Toxicity testing with fish, zooplankton and mussels


Material Information

Toxicity testing with fish, zooplankton and mussels a comparison of sensitivities
Physical Description:
ix, 212 leaves : ; 28 cm.
Keller, Anne E., 1952-
Publication Date:


Subjects / Keywords:
Unionidae -- Effect of water pollution on   ( lcsh )
Zooplankton -- Effect of water pollution on   ( lcsh )
Freshwater fishes -- Effect of water pollution on   ( lcsh )
Water -- Pollution -- Toxicology   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1989.
Includes bibliographical references (leaves 190-211).
Statement of Responsibility:
by Anne E. Keller.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001529328
notis - AHE2688
oclc - 22287966
System ID:

Full Text







Copyright 1989

Anne E. Keller

To my parents,
Robert and Lucille Keller,
whose support and confidence in me
made all the difference


I would like to thank Dr. Thomas L. Crisman, my

committee chairman, for offering me many of the

challenges that have been part of my education. My other

committee members, Dr. Gabriel Bitton, Dr. Frank Nordlie,

Dr. Stephen G. Zam and Dr. Clay Montague, broadened my

horizons via the different perspectives they hold. I am

grateful for that.

Many other people in this department and elsewhere

were instrumental in the completion of this work. I

thank them for their generosity and care. Friendship

often makes frustration and hard work less difficult.

Though they deserve much more, I thank my family and

friends for their faith and trust in me.



ACKNOWLEDGMENTS ...................................... iv

ABSTRACT............................................. viii


1 INTRODUCTION ............................... 1

2 LITERATURE REVIEW.......................... 6

Toxicity Testing with Pelagic Biota........ 6
Rationale for the Use of Freshwater
Molluscs in Acute Toxicity Tests......... 10
Distribution and Life History of Unionid
Mussels.................................. 11
Habitat Destruction and Faunal Decline..... 13
Propagation of Freshwater Mussels in
Artificial Media.......................... 16
Sensitivity of Unionid Molluscs to
Environmental Pollutants................. 18

150 and 250 C.............................. 29

Introduction............................... 29
Materials and Methods..................... 33
Test Organism ............................ 33
Test Organism food........................ 33
Dilution Water............................ 33
Test Chemical ............................ 34
Reference Toxicant....................... 34
Range-finding Test........................ 35
Seven-day Survival and Growth Toxicity
Test.................................... 35
Statistical Analysis..................... 36
Results..................................... 37
Dilution Water Quality................... 37
Survival and Growth of Fathead Minnow
Larvae................................. 37
Discussion................................. 49

REPRODUCTION TEST.......................... 56

Introduction .............................. 56
Materials and Methods...................... 58
Test Organism........................... 58
Test Chemical............................ 58
Dilution Water........................... 58
Test Organism Food...................... 60
Reference Toxicant Tests................ 60
Range-finding Test....................... 61
Preparation for Chronic Toxicity Tests.. 61
Chronic Toxicity Tests.................. 62
Statistical Analysis.................... 63
Results..................................... 64
Reference Toxicant...................... 64
Acute Toxicity........................... 64
Chronic Toxicity......................... 67
Discussion.................................. 74


Introduction............................. 79
Materials and Methods................... 81
Test Organism......................... 81
Plasma Substitutes................... 82
CO2 Incubator......................... 84
Use of Commercial Media.............. 85
Species Collected.................... 86
Results.................................. 87
Discussion............................... 95

FRESHWATER MUSSELS........................ 98

Introduction................................ 98
Materials and Methods...................... 104
General Conditions..................... 104
Physical Conditions..................... 105
Water Quality............................ 106
Feeding Tests............................ 108
Test Organisms........................... 109
Results .................................... 114
Discussion.................... .............. 118



Introduction ............................... 122
Materials and Methods...................... 127
Test Organisms .......................... 127
Test Methodology........................ 127
Dissolved metals..................... 127
Metal mixtures....................... 128
Sediment tests.......................... 129
Metal effluent........................ 130
Test chemicals...... ....... ........... 131
Data Analysis......................... 131
Results........... ..................... 133
Dissolved Metals........................ 133
Metal Mixtures........... .... .......... 143
Sediment Tests ................. .......... 147
Effluent Toxicity....................... 148
Discussion.................................. 150


Introduction............................... 156
Materials and Methods....................... 162
Test Organisms.............. ... ......... 162
Test Conditions......................... 163
Aqueous exposures.................... 163
Karate, atrazine and carbaryl......... 164
Toxaphene and chlordane tests........ 165
Effluent toxicity test................ 167
Data Analysis............................ 168
Results..................................... 168
Aqueous Tests ........................... 168
Karate, Atrazine and Carbaryl........... 174
Toxaphene and Chlordane Tests........... 174
Effluent Toxicity Test................... 178
Discussion.................... .............. 178

9 CONCLUSIONS. ............................... 184

REFERENCES... ........................................ 190

BIOGRAPHICAL SKETCH................................. 212


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of
the Requirements for the Degree of Doctor of Philosophy



December 1989

Chairman: Thomas L. Crisman
Major Department: Environmental Engineering Sciences

Toxicity testing with aquatic organisms is common-

place today. Prior to their manufacture and use, the

effects of pesticides, herbicides and toxic substances on

biota and the environment must be assessed. However,

the focus has been primarily on the fate of pelagic

fauna, particularly fish and zooplankton. Little

attention has been focused on the responses of

invertebrates, other than insects, to pollutants.

In recent years, there has been a significant

decline in the once abundant freshwater mussel fauna,

purportedly due to dam-building and pollution. Since

little is known about the sensitivity of freshwater

mussels to metals and pesticides, there has been no way

to establish protective measures. Currently, the United

States Environmental Protection Agency is using

zooplankton as surrogates for freshwater mussels in


toxicity tests with no verification that the two are

comparably sensitive.

This dissertation was designed (1) to determine how

sensitive mussels are to metals and organic, (2) to

compare the sensitivity to freshwater fish such as the

fathead minnow and (3) to determine by comparison whether

zooplankton are good substitutes for mussels in toxicity

tests. Anodonta imbecilis was chosen as the test species

because it was locally available, has a relatively long

reproductive period and has been previously cultured in

the laboratory.

Acute toxicity tests were performed with juvenile

mussels in reconstituted freshwater. Copper, cadmium,

chromium, mercury, zinc and nickel were the metals used.

It was found that mussels were about as sensitive to

metals as were zooplankton. Organic compounds assessed

included lindane, toxaphene, chlordane, Hydrothol-191,

PCP, carbaryl, atrazine, an unregistered pyrethroid

pesticide, acetone, methanol and SDS. Anodonta imbecilis

was not sensitive to any of these substances except PCP.

It appears that the use of zooplankton species, e.g.

Daphnia magna or Ceriodaphnia dubia, as surrogates for

freshwater mussels is appropriate in tests for metal

toxicity, but may not be so for organic pollutants.


Although the use of aquatic organisms to test

impacts of industrial, agricultural or wastewater

effluents on the biota of streams and rivers is a common

occurrence today, this is a relatively new development.

