourage the rearing of clams
under controlled conditions.

1 Contribution No. 202, Oceanographic Institute, Florida State Univer-
sity.

2 Present address: Florida State Board of Conservation.

-103-

Much of the necessary basic scientific knowledge is available.
Remaining to be developed are techniques of application and financial
investment. It has been shown previously (Menzel, 1963) that trans-
planted laboratory-reared northern quahogs , initially less than 0.2
inch (3 mm) long will reach a market size of 2 to 2.5 inches (50—63.5
mm) in two years and that mortality is negligible if the clams are pro-
tected from predators .

We wish to thank Dr. V. L. Loosanoff and his associates for
the original laboratory-reared clams that were used to demonstrate that
clams would grow well in this area, and for other aid. We also wish to
express our appreciation to: The Fire Island Sea Clam Company, Inc.,
whose financial support made the present study possible; Mr. Don
McKee who sold those clams that were harvested, thereby providing
us additional capital to continue the project; and the Florida State Board
of Conservation who encouraged and aided us in this work.

METHODS

In Alligator Harbor, Florida, pens were constructed varying in
area from 100 to 125 0 square feet. The pens were situated on a firm
sand-mud flat about 100 feet from the low water mark in water from one
to one and one-half feet deep at low water. Six-foot-high fences were
constructed of one-half inch mesh plastic -coated wire or nylon netting,
either around wooden piling or fastened to metal frames. Care was
taken to sink the fencing material several inches in the substratum to
prevent the entrance of predators, which we had found to be very des-
tructive to unprotected clams. Despite the fences, blue crabs, Calll -
nectes sapidus , entered the pens; most of these were removed with
baited traps .

The clams were planted at concentrations of 10 to 75 per square
foot in the fenced areas. Unprotected plots of 25 square foot area were
planted as controls on 10 October and 9 November 1961 at densities of
10, 25, and 75 per square foot. Random one-square-foot samples
(enough samples in each planting to equal at least 1% of the area) were
examined from each planting at approximately monthly intervals. The
clams were measured to the nearest 0.5 mm in length and the number
dead was recorded .

RESULTS

Growth of the clams of each planting is shown in Table 1 . All
had an average growth rate better than one millimeter per month except

-104-

o

s-

(D
+->

c

fO

Dm
CO

6

fO

O

+->
rg
E

■ r-t

x
o

u

a.
a

<

<u
a

S

s

<o ti.

CD C

> 0)

< -J

10

Q

CD

0) C

> (1)

< -4

(0

4-1

o

H

CO

c

-r-l
-t-»

c
(0

<3*

in

LO

CNI

to

co

0)

4-"

(0
P

o

a)
-a
o

O

IX)

LO
LO

CM
CD

CO

en

LO

CO

«tf

o

CD
LO

CN]

CD

00

<NI
LO

CD

JD

T3

(0

-a

CM

CD

CD
\

CD

CM
LO

CM

CD

>*

CO

CO
CO

LO

CM

CD

CD
\

CD

CNI

CD

CD
\

CD

CM

CD

CO
\

CO

CO

o

CO

CD

CO
\

o

CO

\

CO

i—l I— I CO

CD
\
CD

CO
\

CD

CD

0Q

00

00

o

< CQ

I— I ►— I (— I 1— » •"*>

►— I I— I I— I I— I EL

1—1 "

oo

CO

CM

CD

CD
\

CD

•

o

LO

O

LO

LO

O

o

LO

00

LO

o

• — i

CM

i — 1

CM

CNI

OO

LO

c-^

^r

r--

£

LO

LO

LO

O

LO

O

O

o

CNI

LO

t^.

CM

CNI

CNI

LO

CO

O

o

o

i — 1

<NI

CD

CO

CD

CD

CM

CO

i—l

I—l

LO

CO

CM

CO

CD

O

-105-

those planted at the highest concentrations of 75 per square foot. This
growth rate compares favorably with that obtained in experimental
baskets (Menzel, 1963). In the control (unprotected) plots all clams
were killed before they had time to show much growth .

Mortality in all pens ranged from less than 5% to 18%, except
for planting IIB . Clams planted in I IB were under 10 mm in length and
were planted in August after two days in transit from their source,
without refrigeration. Over 15% of the clams were culled out as dead
before planting; other clams were probably dead when planted but death
was not detected. In the pens there were small crabs that had not been
removed by trapping, and these undoubtedly destroyed many of the small
weakened clams that might have survived otherwise. Plantings IA, IB
and III were harvested in March, April and May, 1962. Of the almost
25, 000 clams planted, 18, 000 were sold . It is estimated that over 2 0%
of the clams had not reached market size. The overall mortality was
under 10%. The unprotected clams had 100% mortality. Examination
of the shells indicated that predation was the reason for the mortality
and over 90% of the shells examined were cracked, indicating blue
crab predators . A few of the clams were killed by lightning whelks,
Busycon contrarium.

It was found that circulation of water in the pens was retarded
by the fencing . After several months the bottom accumulated a "gooey"
mud one-half to one inch deep. It is assumed that the mud was caused
by the biological activities of the planted clams, coupled with the
retarded circulation, because unplanted pens did not have this mud.
As far as we know, this did not affect the clams; at least, there was
no mass mortality resulting from it. However, the effect of a heavy-
deposition of silt may be very detrimental. This was illustrated by a
dramatic but unwanted "experiment" which took place in June, 1962.
Dredging operations for a boat basin and channel adjacent to the clam
pens were begun about the second week in June, filling the water with
silt . Loose silt had settled up to eight inches deep near the inshore
edges and from two to four inches in most other areas by the July samp-
ling. Sampling showed 100% mortality on the inshore edges and an
overall mortality of from 5 0% to 75%, with the least on the outer edges,
where very little silt had settled.

DISCUSSION

The excellent growth of clams when not planted too densely and
the low mortality indicate that there is a possibility for clam farming
in these areas. We would like to suggest some prerequisites and pre-
cautions:

-106-

1 . Available seed clam supply . Small clams can be obtained
from the culls of commercial clam fishing but this source would not be
reliable because of variations in natural set and various local regula-
tions. Clam hatcheries would be the more feasible and economical
source of seed clams (Loosanoff and Davis, 1963).

2 . Suitable planting area . There are millions of acres of
suitable bottom having the proper salinity, depth, and firmness of sub-
strate, not only in Florida, but in other coastal states as well. The
following conditions should exist: salinity above 25 ppt; bottom firm
enough to hold the clams; areas protected from storms; shallow water;
not over two feet at low water, for ease of operation and erection of
fences; and good growing and low mortality conditions. Some indica-
tion of the suitability of an area can be obtained from a superficial
examination of the biota but it is recommended that trial plantings be
made and observed for growth and mortality for upwards to a year,
especially during the warm season. For example, a trial planting was
made in an area about ten miles from Alligator Harbor in the spring of
1962. Conditions seemed ideal; the bottom was firm, the salinity was
25-30 ppt and the planting was on a large shallow flat . For the first
two or three months growth was excellent with negligible mortality.
This planting was not examined regularly. By late August, there had
been a cumulative mortality of 50%, not from predation. It is suspected
that very high temperatures in July and August, accompanied by very
low daytime tides, was the cause of death . The water was shallower
than at Alligator Harbor and very shallow water extended for great dis-
tances, allowing high water temperatures to occur.

3 . Protection from predators . This is absolutely necessary, at
least in our area of Florida . It has been repeatedly observed by us
that unless clams are protected from predators mortality is nearly 100%
within a few months .

4. Constant attention. Unless frequent examinations are made,
with attention given to rectifying adverse conditions, no attempted
clam farming will be successful. Despite precautions, predators will
enter the pens and should be removed . If mortality occurs information
should be obtained to take precautions against its recurrence.

Judging from our limited observations, clams should be at least
one-half inch (12.5 mm) long before they are planted in pens, other-
wise there is too much danger of mortality from small predators. The
problems of rearing clams in a hatchery from a length of about 1 mm to
the proper planting size are considerable because of the space and the
large amount of food needed. The use of dried algae (Hidu, 1964) may

-107-

alleviate the food problem somewhat. We have found that it is possible
to grow clams of less than 5 mm initial size up to 10 to 15 mm in con-
centrations of up to 500 per square foot in sand-filled one -fourth -inch
mesh wire-covered boxes in the open water. The expense in labor and
materials is considerable for large scale operations.

Seed clams have not yet been produced on a commercial scale
from a hatchery or from natural sets . The development of genetically
faster -growing clams merits attention (Chanley, 1961). The use of
hybrids between the northern and southern species (M. campechiensis),
which are faster growing than the northern species, is another possi-
bility (Haven and Andrews, 1957; Chestnut , Fahy and Porter, 1957;
Menzel, 1964). Although the southern quahog is a fast growing clam,
it does not keep well out of water. Preliminary experiments by us
showed that the southern species gaped in less than a week in a house-
hold refrigerator. The hybrids between the two species "kept" up to two
weeks, although not as long as the northern species . It would be more
economical to harvest by machines, such as the escalator dredges
already used commercially, and it would be an advantage to develop
by genetic selection clams that grow at a uniform rate to eliminate cul-
ling . Fencing is one of the more expensive items in clam farming, and
the more crowded clams can be grown, the more economical it will be.

CONCLUSIONS

In conclusion, we feel that the cultivation of clams will become
practical. Most of the basic techniques have been established on a
pilot-plant scale and await application.

LITERATURE CITED

Chanley, P. E. 1961. Inheritance of shell markings and growth in the
hard clam, Venus mercenaria . Proc . Nat'l Shellfish . Assoc .
50: 163-169.

Chestnut, A. F. , W. E. Fahy and J . J . Porter. 1957. Growth of young
Venus mercenaria , Venus campechiensis and their hybrids .
Proc . Nat'l Shellfish . Assoc .47: 50-56 .

Haven, D. and J. D. Andrews. 1957. Survival and growth of Venus
mercenaria , Venus campechiensis, and their hybrids in sus-
pended trays and on natural bottoms. Proc. Nat'l Shellfish.
Assoc . 47:43-49.

-108-

Hidu, H. and R. Ukeles . 1964. Dried unicellular algae as food for
larvae of the hard shell clam, Mercenaria mercenaria . Proc .
Nat'l Shellfish. Assoc. 53:85-101.

Loosanoff, V. L. 1954. New advances in the study of bivalve larvae
Am. Scientist 42:607-624.

Loosanoff, V. L. and H. C. Davis. 1963. Shellfish hatcheries and
their future . Comm. Fish. Rev. 25:1-11.

Mangelsdorf, P. C. 1961. Biology, food and people. Econ . Bot .
15:279-288.

Menzel, R. W. 1961. Shellfish mariculture . Proc. Gulf and Caribb .
Fish. Inst. 14th Ann. Sess .: 195-198.

Menzel, R. W. 1963 . Seasonal growth of the northern quahog, Mer-
cenaria , and the southern quahog, M_. campechiensis , in
Alligator Harbor, Florida. Proc. Nat'l Shellfish . Assoc . 52:
37-46.

Menzel, R. W. 1964. Seasonal growth of northern and southern

quahogs, Mercenaria mercenaria and M_. campechiensis , and
their hybrids in Florida. Proc. Nat'l Shellfish. Assoc. 53:
111-119.

-109-

SEASONAL GROWTH OF NORTHERN AND SOUTHERN QUAHOGS ,

MERCENARIA MERCENARIA AND M. CAMPECHIENSIS, AND

THEIR HYBRIDS IN FLORIDA

R. Winston Menzel

1

Florida State University'

ABSTRACT

Growth (increase in shell length) of the northern quahog, Mer-
c en aria mercenaria, the southern quahog, M. campechiensis, and their
reciprocal hybrids was measured monthly for 19 months in Alligator
Harbor, Florida. The fastest growth for the four groups of clams occurred
during the spring and fall. The northern quahog had the least overall
growth (8 mm clams growing to 41 mm average length) and showed least
growth during the hottest period of the year. The southern quahog had
the best overall growth (growing from 8 mm to 57 mm in average length)
and had the least growth during the coldest period. The growth of the
hybrids was intermediate between that of the two parents but closer to
the southern parent than to the northern parent, 8 mm clams growing to
average lengths of 52 and 55 mm in 19 months. The growth rate of the
hybrids was better than that of either of the parents in late spring and
early fall.

INTRODUCTION

In the past several years there has been increased emphasis on
the study of the hard clam called the quahog . The northern quahog,
Mercenaria mercenaria , has been successfully hybridized with the
closely related southern species, M_. campechiensis , by Loosanoff (19 54)
and his associates at the Milford, Connecticut, Shellfisheries Labora-
tory. Haven and Andrews (1957) and Chestnut, Fahy and Porter (1957),
using seed clams reared at Milford, observed the comparative growth of
the two species and their hybrids in Virginia and North Carolina . The
present paper concerns the monthly growth rate of clams of the two
species and their hybrids in Alligator Harbor, Franklin County, Florida.
I wish to thank Dr. V. L. Loosanoff and his associates for the laboratory-
reared seed clams and for their continued interest. Dr. R. G. Cornell,
Department of Statistics, Florida State University, kindly programmed the
data for the IBM 709 computer, supported by NSF Grant G17467, and
aided in the analysis .

1 Contribution No. 2 01, Oceanographic Institute, Florida State University

-111-

METHODS

Laboratory -reared northern and southern quahogs, and off-
spring from male northern X female southern and female northern X male
southern quahogs, v/ere planted in sand -mud bottom about 10 inches
below mean low water in Alligator Harbor and covered by wire cages .
Initially the clams were between 5 and 10 mm in length. At monthly
intervals (around the middle of the month) lengths were measured to the
nearest 0.5 mm. Growth has been calculated as previously described
(Menzel, 1963). The experiment began in November 1960 and was ter-
minated at the end of June 1962 . Growth data were analyzed on an IBM
709 computer for the regression coefficient (slope). Seasonal differences
of the four groups were considered, reducing the standard error. After
obtaining the slope and the standard error the groups of data were com-
pared by the standard normal deviate.

RESULTS

Fig. 1 shows the growth of the two species and their hybrids.
M . campechiensis grew more rapidly than M_. mercenaria or the two
hybrids . Growth rate of the two hybrids was intermediate between the
rates of the parents, but closer to tnat of the southern parent than to
that of the northern one . Fig . 2 shows the average daily growth rate of
the two species and of the two hybrids. Generally speaking, both
species and their hybrids grew best in spring and fall. Table 1 gives
the number of clams measured each month, average sizes, and ranges.

The statistical analysis given in Table 2 corroborates the data
as given in Fig . 1 . The slopes of M_. campechiensis and of the two
hybrids are clearly different from that of M_. mercenaria . The two
hybrids are not significantly different from each other. The two hybrids
are closer to the southern parent but analysis shows that they differ on
the 99% confidence level. The standard error may have been under-
estimated because the same clams in each group were measured each
time and no records of individual clams were kept; however, the analysis
used could have had a much greater standard error and still have given
significant differences .

The daily growth rate (see Fig . 2) of the northern species was
above that of the others from the middle of January through the middle
of February 1961, from the middle of January through the middle of April
1962, and from the middle of September through the middle of October
1961. This species showed no increase in shell length when measured
in mid-August 1961.

-112-

60

30-

45-

40-

E

x35

I-
O 30

z
w25

_i
20

10

/

M «lftmorlo

X X M compichlimli

? M. m.rc.norla X c*M camp.chl.n. I.

X--X d*1 "iremorlo X ? M c am p.chi.n. i .