Initial concern centered on the safety of chemicals to

humans and domestic animals relative to their efficacy on

target organisms (Casarett and Bruce 1980). However, as

concern for the environment has increased, so have the

number and uses of aquatic toxicity tests. There are now

test methods for many vertebrate and invertebrate aquatic

animals (Peltier and Weber 1985).

Aquatic toxicology arose as an outgrowth of the

chemical revolution of the 1940s. Biologists, seeing

adverse changes in the biota of streams receiving human

and industrial wastes, advocated the use of fish or other

aquatic species as a means of predicting the response of

stream organisms to industrial wastes (Buikema et al.

1982). Regulatory control of water quality was

established in the U.S. with the passage of the Federal

Insecticide, Fungicide and Rodenticide Act (FIFRA) in

1947 and the Federal Water Pollution Control Act (FWPCA)

in 1948. These pieces of legislation regulated the input

of both conventional municipal and industrial waste into

lakes, rivers and the ocean.

In 1970, when United States Environmental Protection

Agency (USEPA) became the agency responsible for

improving and protecting the nation's water resources,

the major concern was still the impact on both drinking

water supplies and harvestable aquatic species.

While sewage treatment and receiving water quality

improved over time, as recently as 1982 some 30 states

were cited for water quality standards violations due to

toxic pollutants (Wise 1985). Biomonitoring of effluents

via acute toxicity tests became the standard method for

assessing environmental impacts because such tests are

inexpensive and directly measure biological response.

Acute toxicity tests are also required for all new

chemicals posing a potential risk to human health or the

environment, and as supporting documentation for

pesticide registration (Zucker 1985a, 1985b).

To date, over 165 species of aquatic organisms have

been used in acute toxicity tests (Buikema et al. 1982).

Rapid methods to assess chronic toxicity have recently

been developed for several species, including Pimephales

promelas, Ceriodaphnia dubia and Selenastrum

capricornutum (Horning and Weber 1985). These new

methods measure sublethal effects of toxicants on biota,

but acute toxicity tests are still the more commonly

used. While it would be convenient to use only one or a

few species to assess the impacts of various pollutants

on aquatic organisms, such a practice would not be

adequate. Species specific tests are necessary because

of differences in biological sensitivity and because

organisms from different habitats may be exposed to a

wide range of pollutant concentrations. The need for a

variety of test methods has been demonstrated by results

of testing with many pollutants (Johnson and Finley 1980,

Mayer and Ellersieck 1986).

Most of the accepted test organisms are pelagic

species (fish and zooplankton) which are well-known, easy

to culture, economically important or of public concern

(Buikema et al. 1982). However, benthic organisms are

more appropriate for use in tests assessing the impacts

of pollution in flowing waters because they typify the

fauna of lotic systems. Pelagic fauna are more

representative of lacustrine habitats. To date, toxicity

data for benthic organisms including aquatic

oligochaetes, turbellarians, pelecypods and gastropods

are still extremely limited, and only one toxicity test

protocol for these organisms exists, i.e., for the exotic

Corbicula fluminea (Foster 1979). Because its life cycle

differs significantly from that of native North American

freshwater clams (Unionidae),the use of C. fluminea as a

model for other mollusk species is questionable.

The current dissertation research is divided into

two large parts. The first part deals with the

assessment of the chronic toxicity of Hydrothol-191, an

aquatic herbicide, to Pimephales promelas (the fathead

minnow) and Ceriodaphnia dubia, using recently developed

EPA protocols. The Florida Department of Environmental

Regulation foresaw an increased demand for Hydrothol use

in Florida and wanted to know what impact this might have

on nontarget organisms.

While both the fathead minnow and C. dubia are used

to monitor the toxicity of wastewater effluents and

assess the impact of pure compounds on aquatic organisms,

neither of them is native to Florida (Lee et al. 1980).

In addition, their responses have not been compared to

those of benthos that inhabit canals, streams or rivers

where Hydrothol is widely used for macrophyte control.

Therefore, the second part of the dissertation contains

results of test development work with a representative of

native benthic fauna, the freshwater mussel, Anodonta

imbecilis. The need for a toxicity test for native

freshwater mollusks was apparent based on their

importance in flowing waters, their taxonomic distinction

from insects and other benthos that have been tested, and

the need to corroborate the use by EPA of tests with

Daphnia magna to estimate the sensitivity of mussels to



My research with A. imbecilis was designed to (1)

simplify the culture techniques permitting easier

production of test organisms, (2) develop an acute

toxicity test protocol for use in assessing the

sensitivity of freshwater mussels to pesticides, metals

and wastewater effluents, and (3) determine the toxicity

of a number of pure compounds and effluents to A.

imbecilis. Results from this work could then be used to

determine whether it is appropriate for EPA to use D.

magna in toxicity tests as a surrogate for freshwater



Toxicity Testing With Pelagic Biota

The bulk of information on the toxicity of

pollutants to aquatic biota was derived from tests with

pelagic organisms. In particular, several species of

economically important fish, e.g., Salmo gairdneri,

Oncorhynchus tshawytscha, Lepomis macrochirus and

Ictalurus punctatus, and a number of zooplankton species,

e.g. Daphnia magna, Daphnia pulex, and Simocephalus spp.

have been the most common test organisms (Johnson and

Finley 1980, Mayer and Ellersieck 1986, Buikema et al.

1982). The latter group has been well studied because

they are both easy to rear in the laboratory and

important links in the aquatic food chain leading to


Since the late 1970s there has been increasing

interest in the development of short term chronic

toxicity tests for fish that combine the simplicity of

acute methods with the estimation of sublethal effects

provided by lifecycle toxicity tests (Horning and Weber

1985). The latter can require months or years to

complete, depending on the lifespan of individual

species. Chronic exposures to low concentrations of

pollutants can affect reproduction, growth, behavior or

species interactions, any of which may alter the

structure of the aquatic community (Alabaster and Lloyd

1982, Rand 1985).

Studies by McKim (1977) and Macek and Sleight (1977)

proved that exposure of critical life-stages of fish

(embryos or larvae) to toxicants for 30-60 days provided

toxicity estimates comparable to full life cycle tests.

These early life stage (ELS) tests were soon adopted as

the standards for estimating water quality criteria

because they were faster and cheaper, as well as accurate

(Horning and Weber 1985).

Further simplification followed as data from ELS

tests showed that larval growth could be as sensitive a

measure of sublethal toxicity as larval survival (Benoit

et al. 1982, Woltering 1984, Birge et al. 1981). As a

result, a seven-day fathead minnow larval survival and

growth test was developed for effluent and single-

compound toxicity evaluations (Norberg and Mount 1985).

The method, published by EPA (Horning and Weber 1986), is

described as a static-renewal subchronic toxicity test

that uses larval growth as a measure of sublethal


Larval fathead minnows (< 24 h old) are exposed to a

series of toxicant concentrations (usually 5) and a

control comprised of dilution water. The test solutions

are changed daily, after a count of survivors has been

made. Larvae are fed rinsed brine shrimp. At the end of

the test, all surviving larvae are preserved in formalin

until their growth can be assessed based on weight gain

compared to that of controls. An LC50 is calculated

using survival data.