1

-1-

I I

1 ' '

_i_

■

I I

N 0 J r MAMJJA80N0 J F MAMJJ

•960 I 9 6 I 1962

MO NT H S

Fig . 1 . Monthly size of M_. mercenaria, M. campechiensis
and their hybrids .

-113-

Fig . 2 . Da ily growth of M_. mercenaria , M_ . campechlensls
and their hybrids .

-114-

co

C
CD

x:
o

CD
Q

s

a

o

2

5

.. n

m

C

-.-*

01

\~

o

<U

i~

c

a;

CD

e

(>

li

•

(i)

6

21

CM
ID
CD
i— I

CD

c
3

. a) lo

1 P O

° XI O

w G co

^ B en

CD ■—!

CD

to

B e

Z? CD

?^ x:
6 o
a)

>-■

0)

E
cd
>

o

Jigs

m • E

£S s

c cd
ro

-a

ro O
_T E i_

X. CD

*-* 44 «

0> * e'-
en

c

1-1

3
T)

to

en
C

0)

C f0

® t,

(0

CD C

Cn CD

(0 O

I-. l-i

CD CD

> E

(0

CO 2|x

CD CD £

X) T« D.

| 1 f

•5 CD <°

CD

ro
H

co

C
CD

-r-1

x:
u

(D

E

ro

o

CD

c

ro
OS

CD
CJi
(0
u
CD
>
<

o

CD

en

c

rC
OS

CD
Di
ro

!_

CD
>

<

O

2

c

o

o

1/5

O

o

o

o

o

LO

o

LO

o

LO

o

o

o

o

LO

o

i— 1

CM
i—l

LO
i—l

LO

CO

1 1 ~>

o

o

<3*

1—1

00

1— 1

LO

1—1

LO

CM

LO

LO
LO

LO

1—1

CD

LO

CD

o

O

LO

LO '

1 ■ o

o

LO

LO

o

LO

LO

LO

o

o

LO

o

o

LO

cd

i—l
i—l

i—l
i—l

CO
CM

CD
CM

CO
CM

cd

CM

o

CO

00
CO

00

co

00
CO

CD
CO

1— <

co

00

1— 1

LO

o ^
00 i-i

o ai

LO t

o

LO

•

•

o

CO

1— 1

. — 1

o

LO

O 00

oo o

o o

LO LO

o

CD
CD

00 CM
CM CO

I I

I I

CD LO CO •*? CD 00 CD

I CO i-l ■— I CD CD CO LO
CM CO CO CO CO ^J* ^

CD CO I

I LO LO LO LO LO LO <3*
I ^ **J* *3* ^ ^T ^ *3*

OOOOOOLOOOOOO

^lomo^iomONOr ii— i
I I I I I I I I I I I i

loOlOlOOOlOloOlOOO

coooooj^oooo'-i'^rr^f^

i— ii— ii-Hi— ii— I i-l i— I CM <M CM <N

^TLOLOOCDCMOO^rcvlCMi— I

■— ICMCO^LOOOOOcoOCMcO
i— I ■ — I i — Ii— I i— I i-H CM CM CM CO CO CO

moimNNNiDmnconN

^J* ^^ ^* ^J* ^^ ^3* ^^* ^J* ^^ ^J* ^5* ^*

•

•

i—i

>

o

co

0

CD

CD

£

u

i — l

X3
CD
Dl,

ro

i-i

a
<

>, CD

ro C

-3 3

i? o> p. *t

•^ d (1) U
2.-5: M O 2 Q

>

o

co r- lo lo co ■*?

IO ID CO O ^ N

^r tt ^r lo lo lo

"<*■ ^r ^ "sf cm i-i

^J* ^J* ^J* ^J* ^J* ^J*

o o o o LO o

i—l i—l CM CO t*» CD

^3* ^^ ^* *>1* ^J* ^^

I I I I I I

LO LO LO LO O O

N N O H Tf Ifl

CM CM CO CO CO CO

■^< lo O r^ ^J* oo

CO CO CD 00 O O
CO CO CO CO ^ ^

CM i—l i—l i— I i—l
^j* ^3* ^3* ^J* *^*

T

CM
CD
CD

a *-
< 2

>■ CD

ro C

3

-115-

3 0) fc| qj
CD ■< . ~>

>- g

^ O Z O w >^
10 a> o o n> g c

N5 . . . r+ •

CO CO ,b ,b ,b ,b
CD 00 O O O O

U1 ib ^ ib >b ib

w (£ ai en w w

O M O W W M

b ib b u u u

Ol N O 00 ib W

en en en en o O

I I I I I I

CD en en en ,b ib

o oo en kj to eo

en en o O O CD

lb lb lb lb lb lb
,b ib ib en en en

en en ,b ib ib ib

b M sj Ji to W

-v] to ib 00 CD -v]

lb

.b

.b

lb

CO

CO

en

ro

i — «

o

CO

VI

en

en

o

o

o

o

en

en

en

en

en

lb

i— i

■vi

lb

to

o

uo

o o en o o en

2 >

cr 3

1— '

n

£

eo

m

o

en

n

<

i— '

•

^,bibibibibibibibibiben

oot^totocococnooooooo

ibibUUlONNHHHMH

NOOlNJsJMUCOUlibibU
•-JtOODMlDsllDvJSlUltOsl

WWNMMMHHHHHH
OOenencnoenenooenen

I I I I I I I I I I I I

^iblbibWUMWMMMM

eocneoi— 'cn^tDcooooovjcn
oenoenoenoocnoenen

bbiblblbibibibibibibib

encncncncncncncnoooocooo

►bibcoeotoooc-ot— 'i— > y- >i— 'i— •

NOOlUlOsjUCDCn.bUU

ooiiDsjoooMWMMrocn

UUNWMNMMHI-'HH

OlHlDUlUNM^UMMO

oenoooenoenenooen
I i I I I I I I I I I i

ibibbibWUtOtOHMMM

CDyibHOiwcoMiOMOiai
ooenoenenoenoooen

en en

o o

to oo

CO o

o en

en

o

1 — 1

1 — 1

en

o

•

•

en

o

ib en
eo o

to oo
to o

oo en

en

o

1— 1

1— 1

en

o

•

•

en

o

2

o

>
<

CD
►-I
QJ
iQ

JO

QJ
3
iQ
CD

o

>
<

CD
►1

OJ
iQ
CD

JO

QJ
3
iQ
CD

g -a

CD Q)

CD

O

CU

3
ra

CD
O

tr

fD

a

w

CD ^

3 »

QJ

CD

H

QJ

cT

o
o

a

c+

a
c

CD

-116-

Table 2. Regression coefficient, standard error and confidence limits
of the growth of M. mercenaria, M . campechiensls , female
M . mercenaria x male M.. campechiensls (female _M . m_. x
male M^. c .), and male M_. mercenaria x female M_. campechien-
sls (female JVL m_. x male M . c . )

Group

Regression
coefficient

Standard
error

99% confidence
limits

M . mercenaria 0.06131

M. campechiensls 0.08958

Female M^. m_. x

male M_. c. 0.08241

Male M^. m_. x

female M. c. 0.08344

0.00049 0.06005 and 0.06227

0.00064 0.08793 and 0.09123

0.00056 0.08208 and 0.08502

0.00057 0.08097 and 0.08385

The growth rate of the southern species exceeded that of the other
three groups except in the period from November 196 0 to February 1961, in
September and October 1961, and in June 1961 and 1962.

The growth rate of the two hybrids was greater than that of either
parent from time of planting in November 1960 through January 1961 and
also in June 1961 and 1962 and September 1961.

DISCUSSION

Growth rates of the northern and southern species during the present
study agreed closely with those reported previously (Menzel, 1963) for the
same locality. Both species had the best growth in spring and fall and the
annual growth of the southern species was greater than that of the northern
species. In both studies, the northern species continued growth during
the colder period but had little or no growth during the middle of the sum-
mer. The southern species had less growth than the northern in the winter
but did continue growing during the summer.

The hybrids in the present observations had essentially the same
seasonal pattern of growth, i.e., more growth in the spring and fall and
less in the summer and winter. The initial faster growth of the two hybrids
cannot be explained unless they received better treatment in shipment and

-117-

transplantation. As far as can be determined, all clams received
identical treatment. The faster growth rate of the hybrids in late spring
and early fall is very interesting. If it had happened only once, it
could have been the result of error in measurement or some other fac-
tor in the experiment itself, e.g., mortality of faster growing clams in
the two species but not in the hybrids. (But there was little mortality
during these observations, except from occasional predation.)

Chanley (1961) has shown that individual clams differ widely
in inherent growth rate. Our observations support Chanley' s findings,
e.g., growth experiments after a period of six months or so will show
some clams that have grown several hundred per cent more than others
in the same cage. At the termination of the present observations, the
hybrid of the female southern quahog and the male northern quahog was
slightly larger than the reciprocal cross; however, the growth curves
crossed several times (Fig. 1). The faster growth rate with the southern
species as the female parent probably is not significant.

It is possible that the difference in growth rate of the two species
and their hybrids was not a specific trait at all but was a result of the
genetical constitution of the particular parents . This would seem un-
likely, however, and the previous observations (Menzel, 1963) also
showed that the southern species had the greater annual growth rate.
The observations of Haven and Andrews (19 57) and especially Chestnut,
Fahy and Porter (19 57) corroborate that the southern species has the
fastest growth .

The possibility of producing faster growing clams for mariculture
(Menzel, 1961, 1963) merits attention. Thurlow Nelson (1949) com-
mented years ago that techniques for rearing mollusks in the laboratory
give us every reason to hope that faster growing clams and oysters can
be produced. Loosanoff (1954) discussed the possibilities of producing
strains of mollusks not only with faster growth but also with higher
glycogen content and resistance to disease.

LITERATURE CITED

Chanley, P. E. 1961. Inheritance of shell markings and growth in the
hard clam, Venus mercenaria. Proc . Nat'l Shellfish. Assoc.
50: 163-169.

Chestnut, A. F., W. E. Fahy and H. J . Porter. 1957. Growth of young
Venus mercenaria, Venus campechiensis , and their hybrids.
Proc. Nat'l Shellfish. Assoc. 47:50-56.

-118-

Haven, Dexter and Jay D. Andrews. 19 57. Survival and growth of

Venus mercenaria, Venus campechlensls, and their hybrids in
suspended trays and on natural bottoms. Proc . Nat'l Shellfish.
Assoc. 47:43-49.

Loosanoff, V. L. 1954. New advances in the study of bivalve larvae.
Am. Scient. 42:607-624.

Menzel, R. W. 1961. Shellfish mariculture . Proc . Gulf and Caribb .
Fish . Inst . 14th Ann . Sess . : 195-199 .

Menzel, R. W. 1963. Seasonal growth of the northern quahog,

Mercenaria mercenarla, and the southern quahog, M_. campechi-
ensis, in Alligator Harbor, Florida. Proc. Nat'l Shellfish .
Assoc . 52:37-46 .

Menzel, R. W. and H. W. Sims. 1964. Experimental farming of hard
clams, Mercenarla mercenaria, in Florida. Proc. Nat'l Shell-
fish. Assoc. 53:103-109.

Nelson, T. C. 1949. What can science offer the oyster grower.

Convent. Address Nat'l Shellfish. Assoc. 1949 (Proc. Nat'l
Shellfish Assoc. 40:1-9).

-119-

SEASONAL GONADAL CHANGES IN FEMALE SOFT-SHELL CLAMS,
MYA ARENARIA, IN THE TRED AVON RIVER, MARYLAND

William N. Shaw

Fishery Research Biologist, U.S. Department of the Interior,

Fish and Wildlife Service, Bureau of Commercial Fisheries

Biological Laboratory, Oxford, Maryland

ABSTRACT

Changes in the gonad of female soft-shell clams were observed
at regular intervals from May 1961 through May 1962. From May to
August 1961 ovaries consisted of follicular cells with many inclusions
and no gametogenesis was seen. In August gametogenesis began and
small ovocytes were observed developing. By late September most clams
had mature ova. Spawning began around September 27 and continued
through October. A second cycle of gametogenesis started shortly after
fall spawning was completed. Histological examination revealed that
ovocytes in many clams were half-grown by March 1962, but failed to
grow to maturity Most ovocytes underwent cytolysis, and by May 1962
clams were in the summer or inactive stage as in the summer of 1961.
Mature ova were found in a few clams, but no clam reached the degree
of ripeness found in autumn 1961. Lack of Mya larvae and set support
the conclusion from gonad examinations, that Tred Avon River clams
failed to spawn or to produce viable larvae in the spring of 1962.

INTRODUCTION

Seasonal changes of the gonad of many species of pelecypods
have been described, including the northern quahog, Mercenaria mer-
cenaria, by Loosanoff (193 7); the eastern oyster, Crassostrea virgin lea,
by Loosanoff (1942); and the European oyster, Ostrea edulis by Loosan-
off (1962). Only incomplete information is available on the seasonal
gonadal cycle of the soft-shell clam, Mya arenaria . Rogers (1959)
gave a brief description of the fall changes that occur in the gonad of
Mya, but he failed to complete the yearly cycle. Coe and Turner (1938)
describe in detail the development of the gonad of Mya, but mention
only briefly the seasonal changes.

It appears that north of Cape Cod Mya spawns once a year,
while south of Cape Cod it spawns twice a year. Landers (1954) found
peaks of abundance of Mya larvae in Narragansett Bay, Rhode Island,
in the spring and in the fall. He noticed that spawning ceased during
late summer. In Maryland, Pfitzenmeyer (1957) observed a successful
set of Mya in October 1956 and again in May 1957, and reported that

-121-

in previous years spring spawning was known to have occurred with
little successful setting, so that the annual set appeared to be a result
of autumn spawning. Manning (1957) reported that the Maryland soft-
shell clam sets during the autumn. In a more recent paper, Pfitzenmeyer
(1962) found setting in both spring and fall in 1957, 1958, and 1959. He
theorized that the two peaks of spawning of the soft-shell clam each
year in Maryland are the result of two separate periods of gonadal matura-
tion, and not one period interrupted by temporary cessation.

The purpose of this paper is to describe the seasonal cycle of
gonad development of the female soft-shell clam in the Tred Avon River,
Maryland. Histological study has shown for the first time that in Mary-
land there are two separate periods of gonadal development each year.
This study is part of a project concerned with the ecology and distribution
of clams and oysters in local waters .

METHODS

The Tred Avon River is about 10 miles long, beginning near the
town of Easton and emptying into the Choptank River which in turn flows
into Chesapeake Bay. In the mouth of the Tred Avon there are several
clam beds which are fished commercially with hydraulic dredges .

Salinity, temperature, and amount of dissolved oxygen (Fig. 1)
were measured in front of the Oxford laboratory in seven feet of water.
Temperature was recorded daily by a thermograph while salinity and
amount of dissolved oxygen were recorded weekly. Salinities fluctuated
from a low of 8 . 02 ppt on 26 July 1961, to a high of 15 .52 ppt on 26
December 1961. In general, lower salinities were found in summer than
in winter. During the first four months of 1961, the salinity averaged
2.25 ppt lower than during the corresponding period in 1962. Tempera-
tures ranged from a low of -0.5 C on 26 January 1961, to a high of 3 0.0 C
on 23 July 1961. Amount of dissolved oxygen followed a normal seasonal
pattern with highest readings occurring in winter (13.46 ppm) and lowest
readings in summer (4.55 ppm). No deficiencies in amount of dissolved
oxygen were found during the study .

Mya was first sampled 19 January 1961. Because the river was
frozen over, only sporadic samples were taken during February and part
of March. Starting in May sampling was done each week and continued
without interruption until 18 May 1962, when the first phase of this
study was terminated .