The second subchronic toxicity test to be published

by the EPA was the Ceriodaphnia dubia survival and

reproduction test (Horning and Weber 1985). The impetus

behind the development of the 7-day Ceriodaphnia survival

and reproduction was somewhat different from that of the

test with fathead minnows. While test duration was a

factor, it was perhaps more related to the length of the

work week than to expense since a cladoceran life cycle

test may be completed in about 30 days (Mount and Norberg

1984). If the test is begun on a Friday, little

maintenance time is required over the weekend. More

intense effort is required as the test progresses.

Historically, Daphnia magna has been the most used

species for the estimation of zooplankton acute

sensitivity to pollutants (Mount and Norberg 1984,

Buikema et al. 1982, Anderson 1980). Anderson (1980)

described a series of 17 papers produced by Einar Naumann

in 1933 and 1934 detailing various aspects of toxicity

testing with D. magna. This was perhaps the real

beginning of the use of D. magna in such tests.

With movement toward the use of chronic test methods

in aquatic toxicology, both a lifetime (Buikema 1973,

Winner and Farrell 1976) and 21-day chronic test

(Biesinger and Christensen 1972) were developed for D.

magna. As designed, these methods provided estimates of

sublethal effects based on changes in fecundity (Buikema

et al. 1980), but they were still too lengthy.

Under the auspices of the EPA, Mount and Norberg

(1985) developed a 7-d subchronic toxicity test with a

different species, Ceriodaphnia reticulata. C.

reticulata was chosen (over D. magna) because it is

widely distributed in North America, it was easier to

culture than was D. magna and it produces three broods of

young in seven days (Mount and Norberg 1984). These

characteristics facilitate the performance of many tests

in a short time, virtually anywhere. Since the

developmental work by Mount and Norberg (1984), EPA has

suggested the use of C. dubia in their protocol manual

(Horning and Weber 1985).

A Ceriodaphnia dubia survival and reproduction test

is begun with the collection of neonates (<24 h old) from

the adult culture. Neonates are placed in individual

test vessels consisting of 30 ml plastic cups containing

15 ml of solution. Isolation of individuals is necessary

so that separate tallies of fecundity can be maintained

for each animal. A daily count of survivors is made.

Beginning on Day 3 or 4 of the test when the first brood

is produced, offspring are also counted. Adults are then

moved to new test vessels and fed. This daily counting

and transfer to new solutions continues until the test is

terminated at seven days. A 7-d LC50 for adults is

calculated, and their fecundity is used to measure

sublethal effects.

Rationale For The Use of Freshwater Molluscs In Acute
Toxicity Tests

To date, over 165 species of aquatic organisms have

been used in acute toxicity tests (Buikema et al. 1982).

Species specific tests are necessary because of

differences in biological sensitivity and because

organisms from different habitats may be exposed to a

wide range of pollutant concentrations. The need for a

variety of test methods has been demonstrated by results

of testing with many pollutants (Johnson and Finley 1980,

Mayer and Ellersieck 1986).

Most of the accepted test organisms are pelagic

species (fish and zooplankton) which are well-known, easy

to culture, economically important or of public concern.

Little attention has been given the response of benthic

macroinvertebrates, other than insects, to pollutants.

Benthic invertebrates are more appropriate test organisms

for flowing waters than are zooplankton because the

latter are not typically found in such systems, and are

therefore not good indicators of the impact of pollutants

on lotic invertebrates. However, toxicity data for

benthic organisms, including aquatic oligochaetes,

turbellarians, pelecypods and gastropods, are still

extremely limited, and only one toxicity test protocol

for these organisms exists, i.e., for the exotic clam

Corbicula fluminea (Foster 1979). Non-insect benthos

have been considered either unimportant or their

responses have been extrapolated from those of the common

test organisms. However, use of zooplankton or fish as

surrogates for non-insect benthic fauna is questionable.

Not only are conditions at the water-sediment interface

different than those in the open water and difficult to

assess with pelagic organisms, there ought to be specific

information on the response of benthic organisms to

pollutants since they represent distinctly different taxa

(Buikema et al. 1982). One of the most widely

distributed groups of macrobenthos native to streams in

Florida is the unionid mussels. Little is known about

their sensitivities to various pollutants entering their


Distribution and Life History of Unionid Mussels

The vast majority (36 genera and 250 species) of

bivalve mollusc species in North American continental

waters belong to the family Unionidae (Burch 1973). The

group as a whole is endemic to North America, but many

species have limited ranges. Unionid mussels generally

prefer lotic habitats with stable substrates and some

silt. They are distributed from southern Ontario to

Florida and west to Washington and Oregon. However, the

best studied and perhaps richest mussel (or clam) fauna

is found in the eastern United States between the

Appalachian Mountains and Mississippi River.

Several unique life history features were key to the

evolutionary development of freshwater mussels from their

marine ancestors (Stein 1971). These included the

production of a parasitic glochidia larva rather than the

free-living veliger, incubation of the larvae in the

marsupia (gills) of the female and the requirement for a

fish host during the 9-30 day parasitic phase.

Reproduction begins when male mussels shed sperm into the

water. Sperm cells are drawn into the incurrent siphon

of the female mussel and become lodged in the gills on

each side of her body that are specifically modified for

incubation. Development to the bivalved glochidia

occurs inside the female. During this time, the larvae

may become infected by any number of bacterial, fungal,

protozoan or water mite species, which may reside in the

mussel permanently or temporarily.

When mature, glochidia are shed into the water via

the excurrent siphon either directly onto fish hosts

whose presence stimulates release or randomly broadcast

into water currents (Buchanan 1980, Parmalee 1967).

Complete development into juveniles requires a period of

parasitism on fish during which the organ systems

develop. Distribution of unionid mussels is facilitated

by their attachment to mobile hosts instead of by the

production of mobile larvae as in marine bivalves (Fuller

1974). After encystment periods of varying times,

glochidia drop off of their hosts and become free-living

filter feeders (Arey 1932).

Habitat Destruction and Faunal Decline

Mussel fishing was a thriving business in the

Illinois, Tennessee and Mississippi Rivers from the late

1800's to the mid-1960's (Isom 1969, van der Schalie and

van der Schalie 1950, Starrett 1971). They were a source

of freshwater pearls for jewelry, and their shells

supplied the button industry with raw materials. Later,

shell slugs were used as seeds in the Japanese cultured

pearl industry (Parmalee 1967).

Ten thousand tons of mussel shells per year were

harvested from the Tennessee River in the 1940's and

1950's, but harvests declined steadily during ensuing

years (Starrett 1971, Parmalee 1967). Similar changes in

abundance were observed in other eastern rivers during

the same period (van der Schalie and van der Schalie

1971, Isom 1969, Starrett 1971). Several factors have

been suggested as causes for the decline including

overharvesting, habitat destruction by damming and


Overharvesting of mussels in the Tennessee, Illinois

and Mississippi Rivers may have contributed significantly

to the decreased abundance of some species (Starrett

1971, Forbes and Richardson 1919, Danglade 1914). In

1910, there were over 2,600 boats engaged in mussel

fishing along the lower half of the Illinois River alone

(Starrett 1971). Similar intensive harvests were made

by the well-developed mussel industry of the Tennessee

and Mississippi Rivers. Not only did such harvesting

reduce populations directly, but it also may have reduced

breeding stock below the replacement capacity of

remaining stock, destroyed stream habitat and resulted in

the death of disturbed but uncollected animals (Fuller

1974). While the button industry switched from pearl to

plastic in the 1930's and 1940's, another use for mussel

shells was found. Spheres of mussel nacre (pearlized

shell) were used as nuclei by the Japanese cultured pearl

industry beginning in the late 1950's. The use of SCUBA

gear permitted the harvest of whole beds of mussels

leading to localized extinction (Fuller 1974).

A second major contributor to the declining mussel

fauna was the extensive habitat destruction resulting

from damming activities of the Tennessee Valley Authority

(TVA) beginning in the 1930's (Isom 1969). While adult

mussels of some species prefer quiet water (Wilson and

Clark 1912, Danglade 1914), juveniles and adults of many

species need riffle water. Thus, damming may have

provided more habitat for some species, but reduced the

area suitable for others. Decreased water flow also

hinders reproduction by limiting dispersal of sperm and

later, glochidia. Further, impoundment changes fish

species distributions which may affect mussel recruitment

since they are briefly parasitic on fish (Fuller 1974).

Other consequences of damming include increased siltation

which can lead to suffocation of mussels, as well as

general loss of habitat and decreased recruitment success

due to release of tailwaters into otherwise suitable

stream reaches. The latter results from both low

temperatures and low oxygen levels of tailwaters (Fuller

1974, Marking and Bills 1980, Ellis 1936).

Finally, the effects of water pollution by human

waste, industry and agricultural activities have damaged

the mussel fauna in many areas. Pulp and paper mills

which release sawdust and process effluents destroyed

mussel populations in Minnesota (Danglade 1974), the

upper Tennessee River drainage (Ortman 1918), panhandle

Florida (Heard 1970) and in areas around Ottawa, Canada

(Mackie and Qadi 1973). As noted earlier, siltation is a

problem for mussels and with its increase along with

agricultural activities, the molluscs declined steadily.

The effects on mussel populations of pesticides and

herbicides used in farming and aquatic weed control, and

metals released in acid mine drainage and industrial

effluents have only recently been examined. Little

conclusive evidence is available since the specific

sensitivity of mussels to such pollutants is difficult to

determine from field data, and laboratory exposures have

been limited to a few studies of adults (Imlay 1971,

Imlay 1974, Foster and Bates 1978). Development of

better culture techniques and toxicity test procedures

have begun to make experimental work possible.

Propagation Of Freshwater Mussels In Artificial Media

The unusual mode of reproduction of unionid molluscs

makes their culture in the lab more difficult than it is

for other molluscs that have free-living veliger larvae.

The life cycle of unionid mussels includes a parasitic

larva (glochidia) which normally attaches to fish gills

or fins during early development. This stage must have

fish to parasitize or a culture medium that would provide

the necessary nutrients.

Earliest efforts to propagate freshwater mussels

(LeFevre and Curtis 1912) in fish plasma were

unsuccessful. Glochidia did not transform. By 1926,

transformation of glochidia to juveniles had been

accomplished with the use of an artificial medium (Ellis

and Ellis 1926). However, glochidia were permitted to

encyst on fish gills for 18-96 hours before being

dissected out to incubate through transformation. The

contents of their growth medium were described as

including NaC1, KC1, CaC12, NaHCO3, dextrose, a mixture

of amino acids, small quantities of phosphates and traces

of magnesium salts (Ellis and Ellis 1926).

Research into mussel propagation lost its impetus as

the button and mussel-fishing industries dwindled in the

1940's and 1950's. However, in the hope of replenishing

declining natural populations, the Tennessee Valley

Authority funded research to develop methods for in vitro

propagation of these freshwater molluscs in the early

1980's (Isom and Hudson 1982, Hudson and Isom 1984). The

goal was to develop a complete culture medium that

eliminated the need for fish hosts during the larval

stage and to produce a large number of juvenile mussels

at one time. As a result, a culture medium containing a

Ringers solution, vitamins, glucose, amino acids,

antibiotics and fish plasma was developed as a substitute

for live fish, (Isom and Hudson 1982). Glochidia were

removed from the gills of ripe female mussels, rinsed

several times in sterile water and put in the medium.

Culture dishes were then placed in a temperature

controlled CO2 incubator (230 + 30 C). The

transformation of glochidia to juveniles takes 9-30 days

(230 + 30 C) depending on the species, culture

temperature and degree of glochidia maturity at the start

of incubation. While the Hudson and Isom method (1982,

1984) is far better than that of Ellis and Ellis (1926)

which relied on the use of fish hosts for encystment of

glochidia during transformation, further simplification

is desirable. The old method (Hudson and Isom 1982,

1984) is laborious, still requires the use of fish plasma

which may not be readily available nor of consistent

quality, and a CO2 incubator. A simplified method for

mussel culture is necessary before they can be available

in the numbers needed for replenishment of declining wild

stocks or for other purposes, e.g. toxicity tests.

Sensitivity of Unionid Molluscs to Environmental

Interest in the effects of toxic pollutants on

mussels directly and their use as environmental

indicators in general has increased following the

decrease in population sizes. Freshwater mussels have

been suggested for use as biological monitors in lotic

environments for many years. Biomonitors are important

because analysis of water does not reflect biologically

available concentrations of toxicants (Leard et al.

1980). Sometimes biota can be affected by concentrations

below the detection limit of analytical instruments, and

at other times, high ambient concentrations are benign

because they are refractory or are adsorbed to

particulate matter. The utility of freshwater mussels as

biomonitors is enhanced by their sedentary lifestyle,

their long lifespan compared to other invertebrate

species and the fact that they live in the sediments

while being filter-feeders. Thus, mussels are exposed to

dissolved, particulate and sediment-sorbed contaminants

(Havlik and Marking 1987). However, current information

on mussel sensitivities consists largely of species

presence-absence data for locations impacted by toxics,

and measurements of contaminants in shells or tissues

(Havlik and Marking 1987). Experimental data on specific

sensitivities are very limited.

Simmons and Reed (1973) used the presence and

abundance of freshwater molluscs as indicators of

biological recovery in the North Anna River, Virginia.

The North Anna River was receiving acid mine drainage

from a defunct coal mine. While aquatic insects re-

established quickly below the confluence of the river and

an unpolluted creek, molluscan species were absent for

another 50 miles downstream. In this example, molluscs

were more sensitive indicators of biological recovery

than were insects, traditionally regarded as good

biomonitors. Simmons and Reed (1973), however, suggested

that the lack of mussel fauna in the acidified river

reach may have been caused by siltation and loss of host

fish for the glochidia.