Each sample, consisting of ten clams two inches or more in
length, was collected from a shallow bottom opposite the Bureau of

-122-

o

•o

7*

UJ

o

s

nr<-0&coNO^<f »n <f n

I?

O a

>- a
X ~"
0

o o co o •* O m O *n O

— n a rt — —

5a

<

a.
S

>

S

c
o
>

<

CD

)-.
H

cu

4=

C

a)
X

o

CD
>

"o

w
en

C

o
S

(0

T3
C

(0

>^

c

-1—1

10
w

CD
u

fO

s-
CD

a

i-.

cd

(0

E -a

^ c

(0

(0

-123-

Commercial Fisheries Laboratory. The clam meats were fixed in Kahle's
fluid (Guyer, 1953), dehydrated in alcohol, changed to xylene, embedded
in paraffin, and sectioned at about 8 microns with a razor blade mounted
on a standard rotary microtome. Sections were stained in Harris1 or
Delafield's hematoxylin and eosin. A total of 607 clams were processed
over a continuous period of 17 months. Of these, 3 06 or 50.4% were
males, and 3 01 or 49.6% were females. No hermaphrodites were found.
In this phase of the study only the female clams were examined . The
seasonal gonadal changes of the male will be covered in a later publi-
cation .

Weekly plankton samples were taken at four stations in the Tred
Avon River from 23 June to 3 0 November 1961, and from 11 April to 3 0
June 1962, in order to determine when planktonic My a larvae first
appeared, when they were most abundant, and when they disappeared.
Bottle collectors like those used by Thorson (1946) were placed at the
same stations to determine when Mya was setting . These bottles were
replaced every 15 to 3 0 days from 28 June to 7 December 1961, and
again from 11 April to 16 July 1962.

RESULTS

Following Loosanoff (1942) , I have divided the seasonal cycle
into several stages: summer or inactive, active, mature, partially
spawned, and completely spawned.

Summer or Inactive Stage

Examination of female clams collected in the latter part of May
1961, and again in May 1962, showed each alveolus consisting of
follicular cells with many inclusions (Fig. 2). Those ovocytes present
were degenerating and undergoing cytolysis . Coe and Turner (193 8)
state that the inclusions are reserve nutritive substances derived in
part from the cytolysis of degenerating ovocytes. It was observed that
some inclusions stained bright blue and showed no traces of eosin,
while others were bright red . There is need for further investigation of
these bodies . The clams remained in the summer or inactive stage
until August .

Active Stage

Early in August active gametogenesis began. Small ovocytes
were observed at the base of the alveolar walls (Fig. 3). Gametogenesis
progressed at a rapid rate during August, and by the end of the month

-124-

Fig. 2. Summer or inactive stage. Follicle cells with dark
bodies called inclusions. No ovocytes present. June 1961. X 200

:!t

M

Fig. 3. Female clam with young ovocytes. August 1961. X 200

-125-

many ovocytes were half-grown and attached to alveolar walls by
stalks (Fig. 4). By the end of the first week in September a few ova
had reached maturity and appeared free in the lumen. The follicular
cells were then undergoing cytolysis and inclusions were much less
numerous. Coe and Turner (1938) state that these inclusions supply
the nourishment for the rapidly growing ova .

Mature Stage

In 19 61 the majority of clams examined were mature by the last
week in September or the first week in October. Each alveolus now
contained many mature ova which appeared to be free in the lumen or
attached to the alveolar walls by slender stalks (Fig. 5). The follicular
cells were gone and only a few inclusions were present. The mature
ova measured about 6 0 microns. Each ovum had a large nucleus and the
nucleolus (amphinucleoli, Wilson, 1925) was divided into two zones,
the smaller of which stained much darker.

Partially Spawned Stage

Evidence of spawning was first observed on 27 September. As
shown in Fig. 6, the number of mature ova present in each alveolus of
a partially spawned clam ranged from 0 to 8 . Those alveoli that no
longer had ova were collapsed .

Completely Spawned Stage

By the first of November all clams examined were spawned out.
Only a few unspent ova could be found . The alveoli had collapsed and
only a few inclusions were still present. In general, each alveolus
now consisted of a single row of cells (Fig .7). At the corresponding
stage in M. mercenaria , Loosanoff (1937) found few if any phagocytes
in the lumen. He believed that the undischarged ova showed no indi-
cation of being decomposed and suggested that these eggs are extruded
later in the normal manner. Mya appeared to be similar to Mercenaria
in this respect since no phagocytes were found after spawning and the
unspent eggs showed no indication of being broken down.

Second Cycle of Gametogenesis

Coe and Turner (1938) reported that during the autumn and winter
at New Haven, Connecticut, the gonad of Mya consisted of follicular
cells with an accumulation of inclusions. In Maryland, however, the

-126-

Fig. 4. Active gametogenesis . Ovocytes half-grown with
stalks. September 1961. X 200.

1*4

Br

Fig. 5. Ripe female clam. October 1961. X 200,

-127-

Fig. 6. Partially spawned out clam. October 1961. X 200,

Fig. 7. Spawned out female clam. November 1961. X 200

-128-

corresponding stage was observed during summer but not during autumn and
winter. Instead, shortly after spawning was completed in November,
small ococytes were developing along the base of the alveolar walls
(Fig. 8). Lacking, though, were the numerous inclusions that were so prev-
alent in the early stages of gametogenesis in August 1961 (Fig. 3).

By January 1962, the clams were found in different stages of
gonadal development. Many were in a stage similar to that observed
in November 1961 (Fig. 8) with only small ovocytes along the inner
walls of the alveoli. In other clams the ovocytes were half-grown
(Fig. 9). There was no apparent ecological basis for these differences
since all clams were collected from the same locality.

During February and March little further development was
observed . The water temperature for most of this period was below
5 C (Fig. 1) and the clams' metabolic rate was probably reduced.
Loosanoff (193 7) observed that Mercenaria enters hibernation when the
water temperature falls below 5 C and gametogenesis is considerably
reduced during this time .

During April 1962, instead of an expected rapid gonadal develop-
ment corresponding with an increase in water temperature, the ovocytes
underwent cytolysis . The clams began to show characteristics of the
summer or inactive stage (Fig. 2) rather than developing to maturity.
In a few clams some mature ova were found, but in no case did any of
the clams examined reach the degree of maturity found in the fall of
1961. By May 1962 all clams were in the summer or inactive stage
with only remnants of ovocytes (Fig. 2). From histological examination,
there was no evidence that spawning had occurred in the Tred Avon
River in the spring of 1962.

Examination of Plankton and Contents of Bottle Collectors

Mya larvae were first observed in the plankton on 28 September
1961. This period corresponded very closely to the time when indica-
tions of spawning were observed in histological sections. The peak
abundance of larvae was on 20 October, and the last larvae were found
in the plankton on 22 November. No Mya larvae were found in the
plankton in the spring of 1962.

The first set of Mya in the bottle collectors was observed on
20 October. These bottles had been in the water since 28 September
1961 . The greatest number of Mya spat (3 5) was found in a bottle that
was in the water from 5 October to 7 November. Only two Mya had set
in the bottles that were in the river from 7 November to 7 December. No
Mya set were found in the bottles that were placed in the river in the
spring of 1962 .

-129-

Fig. 8. Early gametogenesis of second cycle. Only a few
inclusions are present as compared to same stage in fall of 1961.
November 1961. X 200.

Fig. 9. Female clam with partially developed ovocytes. Majority
of clams reached this state in the spring and then the ovocytes underwent
cytolysis. March 1962. X200.

-130-

DISCUSSION

Examination of the seasonal cycle of M. arenaria in Maryland
reveals that there are two cycles of gonadal development each year.
Though the spring cycle did not reach maximum development in 1962,
there is evidence (Pfitzenmeyer, 1962) that in past years successful
spawning has occurred in both spring and fall.

From histological examination there is definite evidence that
the two spawning cycles actually involve two different gonadal matura-
tions . The summer or inactive stage which separates the two cycles
appears to correspond with the period in the life cycle of the oyster
which occurs just prior to hibernation, when the organism has stored a
considerable amount of glycogen to carry it over the winter. Further
studies are necessary to determine if inclusions in clams are actually
bodies of stored glycogen.

Loosanoff (1942) found in Crassostrea virginica an indifferent
stage at which time sex reversal takes place. However, Loosanoff
(193 7) did not find such a stage in the female _M. mercenaria, and he
stated that sex reversal seldom occurs in the quahog . Mya was found
to be similar to M_. mercenaria in this respect. At no time during the
female gonadal cycle was there a period of indifference when the sex
was in doubt. It therefore seems unlikely that sex change from female
to male occurs in the soft-shell clam.

Close examination of the two cycles that occur in Maryland
indicates that the inclusions, which are so abundant during the summer,
are supplying a good part of the nourishment (Coe and Turner, 1938) for
the developing ova . When the ova reach maturity only a few inclusions
are present in the alveoli. This is not the case in the spring cycle.
Following fall spawning, gametogenesis starts after only a short period
of inactivity. Mya appears to be unable to build up nutritive reserves
(inclusions) prior to the initiation of the second cycle. Therefore,
nourishment obtained from intake of food must be converted directly into
development of gametes .

It is obvious that when temperature is low, feeding is limited
and metabolic activity is greatly reduced, thus gametogenesis occurs
at a very slow rate during the winter. Why the spring development
was not completed during 1962 is not known, but this is probably not
unusual since evidence (Pfitzenmeyer, 19 57) shows that spring spawning
in Maryland has not been successful in some past years .

-131-

LITERATURE CITED

Coe, Wesley Ros well, and Harry J . Turner, Jr. 1938. Development of
the gonads and gametes in the soft-shell clam (Mya arenaria).
J. Morph., 62:91-111.

Guyer, Michael F. 1953. Animal Micrology. Univ. of Chicago Press,
Chicago . 5th ed .

Landers, Warren S. 19 54. Seasonal abundance of clam larvae in Rhode
Island waters, 1950-52. U . S . Fish and Wildlife Serv ., Spec .
Sci. Rept. Fish. 117: 1-29.

Loosanoff, Victor L. 1937. Seasonal gonadal changes of adult clams,
Venus mercenaria (L.). Biol. Bull., 72:406-416.

Loosanoff, Victor L. 1942. Seasonal gonadal changes in the adult
oyster, Ostrea virginica , of Long Island Sound. Biol. Bull.
82: 195-206.

Loosanoff, Victor L. 1962. Gametogenesis and spawning of the Euro-
pean oyster, O. edulis , in waters of Maine. Biol. Bull. 122:
86-94.

Manning, J. H. 19 57. The Maryland soft-shell clam industry and its
effects on tidewater resources . Maryland Dept . Research and
Educ . , Chesapeake Biol . Lab . Resource Study Rept . 12:1-25.

Pfitzenmeyer, Hayes T. 1957. Good winter survival and spring set of
soft-shell clams. Maryland Tidewater News 13:3.

Pfitzenmeyer, Hayes T. 1962. Periods of spawning and setting of the

soft-shell clam, Mya arenaria, at Solomons, Maryland. Chesa-
peake Sci. 3: 114-120.

Rogers, W. E. 19 59 . Gonad development and spawning of the soft
clam. Maryland Tidewater News 15:9-10.

Thorson, Gunnar . 1946. Reproduction and larval development of

Danish marine bottom invertebrates. Meddelelser Fra Kommis-
sionen for Danmarks Fisheri-Og Havundorsegelser . Ser. Plank-
ton 4: 1-523 .

Wilson, Edmund B. 1925. The cell in development and heredity. The
MacMillan Co., New York . 1232pp.

■132-

INCIDENCE OF MALACOBDELLA IN MERCENARIA CAMPECHIENSIS
OFF BEAUFORT INLET, NORTH CAROLINA

Hugh J . Porter

University of North Carolina Institute of Fisheries Research
Morehead City, North Carolina

ABSTRACT

The nemertean worm, Malacobdella grossa Mueller was found
in the mantle cavity of 83.3% of 2100 hard clams, Mercenaria cam-
pechiensis (Gmelin), from North Carolina offshore beds. The infestation
was higher in areas farther offshore, and lower in those closer to shore
and more influenced by estuarine conditions. M. grossa is a commensal
inquiline that is not known to harm the host. Of the infested clams,
54% had nemerteans in the right and 46% in the left mantle cavity; 93%
had a single nemertean, but during the period of rising water tempera-
tures (March-October), when a new year-class was entering clams, the
number of small nemerteans in one host sometimes rose as high as 21.
Sex of clams made no difference in incidence or degree of infestation.

INTRODUCTION

The nemertean, Malacobdella grossa Mueller, is widely dis-
tributed on the European and North American coasts . Occurring as far
south as the Mediterranean Sea in Europe (Coe, 1951), it has been
found in the mantle cavity of the following bivalves: Mya truncata L.;
Mya arenaria L . ; Mercenaria mercenaria (L . ) ; Pholas crispata L . ;
Cyprina islandica L . ; Cardium aculeatum L . ; Isocardia cor L . , ; and
Mactra stultorum L. (see Guberlet, 1925). In North America , Malacob-
della has been found in New England in Mya arenaria and Mercenaria
mercenaria according to Coe (1951); in North Carolina in Mercenaria
mercenaria by Pearse (1949) and in M. campechiensis (Gmelin) by
Porter and Chestnut (1962); in Florida and the Gulf area in Mercenaria
mercenaria (see Coe, 1951); and along the Pacific coast from Washing-
ton to California in Siliqua pa tula Dixon and Macoma secta Conrad
according to Coe (1940, 1951).

Incidence rates for Malacobdella in Mercenaria species are not
as well documented as in other bivalve genera . 1 Only Pearse (1949)

Since this paper was presented the following thesis has been received:
Twohy, D. W., 1951. The growth, sexual development and incidence
of infection of Malacobdella grossa in the razor clam, Siliqua pa tula .
Master's Thesis, Oregon State College.

-133-

gives any figures on the rate of this association. Kennel (1877-78),
Gering (1911), and Riches (1893) recorded incidence rates for Malacob-
della in Cyprina islandlca . Guberlet (1925) and McMillin (1925)
recorded rates of incidence in Siliqua patula and Takakura (1897) in
Mactra sachalinensis ♦

Kennel (1877-78), Gering (1911), Guberlet (1925), and Riepen
(1933) reported that clams usually have only one large Malacobdella
per clam and assumed that larger Malacobdella kill the smaller ones .

At present the relationship of Malacobdella to its host is
believed to be that of an inquiline commensal (Nicol, 196 0).

Periodic samples of Mercenaria campechiensis taken in life-
history and ecological studies off the North Carolina coast showed an
Incidence of Malacobdella grossa . Number and size of nemerteans was
recorded. This paper presents findings to date.

METHODS

During the 1960-61 clam season, one-bushel samples were
obtained monthly from commercial fishing boats in the Cape Lookout
area (Fig. 1). Beginning in June 1961, samples were dredged monthly
with the laboratory research vessel in area A, east of Beaufort Inlet
at a 40 to 50-foot depth (Fig. 1). Other samples were collected in Drum
Inlet and off Atlantic Beach (Fig. 1).

Each clam was measured and sexed, and its gonad condition
was recorded. The number of Malacobdella in right and left mantle
cavity of each was noted . In order to relax the nemerteans and prevent
violent contraction during killing and fixing, they were narcotized with
a saltwater solution of Chloretone . Afterwards they were fixed in
Bouin's solution.

The Malacobdella samples from Cape Lookout (Fig. 4) and area A
(Fig . 5) were preserved in cedar wood oil after being accidentally hard-
ened in absolute alcohol. These Malacobdella are mostly small and
hard. However, the area A (Fig. 6) samples preserved in 70% alcohol
appear to be in a more natural condition.

In spite of the narcotizing effect of Chloretone, contraction did
take place during fixing. A small series measured live and after preser-
vation in 7 0% alcohol showed a variation of between 10% and 42% con-
traction as a result of fixation and preservation.

-134-

d)

CO

4->

CD

fO

4->

£

ro

-rJ

U

X

-r-l

o

T3

t-<

C

a

-■— 1

o.

_

ro

" «

CO

C£

4->

•

C

Cjl,

CD

•

CO

i — i

CD

=

i_

a

•

CD

^-^

i-.

c

Cn

^

C

CD

S

o

&

4-1

rO

ffi

-i-i

en

#

C

+-»

CD

to

-.—1

ro

-C

2.

O

o

<.

O

CD

a

Ui

ro
C

-T-4

6

O

ro

o

1 — 1

o

O

u

it)

•

ro

o

xi

4->

o

ra

c

CD

u

T3
CD

c

ra

"a

<*

£

0>

w

-J

2

c

K

vt-4
O

ra

M-l I-.

^

4->

o *-•

1-H

ro

CO £
T3 m

O) >

c J2

5

£ i

O

x:

O CD

5°

CO

Q.

M-. <

rO

O ro

S

co £

§ ro

•

y! CD

i—l

ro J

•

°. -c

o>

° &

-w

IJL,

ffshore
Dcation

o

1 — 1

-135-

Length and width measurements were made of preserved Malacob-
della . Because of the 10% contraction, length plus width was determined
to be a more accurate index of size than length alone. Contraction rate
varied between samples, due to handling methods, so the data could not
safely be used for following morphometric size fluctuations of different
year-classes within and between sampling locations . Thus the data
were used mainly to determine the existence of year-classes, or new
recruitment .

RESULTS

Frequency of Occurrence

Malacobdella were found in 83.3% of 2100 clams from the off-
shore beds . In monthly samples from these beds, frequency of occur-
rence varied from 61 to 100% (Fig. 2). Samples taken in 1961 from Cape
Lookout showed less incidence (71.9% of 636 clams) than those from
area A (9 0.4% of 1448 clams) taken in 1961 and 1962. The Atlantic
Beach sample in 1961 showed an incidence of 63% in 127 clams (Fig. 2B),
One sample from Drum Inlet, not included in Fig. 2, had 26% of the 61
clams infested .

No difference in frequency of occurrence was apparent between
male and female clams . Eighty-three per cent of 1534 female clams
and 82.5% of 3 88 males contained Malacobdella .

More Malacobdella were found in the right mantle cavity than
in the left. Six hundred and eighty-five clams had nemerteans in the
left mantle cavity compared to 816 in the right mantle cavity. A statis-
tically significant difference between frequency of occurrence in these
two samples was found (x^ = 5.63; p = 0.02).

Numbers per Clam

The largest number of Malacobdella seen in any clam was 21 .
Ninety-three per cent of 1749 clams had only one Malacobdella . The
number of clams with two or more nemerteans varied seasonally during
the year (Fig. 3). Normally where two or more Malacobdella were
present in a clam, each was small or medium-sized. Large Malacob-
della were found in only one out of 115 cases of multiple inhabitation.

An attempt was made to duplicate some of the experiments of
Kennel (1877-78), Gering (1911), and Riepen (1933), who theorized
that in some manner larger Malacobdella attack and kill smaller or

-136-

Fig. 2. Frequency of occurrence of Malacobdella grossa in
samples of Mercenaria campechiensis . "B" represents a 1961 sample
taken off Atlantic Beach. "O" represents 1960-1961 samples from the
Cape Lookout area. "X" represents 1961-1962 samples from area A.
"Y" represents a 1960 sample from area A. The center of each letter
represents the datum point.

14
12
10

- e

5

in 6

2

<

d 4

b.

o 2 ..

J
il

m

M

i

JUNE JULY AUG. OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY JUNE JULY
6 12 30 10 9 4 16 21 15 9 I 4 9

1961

1962

Fig. 3. Seasonal variation of numbers of Mercenaria campechi-
ensis from area A with more than one Malacobdella grossa . Total rec-
tangle equals number of clams inhabited by more than one. Shaded
portion of rectangle represents number of clams inhabited with more than
two. Number at top of rectangle indicates number of clams in sample.

-137-

weaker Malacobdella . Numbers of small and large Malacobdella were
placed together In small finger bowls . The results were not clearcut.
Injurious attacks by one Malacobdella on another were not seen and
some large Malacobdella lived together in the same bowl for over a
month. However, those Malacobdella living together did not appear
as active as those living singly.

Recruitment of Small Forms

One size group of Malacobdella was present at Cape Lookout
between November 1960 and February 1961 (Fig. 4). Water tempera-
tures during this time varied from 9 to 18 C. In March 1961, a large
group of small forms suddenly appeared. Water temperature at this
time was about 11 C.

Figs . 5 and 6 show size-frequency graphs for nemerteans from
area A collected in 1961 and 1962 respectively. The June 1961 sample
indicates two major size groups. The range in size of the smaller size
group suggests that a recruitment of young occurred earlier and was
still continuing. The July 1961 sample shows that a second major
recruitment occurred during the previous 36 days. Water temperatures
at time of sampling in June and July were 22 and 24 C. The August and
October samples (1961) show recruitment of small nemerteans still
occurring. Water temperatures dropped from 26 to 22 C between August
and October. Samples from November 1961 through February 1962
show one major size group, but the size range indicates that the young-
of-the-year have grown large enough to become absorbed into the
larger size group. Water temperature dropped from 19 in November to
9 C in January. Recruitment began in February 1962 (bottom temperature
at time of sampling was 9 C); this recruitment continued through July
1962.

In November 1961 several bushels of clams from area A were
transplanted to Bogue Sound at the Institute of Fisheries Research
(Fig. 1). In June 1962 about half of these transplants were sampled.
Ninety per cent contained Malacobdella , but no recruitment of young
Malacobdella was noted (Fig. 7).

The one Drum Inlet sample (Fig . 7) of Malacobdella taken in
July 1961 showed primarily individuals of small size.

-138-

II /28/60

iU ■

i

10 -

: !

□

|

H

ft B

SIZE M MM

1 n n

3 0-1

•

"

20 -

1

; .

: :

10 -

'

:

H i i

I I I I I I ' I ' I ' I

8 3 9 5 105 l«5 !•» 22-5

SIZE IN MM

Fig. 4. Size-frequency distribution for Malacobdella grossa
from Cape Lookout area November 196 0 through March 1961. Size is
represented as maximum length plus maximum width.

- h h B a a

mm

o_

i • i i i i i » i . i i i ■ i ■ i ' i > i

2 ft ftft log I4« .0 lift t.A 501 S49 3«B ■ • «J 10$ 14.1 lift II* ••* >° ' 54B "*

Fig. 5. Size-frequency distribution for Malacobdella grossa
from area A, June 1961 through January 1962. Size is represented as
maximum length plus maximum width .

-139-

JO ■

2/21/62

1 ft

10 •

i pi R

j

■
■

S n

•

H

IfJH

1 (2 fl 1

20 -

n

■

< i

■

10 -

<

i ,

J a_

20 i

3/15/62

•

10 -

'

; a . n

< rJ

11

: : : ■ „ n

s

?0 -

H

10 -

. ,

h : "

,

i

'

• '

A P H '

r

,

Jflfl : ■ '

8 . a

n U

11 M LI

«

M

» 20"

«

w

>

3

z 10 -

_

_fl

n 3

-

a „

20 -

7/9/ 62

T

10 -

-

nHnflflflHuflSrl! :

_!. ! ! n n

I I 4 I I I I I I I I I I I I I I I I I

It 65 I0« i4« ItS tl» 1*6 JOS J«B 3«»
SIZE IN MM

I | i | i | i | i | l | l | i I i I i I

IS «fi 105 145 <•» 225 26 5 30 5 345 385

SIZE IN MM

Fig. 6. Size-frequency distribution for Malacobdella grossa
from area A February through July 1962. Size is represented as maxi-
mum length plus maximum width.

a

1 I

±11

P

NOV. 9, 1961

_a_

i,

m , .

■

JUNE 4, 1962

rH

I

HI

11

I

_0_

BOGUE SOUND

JUNE 26, 1962

n W rn\

\ 1 1

K n 1

DRUM INLET
JULY 17, 1961

1 n

05 25 45 65 8 5 105 12 5 14 5 16 5 IBS 205 22 5 24.5 26.5 26 5 305 32 5 34 5
SIZE IN MM

Fig. 7. Size-frequency distribution for different samples of
Malacobdella grossa. Transplanting from area A to Bogue Sound occurred
9 November 1961. Transplants from area A to Bogue Sound were sampled
4 June 1962. Size is represented as maximum length plus maximum width.

-140-

DISCUSSION

Frequency of Occurrence

Incidence rates for Malacobdella grossa in Mercenaria pre-
viously reported from the Atlantic coast of the United States are not as
high as the 62% to 100% found here. Data from Pearse (1949) indicate
a 36% incidence for Mercenaria mercenaria from the Beaufort area and
J. W. Ropes (personal communication) found a 20% incidence in Nan-
tucket Sound, Massachusetts.

Incidence rates for Malacobdella from other areas, and in other
bivalve species, were almost as high as those found in North Carolina.
Kennel (1877-78) noted 65-70% occurrence in Cyprina islandica from
Kiel Harbor, Germany. Gering (1911) reported 55% from the Western
Baltic. Riches (1893) recorded only one Cyprina islandica from the
English coast that had no Malacobdella; unfortunately, his sample
size is not known. McMillin (1925) reported an incidence of 88% in
Sillqua patula (off the coast of the state of Washington); however,
Guberlet (1925) intimated that this figure was too high. Takakura (1897)
recorded a 96% incidence of Malacobdella japonica in Mactra sachalinen-
sis off the coast of Japan.

The highly variable frequencies of occurrence for Malacobdella
in Mercenaria from North Carolina (Fig . 2) may be attributed to dif-
ferences between seasons, sample years, and sampling locations.

Both area A and Cape Lookout series show a low rate of inci-
dence during fall and a high rate during spring. Why the rate became
lower as the recruitment season progressed and how or why it increased
progressively after juvenile recruitment had ceased is not understood.

Insufficient data are available to delimit or explain yearly dif-
ferences in incidence rate. Cape Lookout samples and the Atlantic
Beach sample, taken during the 1961 season, showed lower incidence
than those taken during the 1961-1962 season from area A. One 1961
sample from area A showed a much higher incidence than that from
Cape Lookout and fits in closely with the 1962 area A data . Thus, it
is possible that the difference between the two areas is not simply due
to different year-classes . Some difference in rate of occurrence between
years is shown in the June and July area A samples . The data indicate
that the incidence in 1962 was 7 or 8% higher than in 1961.

Incidence varied between sampled localities. The infestation
rates in area A samples were higher than those recorded from Cape

-141-

Lookout, Atlantic Beach, and Drum Inlet. Cape Lookout and Atlantic
Beach are influenced by estuarine conditions to a greater extent than
area A. Ebb tides at Beaufort Inlet and Barden Inlet flow primarily-
westward in a longshore current thus affecting the area off Atlantic
Beach and just west of Cape Lookout more than area A (east of Beaufort
Inlet). The Drum Inlet sample, taken just inside the mouth of the In-
let, had the lowest incidence rate and was from the area most estuarine
in nature. This sample showed an incidence rate of 26%, similar to the
36% recorded by Pearse (1949) from Beaufort Inlet. However, Pearse
recorded this incidence as being found in Mercenaria mercenaria . The
Drum Inlet sample probably contained a mixture of Mercenaria mercen-
aria , M . campechiensis , and possibly their hybrids. The two species
are often difficult to distinguish in the inlet areas, so it is possible
that Pearse was not sampling exclusively Mercenaria mercenaria .

As pointed out, the distribution of Malacobdella-inhabited clams
in the North Carolina area may be inversely controlled by degree of
estuarine conditions present. It is interesting to note that in the clams
transplanted to Bogue Sound, where Malacobdella has not been found,
incidence remained at the same level over an eight-month period.
During the sample period a new population of small individuals appeared
in the offshore population but none appeared in the transplanted clams
in the Sound . This may indicate that adult Malacobdella are able to
survive in estuarine conditions but not able to reproduce.

The size and incidence of Malacobdella found in Drum Inlet
(Fig. 7) possibly indicate that following an invasion of small forms
into an area, few survive to become full-size; or, that two distinct
year-classes are present, as represented by the one 22.5 mm individual,
and by the group between 4.5 and 10.5 mm. The large one in this case
could be the remnant of a group that had invaded the clams one or more
years before the smaller size group.

Numbers per Clam

The data show that in North Carolina 93% of inhabited clams had
only one Malacobdella (samples varied from 84 to 98%). Kennel (1877-78)
reported that in over 500 Cyprina , 65-70% infested, he found only four
with more than one Malacobdella . Guberlet (1925) found only one in 4 0
clams (97%) from the state of Washington with more than one Malacob-
della . When more than one was found per clam, the largest number
seen in M . campechiensis was 21. The largest number Kennel (1877-78)
and Guberlet (1925) found was two; the largest number Gering (1911)
found was four .

-142-

In only one case did more than one large Malacobdella occur
in a clam. In this case two large specimens were found in the left
mantle cavity and one large and three small ones were in the right
mantle cavity. The large ones had a size of about 19 mm. Neither
Kennel (1877-78) nor Gering (1911) found more than one fully developed
Malacobdella per clam. However, Riepen (1933) stated that when more
than one occurred in a clam, each would be separated from the other by
the clam body. Kennel (1877-78), Gering (1911), and Riepen (1933)
believed that the reason for one large Malacobdella per clam was that
where two or more were present, the larger and stronger would kill the
lesser one. Riepen (1933) also believed that the larger forms would
eat the larval forms. Results from duplication of their experiments so
far have not been as conclusive as their results were.

In area A samples the number of clams with more than one
nemertean varied seasonally (Fig. 3). This seasonal variation, com-
pared with Fig. 5 and 6, appears to be related directly with the influx
of small Malacobdella into the population. Later, when the numbers
of small Malacobdella decreased, so did the percentage of clams
with two or more Malacobdella .

Recruitment of Small Forms

The Cape Lookout and the area A series both show that recruit-
ment by a new year-class of Malacobdella begins in February or March,
continues through summer and early fall, and does not occur after
October. The recruitment season runs from the coldest to the warmest
period of the year, during those months that have rising temperatures.
Those months in which water temperatures are dropping (October through
January) show no recruitment. Whether or not the pattern represents
the spawning season for Malacobdella in this area is not known.

Effect of Malacobdella on Its Host

Malacobdella grossa appeared to have no noticeable effect on
M . campechiensis in spite of the high local intensity of Malacobdella-
inhabited clams . The food for Malacobdella is reported to be algae,
protozoans, small Crustacea, and detritus (Kennel, 1877-78); Riepen,
1933; and Guberlet, 1925). In no case was there any sign of damage
to the mantle or gills that might have been caused by Malacobdella .
Condition of clam gonad or clam meat did not appear to be influenced
by the presence or absence of the nemertean. This is in accord with
what Guberlet (1925) found in Sillqua pa tula .

-143-

Future Investigations

Additional data may provide information on the effect of environ-
mental factors upon incidence, maturation, and growth of this nemertean
The life cycle of Malacobdella has not been fully worked out in the
Western Atlantic .