Three species of mussels from the Illinois River

near Peoria, Fusconaia flava, Amblema plicata and

Quadrula quadrula, were analyzed for metals along with

fish, tubificid worms, river sediment and water (Mathis

and Cummings 1973). The portion of the river studied was

highly industrialized and therefore received metallic

effluents. The goal was to determine if there had been a

loss of species and if the loss could be related to metal

concentrations. Mussels accumulated the metals to levels

exceeding dissolved concentrations by 1-2 orders of

magnitude, but had lower levels than those in the

sediments. No attempt was made to determine species

abundance, but the presence of a thriving mussel-fishing

industry within the study area indicated that these three

mussels were plentiful. Another species, Musculium

transversum, the fingernail clam, was absent from

polluted areas of the river where it had been common

before 1954. While concentrations of metals in the

tested adult clams may not have been lethal, this does

not mean there was no impact on juveniles or on other

species such as M. transversum which were absent.

Anderson (1977) analyzed shells and tissues of

freshwater clams from the Fox River, Wisconsin for

cadmium, copper, lead and zinc. He found that body

burdens generally paralleled sediment concentrations

while being much higher than were found in water. Metal

concentrations were much lower in the shells

latter, samples of gills had the highest levels of


In one of the most important studies of mussel

responses to dissolved metal pollution, Foster and Bates

(1978) compared lethal concentrations of copper in

Quadrula quadrula with state mandated water quality

criteria. Standing crops of mussels in the Muskingum

River during a period prior to increased discharge of the

metal effluent (1967-1970) were compared to a post-

increase period (1972-1973). An 86% decrease in mussel

standing crop occurred between the earlier and later

study at a point 5 Km downstream from the effluent

outfall. Body burdens of copper increased approximately

10-fold during 11-day laboratory exposures of Q. quadrula

to the effluent. Mussels placed in in situ cages

exhibited an even greater ability to concentrate copper.

Their mean body burdens of copper, 20.64 ug Cu-/g after

14 days in the cages, were similar to that found after 11

day exposures to whole effluent in the laboratory.

However, the toxic effluent comprised only 0.004% of the

mean daily river flow at the point where the cages were


Juvenile mussels accumulated copper at a greater

rate than adults. This finding was significant because

it may be the result of increased metabolic rate

associated with the sexual maturation process (Foster and

Bates 1978). In that case, the reproductive capacity of

the species may be compromised by the loss of juveniles,

or by decreased recruitment success. Foster and Bates

(1978) also emphasized that in developing water quality

criteria, consideration should be given to the impact on

a greater variety of stream fauna, including molluscs.

Metals such as manganese, zinc, cadmium, copper and

lead may accumulate in mussels to levels higher than

ambient concentrations depending on sediment conditions.

Radioactive manganese (Mn54) was concentrated at rates

ranging from 11,000-40,000 times water and sediment

levels in soft tissue and about 14,000 times in shells of

Unio mancus var. elongatulus (Gaglione and Ravera 1964,

Ravera 1964). These values were three times higher than

in Anodonta cyqnea from the same area. Levels were

highest in U. mancus gills and lowest in the visceral

sac. Mn54 was present at undetectable levels in water,

sediment and other organisms. Leatherland and Burton

(1974) measured cadmium levels in Anodonta cygnea in the

Thames River, England. The mussels had concentrated Cd+2

from water with 0.49 mg/L to a tissue level of 9 mg/L.

These studies demonstrate the validity of using molluscs

as biomonitors for metals.

Reddy and Chari (1985) found increased production of

enzymes involved in amino acid synthesis in the

freshwater mussel Parreysia rugosa after exposure to

mercury and copper. They attributed this increase to the

higher demand for amino acids in metabolic processes and

energy transfer in the stressed mussels.

One of the factors that can hinder the usefulness of

molluscs as biomonitors is variability in accumulation

rates of metals depending on tissue, age, sex or other

unknown characteristics (Jones and Walker 1979, Bryan

1973, Ayling 1974). Selection of samples can obviate

some of the variability but requires adequate field time

and preliminary studies. However, in some cases, no

explanation for the variability has been found so no

blocking of samples or analyses can be used to overcome

the problem. This may reduce the apparent responses of

mussels to changes in ambient conditions and thus obscure

determination of real responses (Jones and Walker 1979).

Using fractionation procedures that isolated trace

metals from sediments, Tessier et al. (1984) determined

that the accumulation of metals in tissues of Elliptio

complanata was most strongly related to individual

fractions rather than to total metals. Samples were

extracted sequentially at decreasing pH with various

acids to isolate metals bound to Fe-Mn oxides,

carbonates, organic matter and those that were dissolved

or residual. The mantle and gills of the molluscs

contained the highest levels of metal (Cu, Zn, Mn, Fe)

while the foot and adductor muscles had the lowest. This

is a typical pattern observed in molluscs (Hobden 1970,

Gaglione and Ravera 1964, Anderson 1977). The

relationship between sediment levels and body burden

depended on the metal species but were predictable from

appropriate regression equations.

Pace and DiGuilio (1987) assayed the lead content of

peat, sediment and clams (Rangia cuneata) from the Pungo

River estuary in North Carolina. Using fractionation

techniques similar to those described in Tessier et al.

(1984), they determined that lead levels in the clam

tissues were very low (0.2-0.5 ug/g) compared to the peat

(12.8 ug/g) reflecting the presence of lead in non-

bioavailable forms. An analysis of heavy metals in three

estuarine molluscs from the Spanish coast (Lopez-Artiguez

et al. 1989) showed that the concentration of particular

metals by molluscs is species-dependent. Oysters

(Crassostrea angulata) accumulated very high levels of

copper (180.45 ug/g) compared to Tapes decussatus and

Cardium edule, while the latter had generally higher

levels of As, Hg and Sn than did the other two species.

Growth rates of mussels, as determined by changes in

shell growth rings, were lower in rivers polluted with

heavy metals such as silver, cadmium, iron, mercury and

manganese (Imlay 1982) was determined by changes in shell

growth rings.

A few laboratory studies have measured the response

of mussels to specific metals. K+ was toxic to four

species of unionid clams at levels below those found in

some rivers in the United States (Imlay 1974). The LC50

for Lampsilis radiata siliquoidea and Fusconaia flava

was 15 mg/L after 36 hours of exposure. Amblema plicata

was more sensitive, having an LC50 of 15 mg/L at 26 days

7, while 50% of the Actinonaias carinata died in 11 mg/L

K+ in eight days (Imlay 1971). Labos and Salanki (1964)

recorded "abnormal" glochidial activity for Anodonta

cygnea in 3.91 mg/L K Imlay (1971) used such data to

successfully predict mussel distributions in rivers based

on ambient K+ levels (Imlay 1974). Anodonta cygnea was

found to be more sensitive to K+ than to any other cation

in work by Lukacsovics and Salanki (1964). Since K+ is

a common effluent of paper mills, irrigation return water

and petroleum brine, this sensitivity may have

significance to the survival of mussel populations.

Imlay (1971) showed that mussels were as sensitive to

dissolved metals as other invertebrates and fish

following laboratory exposure for several months. In

particular, copper was lethal to the mussels at 25 ug/L,

a value similar to that recorded for fish and non-

molluscan invertebrates.

In a field study of the effects of dissolved

aluminum on Anodonta grandis grandis, mussels suffered no

mortality and only transitory changes in blood ion

composition (Malley et al. 1988). When the mussels were

placed in an acidified lake to which alum was added,

there was no change in Na K or SO2 concentrations, a

decline in Mg+2 and a slight increase in blood Cl.

Increases in Ca+2 levels were attributed to compensatory

mechanisms for maintenance of blood pH and were seen as

more detrimental to mussels during chronic exposures to

low pH than to aluminum.

As with metal toxicity data, information on mussel

bioaccumulation or sensitivity to pesticides is largely

circumstantial with few experimental data being reported.

Bedford et al. (1968) introduced specimens of two mussel

species into the Red Cedar River, Michigan to determine

whether they could be used to detect pesticides at low

dissolved concentrations. Lampsilis siliquoidea and

Anodonta grandis were found to concentrate DDT and its

metabolites, methoxychlor and aldrin to levels many times

greater than were present in the dissolved or particulate

fractions. However, the mussels contained lower levels

of pesticides than were detected in the sediments. These

results indicated the feasibility of using freshwater

mussels for detecting the presence of pesticides in

running water. Leard et al. (1980) found the

insecticides parathion, DDT, chlordane, toxaphene and

their metabolites in seven mussel species from streams

draining agricultural areas where these pesticides were

in use. Pesticide accumulation varied with species, but

there was a decrease in DDT body burden during the period

after DDT use was limited. During that time, an increase

in toxaphene and parathion levels was measured in the

clams. This was a reflection of their increased use in

place of DDT (Leard et al. 1980). Sphaerium corneum

concentrated dieldrin to 1000 times ambient levels in

laboratory experiments and field studies (Boryslawskyj et

al. 1987). Diazinon and parathion are also taken up by

mussels at high rates (Miller et al. 1966). None of

these studies gave data on lethality of the compounds to


Laboratory toxicity experiments determined the 96-h

LC50 of the lampricide 3-trifluoromethyl-4-nitrophenol

(TFM) for Liqumia sp. to be 8.3 mg/L for small

individuals (< 9 cm long) and 11.7 mg/L for larger

specimens (> 16 cm). This is 1.5-4 times the normal

stream water concentration in treated areas and similar

to the sensitivities of the crayfish Orconectes (17.8

mg/L) and Gammarus pseudolimnaeus (22.3 mg/L) (Johnson

and Finley 1980). The lampricide Bayer 73 was lethal to

50 per cent of adult Elliptio dilatatus at 382 ug/L

concentration (Rye and King 1976). The LC50 for rotenone

was 2.7 mg/L in soft water for adult Lampsilis sp.

(Farringer 1972).

The response of adductor muscle activity in

glochidia of Anodonta cygnea was used as a measure of

toxicity by Varanka (1977, 1978). Spontaneous adductor

muscle activity is crucial if glochidia are to attach to

their hosts. In a series of experiments, Varanka applied

the muscle-contraction inducer tryptamine to glochidia

and recorded baseline contraction rates. He then exposed

larvae to tryptamine + pesticide to determine the effect

of the pesticide on this activity. Results of 30-minute

exposures to malathion, 2,4-D and Shell-DD indicate that

the EC50s (effective concentration for 50% of the sample)

based on adductor muscle activity were considerably

higher than the usual environmental levels of these

pesticides. However, the fact that muscle activity

responded to such short exposures suggests that further

study with longer exposure times may be worthwhile.

Though in these and other studies the mussel

bioaccumulation capacity for metals and pesticides is

great, we have very few measures of the lethal or sub-

lethal effects of such exposures. With such a limited

database, it is premature to draw conclusions about the

sensitivity of mussels to environmental exposures to

metals or pesticides. The majority of studies to date

consist of species surveys and measurements of tissue or

shell toxicant levels. How the presence of these

contaminants may affect mussel survival, growth or

reproduction remains to be determined. Laboratory

exposures of mussels of various species and age groups to

pesticides, metals or organic pollutants would provide

extremely valuable measures of direct effects. Such data

would allow us to separate the effects of habitat

destruction, siltation and competition with Corbicula sp.

from response to pollutants.



The United States Environmental Protection Agency

(EPA) has recently advocated the use of short-term

chronic toxicity tests for biological monitoring of water

and wastewater (Horning and Weber 1985). Such tests were

developed by EPA on the basis of their cost-

effectiveness, rapidity and requirement for low sample

volumes over the course of the test. These features

allow the methods to be implemented in on-site effluent

toxicity evaluations, as well as in toxicity assessments

of pure compounds and samples shipped to central

laboratories. Implementation of short-term chronic

toxicity methods was justified by evidence that these

early life cycle tests reasonably approximated more

complete, full life cycle chronic toxicity tests that had

been the standard for many years (Norberg and Mount


One of these tests employs fathead minnow larvae

(Pimephales promelas) in a seven-day, static renewal,

survival and growth test. Test results are based on the

survival and growth (weight gain) of larval fathead

minnows over seven days in the presence of a range of

toxicant concentrations. It is a new evaluative method

for which few test results are available. Norberg and

Mount (1985) determined the chronic toxicity of several

industrial effluents and receiving waters, as well as

zinc, copper and Dursban during their test development

work. The 7-d test with fathead minnow larvae gave

results similar to those from much longer (3-6 months)

early life stage (ELS) tests.

The current study was designed to evaluate the

toxicity of the herbicide Hydrothol-191 to fathead

minnows. Hydrothol (endothall) acts to decrease

photosynthesis and cellular respiration in turfgrass and

to decrease the production of amylase in germinating

barley seeds (Ashton and Crafts 1981). It has a half-

life of 10 days, biodegradation by bacteria being the

primary fate process leading to the decline of Hydrothol

concentrations (Reinert and Rodgers 1987). Its effects,

similar to those of Actinomycin D, are not reversible

with benzyladenine (Penner and Ashton 1968). Since

actinomycin D selectively inhibits the synthesis of m-

RNA, the translator of DNA messages during protein

(enzymes) production, it is hypothesized that Hydrothol

also interferes with m-RNA production (Ashton and Crafts


Support for this assertion was given by the effect

of Aquathol-K (dipotassium endothall, Pennwalt Corp.) on

the smoltification of juvenile Chinook salmon

(Oncorhynchus tshawytscha). When juvenile salmon were

exposed to water with endothall levels as low as 3 mg/L

for four or fourteen days prior to their transfer from

freshwater to artificial seawater, they were unable to

survive (Liguori et al. 1983). Fish that were allowed to

recover in freshwater for 10 days prior to their transfer

to seawater survived well. Histopathologic analyses

indicated that hypertrophy of branchial epithelium

occurred in fish exposed to 10 mg/L or more of endothall.