Further studies are needed to determine possible effects of the
nemertean on clams and why normally only one large Malacobdella is
found in each clam.

REFERENCES CITED

Coe, W. R. 194 0. Revision of the nemertean fauna of the Pacific
coasts of North, Central and Northern South America. Allan
Hancock Pacif. Exped . 2:247-3 23.

Coe, W. R. 1951. Geographical distribution of the nemerteans of the
northern coast of the Gulf of Mexico as compared with those of
the southern coast of Florida, with description of three new
species. J. Wash. Acad. Sci. 41:328-331.

Gering , G. 1911. Beitrage zur Kenntnis von Malacobdella grossa
(Mull.) Wiss. Zool. 97:673-720.

Guberlet, J. E. 1925. Malacobdella grossa from the Pacific coast of
North America . Publ . Puget Sound Biol . Station. 5:1-13.

Kennel, J . von. 1877-78. Beitrage zur Kenntnis der Nemertinen.
Arb. Zool-Zootom. Inst. Wurzburg . 4:3 05-381.

*McMillin, H. C. 1925. The life history and growth of the razor clam.
State of Washington, Dept . Fish. Ann. Rept. Supt. Fish. No.
34. 52 p.

Nicol, J. A. C. 196 0. The biology of marine animals. Interscience
Publ., Inc., New York, 7 07 p.

Pearse, A. S. 1949. Observations on flatworms and nemerteans col-
lected at Beaufort, N. C. Proc . U.S. Nat. Mus . 100: 25-3 8.

Porter, H. J. and A. F. Chestnut. 1962. The offshore clam fishery of
North Carolina . Proc . Nat'l Shellfish . Assoc . 51:67-73.

* Not seen

-144-

Riches, T. H. 1893. A list of the nemertines of Plymouth Sound.
J . Mar . Biol . Assoc . United Kingdom 3 : 1-29 .

Riepen, O. 1933. Anatome und Hlstologie von Malacobdella grossa
(Mull.). Wiss. Zool. 143:323-496.

Takak.ura,U. 1897. On a new species of Malacobdella (M . japonica),
Annot. Zool . Japonensis. 1:105-112.

-145-

TESTS OF INTERNAL TAGS FOR GREEN CRABS (CARCINUS MAE N AS)

John W. Ropes

Bureau of Commercial Fisheries Biological Laboratory
Boothbay Harbor, Maine

ABSTRACT

A technique is described for inserting U-shaped monel metal
tags into the body cavity of green crabs, Carcinus maenas (L.), at a
site between the first and second abdominal segments. In the laboratory,
tagging mortality of approximately 20% was observed. Tags were re-
tained through ecdysis with comparable mortality during tests of disk
tags or U-tags inserted at coxal or lateral spine sites. The internal
tagging technique was developed for use in studies of migration and
growth of the green crab, a predator of soft-shell clams, Mya arenaria L.

INTRODUCTION

Internal tags which were retained during ecdysis by green crabs,
Carcinus maenas (L.), were tested in the laboratory. The internal tag
was developed to provide a means of observing the migration and growth
of individual crabs. Such information is needed to design effective
methods of controlling green crab abundance. Carcinus is a predator
of the soft-shell clam, Mya arenaria L., and is considered responsible
for a decline of this species north of Cape Cod (Glude, 19 5 5).

Various tags and marks have been used on arthropods, but none
were considered suitable for green crabs. Tags and marks applied to
the hard exoskeleton are lost during ecdysis. Some tags and marks that
are retained through ecdysis were reviewed by Simpson (1961). Barbed
and sutured tags have been retained, but could impede an animal's
normal activities or be dislodged from the insertion site when the crab
burrows. Holes punched in the uropods and telsons, a method used on
lobsters, were undesirable because the marks are eventually obscured
by new shell and tissue growth. Dawson (1957), Cargo (1958), and
Costello (19 59) reported successful marking with biological dyes.
However, dyes eventually faded or were difficult to see through the
pigmented exoskeleton of the green crab. Hayes (1961) described a
"spaghetti" loop tag used on king crabs (Paralithodes) , sewn around
the arthral muscle that lies anterior to the joint between the carapace
and abdomen. Such a conveniently located muscle could not be found
in Carcinus .

-147-

To circumvent some of these disadvantages an internal tag and
tagging technique were developed during 1956-1958 at the Bureau of
Commercial Fisheries Substation in the Narragansett Marine Laboratory,
University of Rhode Island, Kingston, Rhode Island. Extensive tests
of the tag and tagging technique were conducted at the Bureau's Bio-
logical Laboratory, Boothbay Harbor, Maine, during 1959 and 196 0.

PRELIMINARY TESTS OF TAGS AND TAGGING TECHNIQUES

The flexible, membranous joints between segments were chosen
for tag insertion sites to reduce exoskeleton damage. Not all joints
were suitable. Many were too small. Leg joints were unsuitable because
Carcinus can autotomize its legs (Waterman, 1960). Tags were easily
slipped into the body cavity through an abdominal joint and through the
joint between the carapace edge and first abdominal segment.

Tag dimensions were delimited by examining internal anatomy.
Two vital structures susceptible to damage, the heart and internal
thoracic skeleton, lay in the path of tag insertion. The heart lies about
in the center of the body mass and anterior to the tag insertion site.
Measurements from the posterior edge of the heart to the tag insertion
site determined the maximum length of a tag. Similarly, the maximum
width of a tag was determined by measuring the distance between two
internal processes of the thoracic skeleton that covers the bases of leg
muscles . Both structures could be avoided in crabs 2 0 mm or larger by
using tags not over 1/4 inch (6.35 mm) in diameter. All size measure-
ments of crabs in this paper refer to carapace widths .

Three types of experimental tags were fabricated. (1) Disk tags
3 . 165, 4.76, and 6.35 mm in diameter were cut from 0. 015-inch (0.381
mm) thick sheet metal using common paper punches. Fish strap tags
(Rounsefell and Kask, 1943), 1.59 mm wide were modified (2) by cutting
off and rounding the ends to form a flat tag 6.35 mm long, and (3) by
cutting off and rounding the ends, then folding the tag to form a U with
a finished length of 4.76 mm. These U-tags were inserted with a
special tool; other tags were inserted with forceps.

Two insertion sites for disk tags were tested. Disk tags inserted
at the joint between the carapace and abdomen were not retained through
ecdysis. Dissections of dead, unmolted crabs revealed that the tags
rested on a projection of the first abdominal segment that is overlapped
dorsally by the carapace. One molted shell was found with the tag
resting on this projection. Burrs on the edges of the disks anchored
the tag to the molted shell. Disk tags were retained when inserted at

-148-

the joint between the first and second abdominal segments. Dissec-
tions of tagged crabs revealed that the tags lay in soft body tissues
without touching the hard exoskeleton. The latter site was chosen
for subsequent tag insertions.

Several tests of tagging mortality indicated the most suitable
tag size. In a test of the three disk sizes, crabs 2 0 to 49 mm were
separated into three 10-mm size interval groups and held for 21 days
(Table 1). Tagging caused high mortality in the crabs 20 to 29 mm,
especially those tagged with the two largest disks. This might be
expected, but not the anomalous situation in crabs 3 0 to 49 mm where
mortality decreased with an increase in tag size. These data indicated
that 4.76 mm and 6.35 mm disks were unsuitable for crabs less than
3 0 mm. Tag retention was not accurately observed because cannibalism
occurred during molting .

Table 1. The percentage mortality of crabs tagged with disks

Crab
size

Per cent mortality

Crabs tagged with

Untagged
crabs

(mm)

3.165 mm disks 4 .76 mm disks

6.35 mm disks

20 to 29*

50 74

86

20

30 to 39*

24 18

16

14

40 to 49*

42 28

22

12

*50 crabs for each disk size were used.

To prevent cannibalism, in an experiment with the flat modification
of strap tags, 120 crabs ranging in size from 20 to 39 mm were held in per-
forated plastic berry baskets 4.5 inches (11.33 cm) square and 3 inches
(7.62 cm) deep. After 120 days, only 4% of these crabs were dead. As
many as 86% molted, but only 9% of the molted crabs retained the tag.
Apparently burrs on these tags prevented complete insertion into the body
cavity, for molted shells were found with the tag stuck in the insertion
site. This tag was not suitable , then, even though few mortalities
occurred .

-149-

The U-shaped modification of strap tags and the following
technique insured more complete tag insertion. A tool was used first
to pierce the abdominal site and then to hold the tag for insertion
(Fig. 1). The shaft diameter of the tool was slightly larger than the
tag end and served as a shoulder preventing insertion beyond 4.76 mm,
a depth twice the length of a U-tag. U-tags were found completely
surrounded by soft body tissues when tagged crabs were dissected.
Mortality from U-tags was not appreciably different from that with other
tags. When 222 crabs of various sizes were tagged and held for 3 5
days, most of the crabs that molted retained the U-tags (Table 2).
U-tags seemed most suitable, then, because they were easily inserted,
caused relatively few mortalities, and were retained by many crabs
through ecdysis .

Table 2. The percentage mortality and U-tag retention through a molt

Number

Percentage

Crab

Molted and

size

of

retained the

(mm)

crabs

Died

Molted

tags

2 0 to 29

17

12

29

100

3 0 to 39

146

16

24

89

40 to 49

28

4

11

100

5 0 to 69

31

23

0

0

TESTS WITH U-TAGS

The following materials and methods were used during subse-
quent experiments. Coded U-tags were 1.98 mm wide and of a finished
4.76 mm length, and made of monel metal, 0.010-inch (0.254 mm)
thick, by the National Band and Tag Company, Newport, Kentucky. The
crabs, held separately to prevent cannibalism, were placed in a tank
16 feet long and 18 inches wide with a water depth of 1 1 inches. Sources
of decomposition such as excreta and dead crabs were removed periodi-
cally. Live soft-shell clams were fed to the crabs each week to promote

-150-

Fig. 1. The tagging technique. An insertion tool (A), as well
as its top (B) and side (C) appearance in relation to U-tags (D).

-151-

growth and molting . Each crab was randomly placed within the tank
with regard to size, sex, and method of tagging .

Tagging mortality and tag retention

During 19 59 two tests were conducted to observe the mortality
and tag retention of crabs bearing U-tags. Crabs, caught just prior
to each test, were separated into 20 size classes with 1 mm intervals.
A male and a female were drawn from each of the 2 0 size classes, to
total 40 untagged controls for each test. A similar selection of 152
crabs was tagged. The first test included crabs 20 to 39 mm,
whereas the second included crabs 25 to 44 mm because smaller crabs
were not available. The intervals 20 to 29 mm, 3 0 to 39 mm, and 40 to
44 mm were used to group the data by crab size, and data from both
tests were combined for analysis .

Crabs tagged for the first test 1 May 1959 were held for 214
days and those tagged for the second test 10 July 1959 were held 131
days. A 13% escapement occurred from the plastic berry baskets used
in the first test and a 7% escapement from compartmented trays used
in the second test. Only 27% of the tagged crabs 2 0 to 29 mm were
alive on the fourteenth day (Fig. 2), indicating that the tags caused
a high mortality. Larger crabs were less affected by the tags during
the same 2-week period: 78% of the 3 0 to 39 mm and 89% of the 4 0 to
44 mm crabs were alive on the fourteenth day. No mortalities of con-
trol crabs 40 to 44 mm occurred until the ninth week at which time 65%
of the tagged crabs were alive. Thus, the larger crabs were less
affected by tagging.

Most tagged crabs died before molting in both tests (Table 3).
Apparently many deaths of crabs less than 40 mm were due to factors
other than tagging, for there was a high mortality in controls . Most
molted crabs retained the U-tag. Though survival was low in the first
test, 11 crabs molted and 9 retained the tag. More crabs molted (66)
during the second test and most (62) retained the tag. A high per-
centage (9 2%) of molted crabs retained the tags, but only a few were
alive at the end of the tests .

Insertion sites and tags

Tagging mortalities during 1959 suggested that other insertion
sites should be investigated . To this end a total of 43 2 crabs, 3 0 to
39 mm, were divided into six groups of 72 crabs each. Each group
contained four crabs of each sex and from 1-mm size categories,

-152-

TAGGED

UNTAGGED

100

75

50

25

in 0

O ioo

<

i- 75

Z 50
ua

V 25

fit

u. 0

Q. 100

L — V

20 to 29 mm

i

1
1
1
1

' "

• '

*
■ i i —

30 to 39 mm

\ »

\ 1
\ '--

' s„^

I I I I I

i i i

8 12

WEEKS

16

Fig. 2. The combined percentage survival of green crabs
tagged with U-tags during two experiments in 1959.

-153-

Table 3 . The mortality of tagged and untagged crabs, and the reten-
tion of tags after molting .

Number of crabs

Test, tagged and
untagged crabs3

Pre-mc

lted that

Post-molted that

Post-molted
that retained
the tag and

Died

Lived

Died Lived

Died Lived

i— i

Tagged 136
Untagged 3 1

125
20

0
2

10 1
7 2

8 1

o

CM

o
2

Tagged 143
Untagged 3 6

65

7

12

5

44 22
10 14

42 20b

Less escapement from the holding cages.

One crab molted twice and retained the tag both times

excepting the sizes 32, 33, 36, and 37 mm that had only three of each
sex. One untagged group served as controls. Three groups carried
U-tags inserted either at a lateral spine, a coxal joint, or the abdominal
site. Two groups carried metal disks 4.76 mm in diameter at either the
lateral spine or abdominal site. The lateral spine sites were made by
cutting off a portion of the carapace and spine large enough to admit a
tag. Tag insertions were made to a depth of 4.76 mm or less and in a
mesiad direction. The coxal sites were made by forcing crabs to auto-
tomize a cheliped . Each U-tag was compressed to make it as small
as possible before insertion into the coxal site. Disks were considered
too large to insert in the coxal site without causing extensive damage
to the hard exoskeleton.

The test began 16 June 1960 and ended 127 days later. New
holding trays, each with compartments that were 4.5 inches (11.33 cm)
long, 3 inches (7.62 cm) wide and deep, were used and only 2% of the
crabs escaped during the test. The survival of tagged crabs ranged
from 67% to 81% in the five groups on the fourteenth day of the test
(Fig. 3). Crab deaths were attributed to tagging since no controls died

154-

CD
CD

x:

39V1N]}U 3d

(0

(0

0)

a

■4->

o

w

CD

■M

CO

Cn

c

I*

3
T3

t/i

2

O

c
0

CD
i-

>

CD

t»

(0

c
u

S-.
<D

a

CD

.c

H

01

c
o

-4->

)-l

0
to

c

155-

during this period. After this initial period, a mortality of both tagged
and control crabs was observed. At termination, survival in the five
tagged groups ranged from 7 to 21% and 3 7% of the controls were alive.
In a chi-square test comparing the mortality data (Table 4) for each
tagging technique, no one technique was found to be significantly
different .

Disks inserted at the abdominal and lateral spine, as well as
U-tags at the lateral spine or coxal site, were retained by 17 to 29 of
the tagged crabs (Table 4). U-tags inserted at the abdominal site were
retained by 3 2 of the crabs after one molt. When the data on tag
retention were tested with chi-square, the differences were found to be
not significant .

Measurements were made to determine if tagging affected the
increase in carapace width after one molt. Crabs carrying U-tags at
the abdominal or coxal site, and crabs with metal disks in the abdom-
inal site, had an average increase in carapace width after molting of
6.8, 7.3, and 7 . 7 mm respectively . Control crabs had an average
increase in carapace width of 7.6 mm. Crabs with either U-tags or
metal disks at the lateral spine site could not be measured after molting
because a portion of the carapace had been cut off to insert the tags.
An analysis of variance indicated that there were no significant dif-
ferences in the average growth of tagged and untagged crabs .