The process of smoltification involves changes in

plasma levels of thyroxine, triiodothryonine, and gill

ATPase activity. Both triiodothryonine and ATPase are

found in gill tissues. Liguori et al. (1983) suggested

that in endothall damaged gill tissue, the levels of

triiodothryonine and ATPase may be depressed, although

neither was measured in their study. Since Hydrothol

interferes with smoltification possibly related to levels

of ATPAse, it may be the result of its inhibitory effects

on m-RNA synthesis (Liguori et al. 1983). Therefore, the

impact of Hydrothol on organisms should be temperature

dependent as are most chemical reactions (Wilson 1972).

In Florida, Hydrothol is registered for use in the

control of algae, Hydrilla verticillata, Myriophyllum

spicatum, Utricularia spp., Valisneria spp. and other

submersed macrophytes, as well as for several

agricultural purposes (Dupes and Mahler 1982, Blackburn

and Weldon 1963). Although there are some data on the

acute toxicity of Hydrothol to non-target aquatic

organisms, only limited data are available on its

chronic toxicity to aquatic biota. The Florida

Department of Environmental Regulation (FDER), seeing the

likelihood of increased Hydrothol use because of its

efficacy in the long-term control of aquatic macrophytes,

requested the evaluation of its potential impact on non-

target organisms.

FDER was also interested in knowing what if any

effect water temperature might have on the toxicity of

Hydrothol to freshwater fish. There were almost no data

indicating the effect of temperature on Hydrothol

toxicity (Walker 1963). Since many physiological

processes are affected by temperature because of the

pivotal role of enzymes in cellular respiration and the

synthesis or degradation of organic compounds (Wilson

1972), it was important to determine whether water

temperature would alter the toxicity of Hydrothol to


The goals of this investigation were: (1) to

determine the chronic toxicity of Hydrothol to the

fathead minnow and (2) evaluate the impact of

temperature on its toxicity. These data would be useful

in the development of better guidelines for the use of

this herbicide in Florida.

Materials and Methods

Test Organism

Fathead minnows (Pimephales promelas) were acquired

from the EPA-Newton Laboratory in Cincinnati, Ohio. Fish

embryos were sent in insulated containers by express mail

and hatched in transit. Newly hatched fathead minnow

larvae preferably less than 24 hours old were used to

initiate a test.

Test Organism Food

Fathead minnow larvae were fed live brine shrimp

nauplii (Artemia salina) raised from eggs in the

laboratory. Brine shrimp nauplii were incubated at 250 C

and harvested when nauplii were less than 24 h old

(Peltier and Weber 1985, Horning and Weber 1985).

Fathead minnow larvae (10/500 ml test vessel) were fed

0.1 ml harvested brine shrimp (approximately 1000

nauplii) three times daily at 4-hour intervals.

Dilution Water

Moderately hard reconstituted freshwater was used

as diluent throughout the test. It was prepared by

adding the following constituents to 1 1 of deionized

water: 96 mg NaHCO3, 60 mg CaSO4 2H20, 60 mg MgSO4 and

8.0 mg KC1. This produced water with a pH of 7.4-7.8, a

hardness of 80-100 mg/L CaCO3 and an alkalinity of 60-70

mg/L CaCO3 (Horning and Weber 1986). Dilution water was

made in bulk and stored in a plastic carboy at 150 or 250

C for the duration of each seven day test.

Test Chemical

The toxicity of Hydrothol-191 (Pennwalt Corp.,

Philadelphia, PA), an agricultural and aquatic herbicide

was evaluated in this study. Hydrothol, the trade name

for the alkylamine salt formulation of endothall (7-

oxabicyclo [2,2,1] heptane-2,3-dicarboxylic acid), is

used extensively to control Hydrilla verticillata and

Myriophyllum spicatum. Test solutions were made fresh

daily by dilution of a stock solution of Hydrothol with

moderately hard reconstituted freshwater (v/v). All

Hydrothol concentrations given are nominal.

Reference Toxicant

Cadmium chloride, obtained from EPA (Quality

Assurance Branch, EMSL, United States Environmental

Protection Agency, Cincinnati, OH 45268) was the

reference toxicant for fathead minnow tests. Three tests

were performed with CdC12 to evaluate the consistency of

test results. Later, CdC12 was used as a check on test

organism quality in each definitive test with Hydrothol.

Range Finding Test

In order to obtain the appropriate range of

Hydrothol concentrations to be used in 7-day tests an

acute toxicity range finding test was conducted. It was

determined that 0% larval mortality occurred at 100 ug/L

and that 100% mortality occurred at 1000 ug/L.

Consequently, the definitive 7-day Hydrothol

concentration range was 50-1000 ug/L.

Seven-day Survival and Growth Toxicity Test

Toxicity tests were initiated by placing larvae in

1 liter borosilicate beakers (test chambers) containing

500 ml control or test water. Larvae were transferred

into duplicate test chambers by a large-bore pipet until

each test chamber contained 10 larvae, for a total of 20

larvae at each Hydrothol test concentration.

Definitive tests were conducted at 150 C and 250 C

in constant temperature rooms, with a photoperiod of 16

hours light and 8 hours of darkness. The test chambers

were randomized at the beginning of the test. Ninety per

cent of the test solution was renewed every day after a

large-bore pipet was used to siphon dead brine shrimp and

other debris from the bottom of the test chambers.

Chemical and physical analyses of test water were

conducted according to standard EPA methods (EPA 1983).

Dissolved oxygen, temperature, pH, conductivity,

alkalinity and hardness were measured at the beginning of

each 24 hour exposure at all test concentrations and in

the control.

The numbers of live and dead larvae in each test

chamber were recorded daily, and the dead larvae were

removed. After seven days of exposure the test was

terminated. The surviving larvae were removed and

preserved as a group in 4% formalin. At a later date,

the groups of preserved larvae were rinsed in distilled

water and dried at 1050 C for a minimum of 2 hours. Dry

weights of each group of larvae were measured to the

nearest 0.001 g.

Statistical Analysis

All LC50 and 95% confidence intervals were

calculated using the TOX-Dat multimethod computer program

(Peltier and Weber 1985, Horning and Weber 1985).

Survival data were arcsine-transformed and analyzed by

Dunnett's Procedure which includes an analysis of

variance (ANOVA), followed by a comparison of each

toxicant concentration with the control. From this

analysis a No Observed Effect Concentration (NOEC) and a

Lowest Observable Effect Concentration (LOEC) were

calculated. In addition, the Chronic Value (ChV) was

determined by calculating the geometric mean of the NOEC

and LOEC.

Growth (dry weight) data were also analyzed by

Dunnett's Procedure (SAS 1986). The average dry weight

of larvae from each replicate test chamber was entered

into the program. The results of the analysis of

variance (ANOVA) and regression analysis were used to

determine statistically significant effects of the

various concentrations of Hydrothol and temperature (150

C and 250 C) on larval growth and survival.