DISCUSSION

Various advantages and disadvantages of the tags and insertion
sites were observed during the tests . It was easier and quicker to
insert U-tags than disk tags, even though U-tags were smaller. The
special tool facilitated insertion of U-tags, whereas the forceps used
with disk tags were unwieldy. Disk tags caused less damage to the
hard exoskeleton of the abdomen because they were inserted between
the segments. The blunt, rounded ends of the U-tags sometimes made
a small hole in the hard exoskeleton of the abdomen. Infections at the
abdominal insertion site were not observed and the wounds were almost
invisible after the crabs molted. Less damage was done to the exo-
skeleton by inserting tags at the abdominal site than by cutting off the
lateral spine or causing a crab to autotomize a claw. A U-tag inserted
at the coxal site did not rest on or near vital organs, as did tags
inserted at the other two sites, but it probably became sealed within
the crab during autotomy (Waterman, 196 0). Some crabs even regenerated
a new claw during these tests . However, the absence of a claw may be
a disadvantage to tagged crabs released into a natural population. In
consideration of the above, the insertion of a U-tag at the abdominal site
was considered the most suitable technique tested.

-156-

to

cd

c
o

CD
03

c

-.-4

<p

CD

4-»

T)
C

(0

03
Cn
(0

03

CD

a
>^

O
5

01
CD

Cn

c

-.-I

(-1

T)

C

o

-I— I

4-J

c

CD

ro

4->

-a
c
ra

ro

o

en
C

•i-l

cn

CO
H

CD
E-i

4-J

CO Di

T3

CD

u

,C (0

>

CM -ST1

CO CM

LO

4-> 4-J

T) CD

i-J

CD jC

4-J 4->

i — 1

O T3

S 2

1 c

0) 2

T3
CD

CD CO

CD O

^r

O %

-f-4

1 — 1 1 — 1

|-H 0O

CM

cu Z

P

4->

•a

CO

CD
>

-»— 1

oo r>-

CD CM

CM

"*

4-J

1 — 1

CM

"O

hJ

CD

4-J

03

o

X!

£

(0
s-
rj

i

4-J

03

T3

CD

00 CD

r-i CD

r>~

i — l

o

t-i

CM 00

00 OO

oo

OO

4-4

O

P-,

Q

U

CD

XI

6

^

2

4-J

-a

ro

CD

^r cd

CO 0O

CM

CM

4-J

>

T3

Z3

CD

4-J

. — 1

o

B
i

T3

CD

CD

CD 00

(X) r^

CD

OO

S-c

a-

O

CM i—l

C-J CM

i — i

i — l

fD

0)

ro -Q

CD CM

CM i-l

o

o

4-j ro

to t>~

r»- t^

p^

o^

o ^

H O

en

Cn

Cn

-a

m ro

M ro

ro

CD

T3

oi v

oi v

4-J

en

C

-i-4 1

-- j 1

1

Cn

ro

Q &

Q 13

|3

ro

cd

4-J

4-J

c

|3

suids

uauiopqv

BXOO

01

CD

^eaaiei

Cn Q.

ro >,

F-i £■

o

01

E

CD

3

T3

03

C

01

o

-1-4

4-J

o

CD

>,

03

X)

o

CD
X

XI

4-J

C

4-J

i-1

o

a

c

CD

4-1

CD

3

T>

X)

T3

4-J

c

03

S-4
-t-4

3
o

M-4

>4-4

CD

03

■s

ro
5

c

o

Cn

ro

en

4-J

ro

4-J

CD
X

CD

H

X!

4-J

.

en

c

ro

c

o

•

ro

CD

03
CD
Cn

4-J

CD

1-4

XI

4-1

ro

«

-a

o

CD

CD

4-J

D>

U

-<-4

ro

c

£

CD

-a

-M

C

TS

0

o

0)

Cn

x:

-t->

CD

.— i

S-.

CD

o

X

4-J

S

. — 1
. — 1

£

4-J

ro

ro

o

x

4-J

1-4

1-4

4-J

<4-l

ro

4-J

X!

a

c

CD

6

ro

TS

u

ro

x

CD

CD

Q.

c

03

ro

o

X!

o

01

ro
u

01

CD

u

CD
01

T3

3

CD
CD

01
CD

O

c

1-4

X

J

t— 1

H

-157-

Adverse holding conditions probably caused many deaths during
the tests . Most control mortality occurred after molting and was nearly
equal to tagged crab mortality after molting . Most tagging mortalities
occurred before molting, particularly during the first two weeks after
tagging when very few control deaths occurred . If the mortality of
tagged and control crabs is compared at the end of each test, the dif-
ferences equal 13, 30, and 19%. An average of these differences (20%)
was estimated as the magnitude of tagging mortality.

Though tagging mortality was high, U-tags could produce valu-
able information about Carcinus . Green crab populations are assumed
to be fairly discrete, occurring within a localized area such as an
estuary, and the recovery of relatively few tagged crabs would serve
to support this belief and provide indications of migration patterns.
External tags can provide information on migrations (Baptist, Smith and
Ropes, 1957; Edwards, 1958), and estimates of longevity (Ropes, 1961),
but internal tags have the added feature of providing observations on
growth because they are retained through ecdysis.

There is a need for a rapid method of recovering internally tagged
crabs, but at present it is only possible to suggest some recovery
methods. Dissecting and searching through each crab is one method,
but using a device that would detect a tag in the intact crab is more
desirable. Moore and Mortimer (1954) have described a metal-sensitive
device for fish tags that might be adapted for use with the U-tags, or,
if the tags were made radioactive, a radiosensitive device might be used.

ACKNOWLEDGMENT

I gratefully acknowledge the assistance of Mr. Louis D. Stringer,
who designed the U-tag inserting tool used in these tests.

LITERATURE CITED

Baptist, J . P., O. R. Smith, and I . W. Ropes . 1957. Migrations of

the horseshoe crab, Limulus polyphemus , in Plum Island Sound,
Massachusetts . U.S. Fish Wildl . Serv. , Spec . Sci. Rept . ,
Fish. No. 220. 15 p.

Cargo, D. G. 1958. Crabs retain dye from stained food. Maryland
Tidewater News, Maryland Dept . Res. Educ . 14(2): 6, 8.

Costello, T. H. 1959. Marking shrimp with biological stains. Proc .
Gulf Caribbean Fish . Inst. 11th Ann. Session: 1-6.

158-

Dawson, C. E. 1957. Studies on the marking of commercial shrimp
with biological stains. U.S. Fish Wildl. Serv., Spec. Sci.
Rept., Fish. No. 231. 24 p.

Edwards, R. L. 1958. Movements of individual members in a popu-
lation of the shore crab, Carcinus maenas L . , in the littoral
zone. J . Animal Ecol . 27:37-45.

Glude, J. B. 1955. The effects of temperature and predators on the
abundance of the soft-shell clam, Mya arenaria , in New
England. Trans. Am. Fish. Soc . 84 (1954):13-26 .

Hayes, M. L. 1961. King crab tagging methods in Alaska . Intern.
Comm. Northwest Atlantic Fish., North Atlantic Fish Marking
Symposium, Woods Hole, Mass., Serial No. 852, (B. Rept. 1),
Contrib. No. 42. 5 p.

Moore, W. H., and C. H. Mortimer. 1954. A portable instrument
for the location of subcutaneous fish-tags. J. Conseil Perm.
Intern. Explor . Mer. 20:83-86.

Ropes, J . W . 1961. Longevity of the horseshoe crab, Limulus poly-
phemus (L.). Trans . Am. Fish. Soc . 9 0: 79-80.

Rounsefell, G. A. , and J . L. Kask . 1943. How to mark fish . Trans.
Am. Fish. Soc. 73:320-363.

Simpson, A. C. 1961. Marking crabs and lobsters for mortality and
growth studies. Intern. Comm. Northwest Atlantic Fish.,
North Atlantic Fish Marking Symposium, Woods Hole, Mass.,
Serial No. 839, (B, Rept. 1), Contrib. No. 29. 10 p.

Waterman, T. H. 1960. The physiology of Crustacea . Vol. I.
Metabolism and growth. Academic Press, New York and
London . 670 p .

-159-

STUDIES ON OYSTER SCAVENGERS AND THEIR RELATION TO THE
FUNGUS DERMOCYSTIDIUM MARIN JM1

Hinton Dickson Hoese

2
Virginia Institute of Marine Science

ABSTRACT

Dermocystidium marinum, a parasitic fungus of oysters, was
demonstrated from the stomach of the snail, Urosalpinx cinerea, from
the stomach, intestine, and body of three fishes, Gobiosoma bosci,
Chasmodes bosquianus, and Opsanus tau, and from the body, especially
setae, of two crabs, Neopanope texana and Rhithropanopeus harrisii.
All animals containing _D. marinum had scavenged oysters infected by
the fungus. A few oysters became lightly infected when kept in aquaria
with fishes that had been fed infected oyster tissue. In one tidal inlet
of Chesapeake Bay, Virginia, Eurypanopeus depressus was the most
abundant scavenger, followed by Nassarius vibex. Gobiosoma bosci, and
Panopeus herbstii. Killed oysters on this reef were consumed by scav-
engers in less tTian one day in temperatures over 24 C. At temperatures
above 18 C, dead oyster tissue never remained long enough to decay.
Theoretical methods of transmission of D. marinum by scavengers are
discussed. It is concluded that nearly alT dying oysters are consumed
by animals during periods of normal mortality, so their parasites must
pass through the digestive systems of scavengers.

INTRODUCTION

Within the past decade there have been several studies on the
biological structure of oyster reefs. These studies, however, have
given little insight into the dynamics of oyster communities . The ex-
tensive studies of Hedgpeth (1953), Gunter (1955), and Parker (1955,
1959) in Texas, Wells (1961) in North Carolina, and Korringa (1951)
in Holland were largely concerned with sedentary forms, and the highly
motile fishes went little noticed . The concept of the oyster biocoenosis
is known widely, but has received little expansion.

The present study was not concerned with the whole community,
but with the role of fishes, crabs, and a few other scavengers in the
community, especially in their relationship to the oyster, Crassostrea
virginlca (Gmelin) and its parasitic fungus, Dermocystidium marinum

1 Contribution from the Virginia Institute of Marine Science, No.

2 Present address: Department of Zoology , University of Texas , Austin ,

-161-

Mackin, Owen, and Collier, 1950. Knowledge of D . marinum has been
reviewed very recently by Johnson and Sparrow (1961) and by Mackin
(1962).

This study started from observations of fishes living in close
association with oysters, and progressed to observations on the
relationship of mortality of oysters with activity of other species in
the community. Mortality of oysters in the study area occurs predomi-
nantly in the warmer months, and most of this mortality is due to Dermo-
cystidium marinum . Some of the oyster associates that are active in
summer are scavengers of dying oysters and consequently ingest cells
of oyster parasites. This suggested that the scavengers might transmit
infections to other oysters .

METHODS

Data on scavengers were collected incidental to studies of
oyster mortality on the Eastern Shore of Virginia. Studies were largely
confined to a small embayment off Chesapeake Bay called The Gulf,
just north of Cape Charles, Virginia . Life history data on scavengers
gathered here and in other areas of Virginia will be presented elsewhere.

The presence of Dermocystidium marinum was determined by Ray's
(19 52) thioglycollate culture method. After culture the enlarged fungus
cells were stained (blue) with iodine. Oysters and the digestive tracts
of fishes were cultured by the standard method, but feces were originally
cultured in petri dishes with 1 0 cc of medium added to about 5 cc of
water containing fecal material . This method has the advantage of not
disturbing the feces, but enhances the growth of molds . Since this
proved generally unsatisfactory, feces were later placed in test tubes
with the medium, and dilute oyster serum from uninfected oysters was
added. Uninfected oysters came from the Seaside of Virginia where
D. marinum has not been found (Andrews and Hewatt, 1957; also
unpublished studies). Fishes and crabs were fed in aquaria or small
bowls with pieces of meat, or with whole oysters that died with heavy
Dermocystidium infections . The fish were then washed in three or
more separate dishes and placed in dishes with Seaside water of a
salinity near 3 0 parts per thousand; or they were placed in aquaria for
infection experiments. Later, after it seemed that the fungus was killed
by Seaside water, Chesapeake Bay water of salinity near 20 pot was
substituted. Feces were collected with a sterile pipette and placed in
culture . After two to five days these cultures were examined under
monobjective and stereoscopic microscopes.

162-

At approximately monthly intervals from 9 June through 7
November 1961, groups of 1 0 oysters were made into "gapers" by
cutting adductor muscles. Each month, these artificial gapers were
placed in individual trays (10 per tray) made of one-inch-mesh rat
wire, with a cover of the same material. These permitted small
scavengers to enter while preventing large crabs from removing the
oysters . Ten control oysters with adductors cut were placed in a cage
of 1 /8-inch hardware cloth, which eliminated most scavengers other
than very small recently metamorphosed gobies and mud crabs (which
ate very little meat). The experimental and control cages were placed
on the top and edge of an oyster reef at The Gulf. This reef is located
near the lower edge of the intertidal zone just inshore from extensive
eelgrass (Zostera marina) flats . The amount of meat taken by scav-
engers was calculated from wet weights of experimental and control
oysters, after 10 minutes drying in the shade. Direct observations
were made on the activities of scavengers on killed oysters in the
shallow, clear water on and near the reef.

Several crude infection experiments were conducted by feeding
fish pieces of infected oyster tissue and then placing them in aquaria
with disease-free Seaside oysters. The habits of oyster fishes were
observed in aquaria for a two-year period (196 0-61).

DEMONSTRATION OF D . MARINUM IN SCAVENGERS

On 12 October 1959 an adult goby, Gobiosoma bosci, from
the hinged shells of a dead oyster from Messongo Creek, was cultured
in thioglycollate medium for Dermocystidium . After culturing and
staining, numerous fungus cells were observed covering most of the
caudal myomeres; most of the remainder of the fish had disintegrated.
Since this observation, Dermocystidium has been demonstrated in the
stomach, feces (Fig. 1), and on the skin of fishes, in the digestive
systems of mud crabs and drills, and covering the body and among
setae on the legs of crabs (Table 1). All of these had just come from
oysters recently killed, or had been fed infected tissue . Nearly all
scavengers from gaping oysters were positive.

Goby feces consist of highly digested remains of oyster tissue
and more definite fecal "pellets" which are apparently the remains of
small animals and scattered sand grains. In a few cases, Dermocystid-
ium cells seen in the digestive system were in eroded oyster tissue
recognizable as gill or mantle, but most fungus cells were found with
numerous colorless fat globules suspended in the liquid intestinal
contents, or in mucus. Dermocystidium was always found abundantly,

-163-

Fig . 1 . Dermocystidium marlnum in feces of Chas modes bos-
quianus . Thioglycollate culture after three days. Iodine stained.

if present at all, in what appeared to be the remains of oyster tissue,
but it was usually scarce in the fecal "pellets."

These observations showed that Gobiosoma bosci, Chasmodes
bosquianus, and Opsanus tau ingest and defecate cells of D. marinum
that respond to the thioglycollate test, and that pieces of infected
tissue or mucus may attach externally to fishes and crabs . Since the
fungus enlarged when cultured properly, and took the iodine stain, it
must have been alive. Mackin and Boswell (1955) concluded that all
stages were infectious .

When small fish were fed Dermocystidium -Infected oyster meat
and then placed in aquaria with disease-free oysters, some of the oysters
developed Dermocystidium infections (Table 2). In experiments 1 through
5 only G . bosci was used, but C . bosquianus and Hypsoblennius hentzi
were added in experiment 6 . In spite of the small number of fish used
and the small amounts of infected tissue they had eaten, the results indi-
cate that G. bosci, at least, can transmit infection to oysters.

-164-

Table 1. Records of Dermocystidium marinum in scavengers. Animals
from aquaria had been fed heavily infected oysters; those
from natural waters had been found in recently dead infected
oysters .

Location of Numbi

5r positive

Species

Locality

fungus for D .

marinum

Gobiosoma bosci

Messongo,

Chesconessex

Skin, stomach

5

Occahannock ,

Cherrystone

Intestine

Gulf

Aquarium

Feces

3

Chasmodes

bosquianus

Gulf

Digestive system

1

Aquarium

Feces

4

Opsanus tau

Nandua

Stomach

1

Aquarium

Stomach, feces

2

Urosalpinx

cinerea

Gulf

Stomach

1

Neopanope texana

Gulf

Covering body &
legs

2

Rhithropanopeus

harrisii

Occahannock

Covering body &
legs

1

-165-

Table 2. Experimental infection of oysters by D . marinum from fishes.
Temperatures were 20-24 C. All experiments were terminated
after approximately 1 month except no. 3 which lasted 6 weeks
All infections were light.

Number
of fish

Experime

ntal

oysters

Control oysters

Exper .

Number

Number

Number

Number

Number

Nu

mber

no .

added

alive

dead

infected

alive

dead

infected

I3

5

23

2

0

25

4

0

2a

5

21

4

2

21

4

0

3a

38

35

12

2

19

3

0

4a

36

5

18

0

17

2

0

5b

34

0

21

3

0

25

0

6°

44

4

21

2

20

5

0

a Seaside water, salinity 29-33 ppt .
b Evaporated Bayside water, 3 2-34 ppt
c Bayside water, 22-24 ppt.

OBSERVATIONS ON SCAVENGING

One of the most ubiquitous and conspicuous scavengers is the blue
crab, Callinectes sapidus . A single adult crab can consume a whole oys-
ter. Both blue crabs and large Panopeus herbstii can carry or drag a whole
killed oyster or one valve with the meat . Large Panopeus were capable of
moving clumps of oysters they were hiding under. Whenever these two
crabs were present, they dominated scavenging. Eurypanopeus depressus
was reluctant to enter killed oysters while larger crabs were feeding.

Sometimes snails, mainly Nassarius vibex, would enter and begin
feeding on killed oysters in experimental wire cages . They seemed to
consume small amounts of meat and were usually the last scavengers to
begin feeding . When killed oysters were placed around the periphery of
a reef, large numbers of N. obsoletus from the nearby flats would feed
on them . Both Urosalpinx cinerea and Eu pleura caudata were found
feeding on recently dead oysters . Although they are widely studied

-166-

predators their scavenging is little mentioned . Demonstration of living
Dermocystidium in the digestive systems of crabs, drills, and fishes
caught in the study area indicates that they had recently scavenged
oysters .

Although most species in the proximity of a reef would eat oys-
ter meat, it may be significant that several would not. Fundulus
heteroclitus , f_. majalis , and a species of Palaemonetes showed interest
in killed oysters but none were observed to eat. However, F_. hetero-
clitus ate loose meat when the shells were pulled apart and Palaemonetes
has eaten meat in aquaria . Fundulus seems afraid to enter partly closed
shells .

Although crabs and snails feed quietly, observations showed
that fishes were the most voracious of scavengers . Due to their mobility
they are often the first scavengers to enter killed oysters . While feeding,
G. bosci tears off pieces of tissue; often several individuals simul-
taneously twist, spin, and turn, scattering bits of meat. A single killed
oyster never failed to attract a few of these gobies, and often they were
very numerous .

All species known to scavenge on the Eastern Shore of Virginia
are listed below . These are included on the basis of direct observations
in natural waters and circumstantial evidence such as the presence of
Dermocystidium . This list is obviously incomplete and probably all
motile animals living with oysters scavenge. However, it seems certain
that a few species (Goblosoma bosci, Chasmodes bosqulanus, Opsanus
tau, Eurypanopeus depressus, Panopeus herbstii, Rhithropanopeus
harrisli, Callinectes sapidus , Urosalpinx clnerea, Eupleura caudata,
Nassarius vibex, and N. obsoleta on native reefs and these plus
closely related species on planted bottoms) account for most tissue con-
sumed in the study area. Fishes, crabs, and snails came to the vicinity
of killed oysters within minutes, regardless of the hour of the day or
night. Most studies, however, were conducted during afternoon hours.

ABUNDANCE OF SCAVENGERS

There is very little information on the density of oyster associates .
As previous authors have noted, relatively few species on oysters are
very abundant. In fact, only Nassarius vibex and Eurypanopeus depressus
were abundant at The Gulf on native oysters, but Gobiosoma bosci and
Panopeus herbstii were not uncommon. The only other scavengers on the
reef were Gobiesox, Opsanus, and Chasmodes, which were comparatively
rare. Other reefs nearby and at other localities varied somewhat but the

-167-

dominance of snails, mud crabs, and gobies was apparent everywhere
on native oysters . Oysters planted on subtidal bottoms acquire a more
varied fauna, but the scavengers are similar. Annelids, which were
not studied, are much more abundant on subtidal oysters .

The reef studied at The Gulf is situated at about low tide level,
but it is a rare tide that exposes all of the reef. Such a tide occurred
on 6 October 1961 and afforded an opportunity to measure the abundance
of G . bosci . Apparently most of the fish in the reef migrated to the
edge and to small pools in the reef. This migration is common on Sea-
side reefs also, during ebb tide. Most fish were then left behind by
the tide; relatively few abandoned the reef for the nearby flats . The
fish were easily captured, and a total of 184 was taken on half of the
reef, an area about 40 feet long and 10 feet wide, by picking up clumps
of oysters along the periphery. All fish were not captured due to rising
tides, but it is believed a majority were. Later observations after the
tide inundated the reef showed no fish attracted to killed oysters as had
always before been the case. Since that part of the reef sampled was
estimated to have 400 square feet of oysters, an estimate of 0.46 fish
per square foot is made . During the low-water period the fish were
concentrated in a narrow band a few inches wide, a concentration of
6 .6 fish per linear foot. As many as 17 gobies were taken under a
single clump of oysters . These figures are probably a fair minimum
index of goby concentrations in autumn .

Nine square-yard samples on 6 and 19 October yielded counts of
15, 16, 16, 18, 18, 19, 24, 25, and 29 2-to-4-inch oysters, an average
of 2 0 oysters per square foot or a total of 17, 000 on the reef. Six square-
yard collections of mud crabs, E_. depressus, from The Gulf on the same
dates yielded counts of 4, 6, 8, 8, 13, and 14. This gives an estimate
of 7,830 E_. depressus on the reef. Estimates of abundance of macro-
associates of the reef are given in Table 3 . These estimates closely
match observations on scavengers, the most abundant forms appearing
to consume the most meat proportional to size . Other than a few barna-
cles, there were no other animals found associated with these oysters .

SCAVENGING RATES AND DETERIORATION OF OYSTERS

A rough idea of the amount of oyster tissue consumed by scavengers
cen be computed from data obtained from trays of live oysters maintained
at a number of stations on the Bayside of the Eastern Shore and examined
at intervals averaging 2 0 days from May through November 196 0. This
encompasses the Dermocystldium mortality season in the area . Of 1338
dead oysters taken, only 156 (11%) had any meats left.

-168-

Table 3 . Estimated abundance of oysters and scavengers on intertidal
reef atThe Gulf, October 1961. a

Species

Average
number
per ft2

Total number
on reef

Biomass,

kg

Crassostrea virginica

20

17,400

7 00b

Eurypanopeus depressus

9

7,830

4.4

Nassarius vibex

1

> 600

?

Gobiosoma bosci

0.5

> 400

0.2

Panopeus herbstii

0.3

> 261

?

a The fishes Chasmodes

bosqu

few to

snail

ianus, Gobiesox strumosus, and

Opsanus tau were too
sapidus, and the mud

estimate .
Nassarius

The blue crab,
obsoletus were

Callinectes
not regular

inhabitants of the reef but invaded it sporadically in unpredictable
numbers .

b Shell weight accounts for 6 00 kg .

Assuming that oysters die randomly between examinations and
that deterioration of oysters tends to be linear, then the tissue of an
average oyster lasted a little over two days after death. Actually only
2% of the dead oysters were taken immediately after death (based on
condition of meats), indicating that an average tray oyster lasted only
0.4 days before it had lost some meat. This seems too fast to explain
by bacterial activity alone, and the destruction probably resulted from
a combination of scavenging and decay. These figures agree with those
of Gunter et al . (1957) who found that oyster meats disappeared in about
two days in the summer at 28 C. Their studies, like these tray observa-
tions, were not made on natural oyster reefs.

Finding recently killed oysters with intact meats on natural
bottoms is difficult. In fact, in all our scavenger studies on oyster
beds we never encountered a gaping oyster, although oysters were
dying .

The results of experiments conducted with killed oysters on a
natural reef at temperatures of 24 to 30 C are shown in Fig. 2. Oyster
meat exposed to scavengers was always consumed in one day, and

-169-

TIME IN DAYS

Fig. 2. Comparison of meat losses from bacteria in protected
oysters and from scavengers plus bacteria in exposed oysters.

half or more was consumed in a few hours . In the controls bacterial
decay destroyed most of the tissue in four days or more, and half in
three days, the curve correlating with many culture growth curves of
bacteria . Within one day oysters kept from scavengers showed no
evident deterioration and weights indicated little had been lost. Brief
studies at 18 to 24 C indicated that both curves shift to the right, but
most meat was still consumed by scavengers within one day and all
within two days . Groups of killed oysters exposed to scavengers on
open bottom, uninhabited by oysters, lost little more meat than con-
trols, presumably because scavengers were not present there.

The figures obtained from these experiments have two sources
of error: (1) Killing ten oysters saturated a small area with a large
amount of meat. Whenever a single oyster was killed, its consumption,
at least to the muscle, was measured in minutes rather than hours .
(2) Dying oysters probably are invaded by decay bacteria some hours
prior to death, so perhaps the decay curve should be shifted to the left
to represent what actually occurs .

170-

The death point of oysters needs further study, using naturally
dying oysters. Oysters die gradually and scavenging (sensu lato) may
begin before the oyster is technically dead. Small gobies (G . bosci)
often enter gaping oysters in aquaria and feed on gill tissue while
these oysters still have the power of complete closure . An oyster
sometimes closes on a goby, rarely catching it at mid-body and killing
the fish, but more often temporarily trapping the fish inside. Pathogens
such as D . marinum which cause lysis of tissue may speed up deteri-
oration, although Ray et al . (1953) did not believe this accounted for
decay of the oyster after death.

In any case, it seems significant that all meat was always
eaten by scavengers in a relatively short time. Observations showed
that the meat was actually eaten, not just removed from the shells .
It is difficult to demonstrate 100% consumption, but the motivation
obviously exists .

DISCUSSION

Since oysters form the basis of an extensive estuarine com-
munity with many dependent organisms, any pathogen of oysters is
significant to numerous plants and animals . A certain amount of
oyster mortality seems to be normal and is of considerable value to
the community. The absence of oyster mortality would limit feeding
and spawning of some associated species. On the other hand, exces-
sive mortality may provide more food than can be absorbed by the com-
munity, and it removes the oysters which are the most important
member, the dominant species on which the existence of the community
depends .

Hopkins (1957) stated that a common effect of marine parasites
is to increase the host's susceptibility to predators. Menzel and
Hopkins (19 56) noted that blue crabs, Callinectes sapidus, destroyed
many spat, but ate only weak and dying adult oysters . The same is
true of mud crabs (McDermott, 196 0). This was also true of oyster
drills, Urosalpinx and Eupleura, and other snails in the study area .
The scarcity of recently killed oysters with intact meat on natural bot-
toms, and the observations on artificially killed oysters, indicate that
nearly all oyster tissue infected with D . marinum is consumed by
scavengers, at least during normal or less extreme mortalities. This
would force almost all oyster tissue parasites to pass through animals
other than oysters .

Spawning of G . bosci in recently killed oysters on the bayside
of the Eastern Shore occurred largely from 15 June to 15 August

-171-

(unpublished data). This is when infections of Dermocystidium buiid
up in live oysters (Andrews and Hewatt, 1957). Ray (1954) noted that
oysters in Louisiana placed in endemic waters in June suffered higher
mortality than those placed there in late August. This is also true of
transplants of highly susceptible Seaside oysters into Chesapeake Bay.

As an oyster has more fungus cells available for release, it
presumably will be more susceptible to attack by other animals .
Andrews and Hewatt (1957) believed that disintegrating gapers account
for most infective material, and Ray (1954) showed that infection by
live oysters was much slower than other methods . Although it is not
certain that live oysters can release large numbers of infective spores,
dead oysters do, and subsequent transmissions could be due, at least
in part, to scavengers, by means hypothesized in Fig. 3. The very
least that scavengers may do is to speed up release of oyster parasites
and prevent production of bacterial metabolites .

LIVE OYSTER-

S-RELEASES SPORES

DEAD OYSTER-

BACTERIAL DECAY
RELEASES SPORES

OYSTER ENTERED
BY SCAVENGER

OYSTER TISSUE
-•-AND MUCOUS-

-mDECAYS LATER

STICKS TO BODY
EATEN BY SCAVENGER

CARRIED BY
SCAVENGER TO
OTHER HOST
INFECTION FROM
FECES

TISSUE /
v SCAT T ERE D\

EATFN

DECAYS

SPORES PASSED
NEAR DEAD OYSTER
INFECTION BY WATER
CURRENTS

Fig . 3 . Theoretical routes traveled by D . marinum in natural

waters

-172-

ACKNOWLEDGMENTS

Many oystermen and dealers graciously gave their time, oysters,
and equipment, andhelped catch specimens, particularly A . M. Acuff,
Ralph Clark, Elwood Gaskins, and W. E. Walker. Mr. A. M. Akuff
kindly loaned an oyster reef for scavenger studies . Dr . J . D . Andrews
introduced the writer to the Chesapeake area and supplied much informa-
tion on Dermocystidlum. Dr. S. H. Hopkins provided information and
advice, and assisted in summer field studies. W. T. Davis, and later
R. D. Hickman and Bonnie Callaway, assisted in both field and labora-
tory studies . Mr . W . T . Davis drew Figs . 2 and 3 . Mr . Tom K . Burton,
Jr. photographed Fig. 1. Mr. Ken Parks of the Accomack County Health
Department sterilized media and equipment and helped in other ways .

LITERATURE CITED

Andrews, J. D. and W. G. Hewatt. 19 57. Oyster mortality studies in
Virginia. II. The fungus disease caused by Dermocystidlum
marinum in oysters of Chesapeake Bay. Ecol . Monographs
27: 1-25.

Gunter, G. 1955. Mortality of oysters and abundance of certain
associates as related to salinity. Ecology 36:601-605.

Gunter, G., C. E. Dawson and W.J. Demoran. 1957. Determination
of how long oysters have been dead by studies of their shells .
Proc . Nat'l Shellfish. Assoc. 47:31-33.

Hedgpeth, J . W. 19 53 . An introduction to the zoogeography of the
Northwest Gulf of Mexico with reference to the invertebrate
fauna . Publ . Inst. Mar. Sci. 3: 107-224.

Hopkins, S. H. 1957. Interrelations of organisms . B. Parasitism,
in Treatise on Marine Ecology and Paleoecology . Mem. Geol .
Soc . Amer. 67:413-428.

Johnson, T. W. and F. K. Sparrow. 1961. Fungi in oceans and
estuaries. J. Cramer, Weinheim, Germany. 668 p.

Korringa, P. 1951. The shells of Ostrea edulis as a habitat. Arch.
Neerland. Zool . 10: 32-152 .

Mackin, J.G. 1962. Oyster disease caused by Dermocystldium

marinum and other microorganisms in Louisiana . Publ. Inst.
Mar. Sci. 7: 132-229.

173-

Mackin, J. G. and J. L. Boswell . 1956. The life cycle and relation-
ships of Dermocystidlum marinum . Proc . Nat'l Shellfish . Assoc .
46: 112-115.

McDermott, J.J. 1960. The predation of oysters and barnacles by

crabs of the family Xanthidae. Proc. Pennsylvania Acad. Sci.
34: 199-211.

Menzel, R. W. and S. H. Hopkins. 1956. Crabs as predators of

oysters in Louisiana. Proc. Nat'l Shellfish . Assoc . 46:117-184,

Parker, R. H. 1955. Changes in the invertebrate fauna, apparently
attributable to salinity changes, in the bays of central Texas.
J.Paleontol. 29:193-211.

Parker, R. H. 1959. Macro-invertebrate assemblages of Central Texas
coastal bays and Laguna Madre . Bull. Am. Assoc. Petroleum
Geol. 42 (9): 2100-2166.

Ray, S. M. 19 52. A culture technique for the diagnosis of infection
with Dermocystldium marinum in oysters . Conv . Add . Nat'l
Shellfish. Assoc . 43:9-13.

Ray, S. M. 19 54. Biological studies of Dermocystidlum marinum, a
fungus parasite of oysters . Rice Inst. Pamph. Monographs in
Biology: 1-114.

Ray, S . M . , J . G . Mackin and J . L . Boswell . 1953 . Quantitative
measurement of the effect on oysters of disease caused by
Dermocystidium marinum. Bull. Mar. Sci. Gulf Carib. 3:6-33.

Wells, H. W. 1961. The fauna of oyster beds, with special reference
to the salinity factor. Ecol . Monographs 31:239-266.

-174-

A TECHNIQUE FOR SEPARATING SMALL MOLLUSKS
' FROM BOTTOM SEDIMENTS

Gareth W. Coffin and Walter R. Welch

U.S. Bureau of Commercial Fisheries Biological Laboratory
Boothbay Harbor, Maine

ABSTRACT

Studies of the soft-shell clam, Mya arenaria, in New England
required frequent examination of bottom samples in which clams ranged
in length from 0.3 to approximately 100 mm. To facilitate separating
smaller clams from sediments, a technique combining the advantages
of screening, elutriating, and decanting was developed. The method has
been used successfully with six other species of pelecypods and gastro-
pods and appears adaptable to other species of unattached mollusks.

INTRODUCTION

Studies of the soft-shell clam, Mya arenaria, at the U.S.
Bureau of Commercial Fisheries Biological Laboratory, Boothbay Harbor,
Maine, from 1950 to I960, required frequent, quantitative examinations
of bottom samples . The size (length) of Mya in the sediments ranged
from 0.3 to about 100 mm. The larger clams (over 5 mm) were not diffi-
cult to sort, count and measure, but the smaller clams required a
special technique for efficient recovery.

Many workers have used flotation or screening methods, or
various combinations of the two, in separating particular fractions.
Pfitzenmeyer (19 56) used screening alone to obtain 0.3 to 0.5 mm Mya .
However, he used specially prepared trays containing fine sediments
to collect new set, then passed the sediments through a fine screen
which retained the clams. Sellmer (1956), working with Gemma gemma,
first screened his samples, then separated mollusks from sand in a
solution of zinc chloride. Anderson (1959) reviewed a number of flota-
tion techniques and used a sugar solution to remove certain inverte-
brates from fresh-water benthic samples, but indicated the method was
not as satisfactory for mollusks. Teal (1960) also used a sugar solu-
tion combined with screening and centrifuging in working with core
samples, but was mainly concerned with recovering nematodes and
arthropods. Williams (1960) utilized the activities of live fresh-water

-175-

invertebrates which would move from the original sediment sample into
a covering layer of fine sand, from which they could be removed by
screening. When organic detritus was present he floated the animals
out in a solution of magnesium sulfate .

COLLECTION AND PRELIMINARY HANDLING OF SAMPLES

Subsamples for processing in the laboratory were taken from the
surface of the clam flats, measured 6 inches square by 2 inches deep,
and included the smallest clams . The clams deeper than 2 inches were
sufficiently large to be readily dug out or screened in the field .

The laboratory sample was divided into two portions . The upper
half-inch ("surface portion") was skimmed off. This layer included the
smallest clams, ranging from 0.3 to approximately 5 mm. The remainder
of the sample, the "subsurface portion," included clams ranging in size
from 5 to 20 mm .

LABORATORY SEPARATION

Step 1

The subsurface portion was placed in a MacDonald hatching jar
for elutriation * (see Figs. 1 and 2). A jet of water was introduced
at the bottom of the jar through a glass tube constricted at the end. The
flow which would remove the maximum amount of sediment without loss
of clams was found by examining the overflow (A in Fig . 2). Any num-
ber of subsurface portions could be started through this initial step at
the same time .

Step 2

The surface portion was first washed down through a series of
seven screens, each 8 inches in diameter. The mesh sizes, from top
to bottom, were 2.6, 5, 10, 18, 38, 52, and 70 meshes per inch. The
first two screens retained pieces of shell or other large particles . All
clams larger than 2 mm were retained on the 10- and 18-mesh screens .
The contents of these screens were washed into separate white enamelled
trays, from which the clams could be readily picked with forceps.

The term "elutriate" is used throughout in its common but more restric-
tive sense, meaning "to separate lighter particles from heavier by
washing ."

-176-

SUBSURFACE

PORTION

SURFACE PORTION

STEP I

i

STEP 3

n

z

UJ
Ul

OE

u
01

STEP 2

ELL

TRIAT

ON

CO

z

UJ
UJ

CE

2.6-MESH

^\ r

5 - MESH

\ r

|

10- MESH

" J COUNTING

___^ 1 TRAY

to

^\ [_

18-MESH

~~ 1 COUNTING

____^ 1 TRAY

58-MESH

"~~" — 1 COUNTING
^^-—] TRAY - 1.4 X

u_ \ r

STORAGE

CONTAINER

52-MESH

"A r

70-MESH

—J STORAGE
___ — - 1 CONTAINER

"lJ"

| J

2.6-MESH

STEP 4

\ 1

COUNTING
CHAMBER-9X

5 ■ MESH

~\ r

STEP 5

COUNTING L-"^"
TRAY ( __

10

-MESH

J [_

\

\ /

OECANT

1 80-MESH SCREEN

|-— — ^_ a WATER JET

COUNTING L- """

TRAY |

18- MESH

\ /

COUNTING
CHAMBER-9X

Fig . 1 . Flow diagram indicating the sequence of separation and
the progression of the two portions of a given sample .

-177-

a.

QJ

QJ

X

3

3
c

CD
r+

CD

3

-i -a

•"-*-

t-h

W ,Q

1 — '

o .

o

•-1

$

2 t^>

ft) .

o
3

3
.Q

rt C

r>

3* P-

3"

cd -a

(D

\ — ..

<* 3

S, CD

QJ
in

QJ 3

O r+

O c

O CO

3 CD

QJ 0-

1—'

O- £•

o

QJ CO

-i CD

CO ,-,

CO QJ

CD i-t

CL CD

i-h a

o 5

l=i o

H

!Lc

C en

,-i en

QJ

.. l_+i

r+ ^

o °

3 3

en

QJ CD

3 a

ex £'

^ 3

> ro

s 3

g. to

CD *

0 nj

O -i

3 O

oT 3

t-*- , .

3 CD

CD t-h

►1 r+

C J?

w O

CD -i

^ id

?T 3d

O rt-

CD

CO r+

r+ 3"

QJ CD

CT" ^

P- 00

*-- 1

to ^.

3" 3

o

3- 3*

CD

-17!

The contents of the 38-mesh screen, which included clams
between 1 and 2 mm long, were also washed into an enamelled tray,
but a 1 .4X magnifier and pipette were necessary to locate and remove
the clams . The 52- and 7 0-mesh screens retained the rest of the clams
in the sample, all of which were in the size range of 0.3 to 1 mm.

Step 3

When the elutriation of the subsurface portion had been com-
pleted, the remaining sediment was washed down through a series of
four screens with mesh sizes of 2.6, 5, 10, and 18. The larger clams
could be picked directly off the 2.6- and 5-mesh screens. The contents
of the 10- and 18-mesh screens were washed out into separate enamelled
trays for examination and removal of the clams . Excess organic detritus
was decanted from the tray and the clams were then removed with forceps.

Step 4

The contents of the 52-mesh screen, obtained in Step 2, were
washed into a 150-mm petri dish which had a 10-mm counting grid
engraved in the bottom. With the aid of a 9X stereoscopic microscope
the clams, ranging in size from 0 . 5 to 1 mm, were removed with a pipette.
When a sample included much organic detritus, the sample was stirred
with water in the petri dish and the lighter detritus was carefully poured
off, leaving the clams and heavier sediments.

Step 5

The material from the 7 0-mesh screen, obtained in Step 2, was
placed in a 3-inch diameter, 80-mesh screen with an 18-mesh screen
on top and washed with considerable force until the water coming through
was no longer muddy. The contents of this 80-mesh screen were then
washed into the grid-marked petri dish and the lighter detritus was de-
canted from the dish as described in Step 4.

When the contents of the 7 0-mesh screen, obtained in Step 2,
included so much sand that the clams could not be found readily, the
contents were washed into a plain 150-mm petri dish. A rapid oscillatory
motion of the dish brought the clams to the surface of the sand and they
were quickly decanted into the petri dish with the counting grid. The
clams, ranging in size from 0.3 to 0.6 mm, were found with the aid of
the 9X microscope and removed with a pipette . The decanting process
was repeated until no more clams were found.

-179-

DISCUSSION

To take fullest advantage of the time-saving features of the
technique described herein, we adapted the various steps to the
characteristics of the animal being sampled and the sediments in which
It was found. With Mya arenaria from Maine, for instance, we found
that in a year-round sampling program the 52- and 7 0-mesh screens
(and therefore Steps 4 and 5) could be omitted from late October to
early June. This was possible because the metamorphosed larvae did
not set before early June, and by late October the setting had ceased
and all clams were at least 1 mm long.

In a limited test to determine the time required for processing
we utilized samples from a variety of bottom types, containing various
numbers and sizes of clams . The processing time (not including removal
of the tiny clams from the 52- and 7 0-mesh screens) for five such samples
varied from 11 to 19 minutes and averaged 15 minutes per sample.

The technique described here has also proved successful in
separating the following mollusks from the sediments: Mercenaria mer-
cenaria , Macoma balthica, Petricola pholadiformis, Hydrobia minuta ,
Gemma gemma, and Mysella planulata . The method appears to be adapt-
able to the separation of most unattached pelecypods and gastropods from
the sediments .

LITERATURE CITED

Anderson, R. O. 1959. A modified flotation technique for sorting bottom
fauna samples. Limnol . Oceanog . 4:223-225.

Pfitzenmeyer, T. H. 1956. Soft clam set observed at Solomons. Mary-
land Tidewater News 13(4): 1,2.

Sellmer, G. P. 1956 . A method for the separation of small bivalve
molluscs from sediments. Ecology 37:206.

Teal, J . M. 196 0. A technique for separating nematodes and small
arthropods from marine mud. Limnol. Oceanog. 5:341-342.

Williams, R. W. 196 0. A new and simple method for the isolation of
fresh-water invertebrates from soil samples. Ecology 41:573-
574.

-180-

ASSOCIATION AFFAIRS
ANNUAL CONVENTION

The 1962 Convention was held jointly with the Oyster Institute
of North America and the Oyster Growers and Dealers Association of
North America, Inc . , on July 29— August 2, 1962, at the Emerson Hotel,
Baltimore, Maryland.

The Secretary-Treasurer reported that the Association had 15 0
members in good standing, including 10 new members. Mr. David H.
Wallace, former director of the Oyster Institute was elected to Honor-
ary Membership. His biography and picture will be printed in Volume 54

The first Thurlow Nelson award, consisting of a scroll and a
five-year membership in NSA, was given to Albert F. Eble of Trenton
Junior College for his work on blood circulation in oysters .

Dr. John Glude announced that Microcard reprints of the Pro-
ceedings from 193 0 to 196 0, each set consisting of 28 cards (4 0 pages
per side), are for sale by the Secretary-Treasurer at $8.00 per set.

The 1961-62 officers were re-elected for 1962-1963. They were
as follows:

President Philip A. Butler

Vice-President John B. Glude

Secretary-Treasurer . .Jay D. Andrews

Members-at-Large . . . Dana E. Wallace and John G. Girard

John B. Glude served as program chairman for the 1962 conven-

tion ,

Sewell H. Hopkins (Chairman), Lawrence Pomeroy, and Daniel
B. Quayle were re-appointed to the Editorial Committee.

-181-

INFORMATION FOR CONTRIBUTORS

i
Original papers given at the Annual Association Convention and

other papers on shellfish biology or related subjects submitted by mem-
bers of the Association will be considered for publication. Manuscripts
will be judged by the Editorial Committee or by other competent review-
ers on the basis of originality, contents, clarity of presentation and
interpretations. Each paper should be carefully prepared in the style
followed in previous PROCEEDINGS before submission to the Editorial
Committee. Papers published or to be published in other journals are
not acceptable .

Manuscripts should be typewritten and double-spaced: original
sheets are required but extra copies will facilitate reviews. Tables,
numbered in arabic, should be on separate pages with the title at the
top. Scientific names should be underlined. Illustrations should be
reduced to a size which fits on 8 x 10 1/2 inch pages with ample margins.
Glossy photographs are preferred to originals . Illustrations smaller than
a page should be carefully oriented and loosely attached to plain white
paper with rubber cement. Legends should be typed on separate sheets
and numbered in arabic.

Authors should follow the style prescribed by Style Manual for
Biological Journals, which may be purchased for $3.5 0 from the American
Institute of Biological Sciences, 2000 P Street, NW, Washington 6, D.C.
In case of a question on style that is not answered by this manual, the
author should refer to the 1961 PROCEEDINGS (Volume 52) or to the present
volume .

Each paper should be accompanied by an abstract which is con-
cise yet understandable without reference to the original article. It is
our policy to publish the abstract at the head of the paper and to dis-
pense with a summary . A copy of the abstract for submission to Biologi-
cal Abstracts will be requested when proofs are sent to authors .

Reprints and covers are available at cost to authors . Master
sheets will be retained for one year after publication. When proof sheets
are returned to authors, information about ordering reprints will be given.
The present agency from which authors may obtain reprints is Bi-City, Inc.,
1001 S. College Avenue, Bryan, Texas.

For inclusion in the PROCEEDINGS of the current year, all manu-
scripts should reach the Editor prior to October 1, and none will be
accepted after December 31. Send manuscripts and address all corres-
pondence to the Editor, Dr. Sewell H. Hopkins, Biology Department,
Texas A&M University, College Station, Texas.

-182-

MBL WHOI LIBRARY

WH 1ACJ 5