Dilution Water Quality

Water quality (Table 3-1) during the 7-day tests

with fathead minnow larvae was within the range expected

for moderately hard reconstituted freshwater (Horning

and Weber 1986). Test temperature, regulated in a

constant temperature room, was maintained at 150 or 250 C

( 10 C).

Reference Toxicant Tests

Results of reference toxicity tests with CdCl2

indicated that methodology was of acceptable quality and

that larvae used in the tests were healthy (Table 3-2).

No trends of increasing or decreasing sensitivity of test

organisms to Cd2+ were noted.

Survival and Growth of Fathead Minnow Larvae

Survival. Comparative data for survival at 150 and 250 C

indicated that Hydrothol toxicity was not changed by test

temperature (Tables 3-3 and 3-4). At 96-h, the LC50 was

Table 3-1. Range of water quality characteristics of
moderately hard reconstituted freshwater used in
replicate fathead minnow tests at 150 and 250 C.

Test pH D.O. Alkalinity Hardness
(mg/L) (as mg/L CaCO3)

150 C

1 7.8-8.0 9.0-9.1 52-56 66-68
2 7.6-7.8 9.1-9.4 44-62 72-92
3 7.8-7.9 8.5-9.1 59-62 87-90

250 C

1 7.7-8.0 6.9-7.1 66-69 80-86
2 7.8-7.9 7.1-7.3 62-63 77-79
3 7.7-7.8 7.2-7.4 58-61 73-82

Table 3-2. Results of reference toxicant tests with
CdCl2 for fathead minnow larvae.

Test No. LC50 95% Confidence Interval
(ug/L as Cd2+)

1 30.2 25.2-37.9

2 20.9 17.7-25.2

3 14.0 10.3-18.8

393 + 198 ug/L at 150 C and 468 + 44.4 ug/L at 250 C.

The values were not different (p<0.05). The 7-d LC50s

were lower at both temperatures than after four days, but

again, the effect of temperature itself was not

significant. The 7- LC50 was 233 + 57.3 ug/L at 150 C

and 304 + 46.4 ug/L at 250 C.

It may be suggested that the temperature increase

from 150 C to 250 C was not sufficient to produce a

change in toxicity. However, chemical reactions

generally double with a 100 C increase in temperature

(Wilson 1972). Furthermore, a measurable increase in the

toxicity of the inorganic salt of endothall (Aquathol-K)

was seen for five species of fish tested by Johnson and

Finley (1980). Although the test species was not given,

Aquathol toxicity increased approximately fourfold in 96-

h tests at 220 C vs 70 C. The LC50 at 70 C was 1740

mg/L, while at 220 C it was only 323 mg/L. In another

study, Walker (1963) found a 13% to 43% increase in the

toxicity of Hydrothol to bluegill sunfish, redear

sunfish, largemouth bass and yellow bullhead with only a

50 C increase in test temperature. Since endothall

compounds are contact type membrane-active herbicide and

affect protein synthesis (Ashton and Crafts 1981), they

should be more biologically active at elevated


The fact that the toxicity of Hydrothol to fathead

minnows did not change with a 100 C increase in

Table 3-3. Ninety-six hour and 7-day LC50s for
larval fathead minnow survival at 150 C.

Test No. 96-hour (95% C.I.) 7-day (95% C.I.)
LC50 LC50
(ug/L) (ug/L)

1 310 (132-530) 203 (164-240)

2 250 (189-32)7 197 (155-253)

3 619 (265-1060) 299 (237-382)

MEAN 393 233
( SD) ( 198) (+ 57.2)

Table 3-4. Ninety-six hour and 7-day LC 50's for
larval fathead minnow survival at 250 C.

Test No. 96-hour (95% C.I.) 7-day (95% C.I.)
LC50 LC50
(ug/L) (ug/L)

1 519 (441-634) 251 (132-530)

2 447 (389-523) 324 (256-400)

3 438 (367-531) 337 (267-417)

MEAN 468 304
(+ SD) ( 44.4) ( 46.4)

temperature may be attributed to species differences.

Interspecific responses to pollutants can differ

remarkably even at a set temperature (Johnson and Finley

1980, Mayer and Ellersieck 1986).

Growth. There was no statistically significant

weight gain in controls at 150 C compared to larvae in

any of the Hydrothol concentrations (Table 3-5). Since

fathead minnows spawn at temperatures above 170 C

(Peltier and Mount 1985), their larvae may simply not be

adapted to grow at temperatures lower than 170 C.

Therefore, at 150 C survival rather than growth data were

used to determine the chronic impact of Hydrothol on fish


Survival was significantly inhibited at Hydrothol

concentrations 265 ug/L (the lowest observed effect

concentration) at 150 C, while no effects were seen at

concentrations I132 ug/L (the no observed effect

concentration) (Table 3-6). Based on these data, the

chronic value (ChV) for survival was calculated as 186

ug/L. The ChV is the geometric mean of the NOEC and LOEC

and is equivalent to the maximum allowable toxicant

concentration (MATC) used by regulatory agencies.

Growth was a more sensitive measure of toxicity at 250 C

than was survival (Table 3-7). Growth was significantly

lower (p 0.05) at concentrations 132 ug/L while

survival was not significantly affected until Hydrothol

concentrations were 2 265 ug/L. Using an LOEC of 265

Table 3-5. Seven-day growth and survival of fathead
minnow larvae exposed to Hydrothol at 150

Hydrothol Mean Mean
percent concentration weight ( SD)
survival (ug/L) (mg)

0.107 (+ 0.024)

0.106 (+ 0.023)

0.087 (+ 0.012)

0.109 (+ 0.038)









SMean values + standard deviation (SD) from three
independent seven-day tests.
b Significantly different from control at the 0.05 level.
c Fish all died before seven days, no weight determined.

Table 3-6. Seven-day growth and survival of fathead
minnow larvae exposed to Hydrothol at 250 C.

Hydrothol Mean Mean percent
concentration weight ( S.D.) survival
(ug/L) (mg)

Control 0.420 (+ 0.074) 100

50 0.348 (+ 0.049) 90

132 0.300b ( 0.016) 100

200 0.318b (+ 0.059) 90

265 0.198b ( 0.084) 70b

530 0.255b (+ 0.078) 5b

1060 c Ob

SMean values + standard deviation (SD) from three
independent seven-day tests.
b Significantly different from control at the 0.05 level.
c Fish all died before seven days, no weight determined.

ug/L and an NOEC of 265 ug/L, the chronic value (ChV) for

survival of fathead minnow larvae in Hydrothol was

calculated to be 230 ug/L at 250 C. This means that

survival of fathead minnow larvae is threatened by long-

term exposures to Hydrothol concentrations that exceed

230 ug/L. The ChV for growth was 81 ug/L. At

concentrations higher than 81 ug/L, growth impairment

could result from chronic exposure to Hydrothol.

The relationship between growth and temperature was

examined further with regression analysis (Table 3-8).

While there was no significant (p-0.05) relationship

between the concentration of Hydrothol and growth at 150

C (R2=0.0002), there was a significant relationship

(R2=0.3940) at 250 C. In addition, a two-way ANOVA (with

factor interaction) indicated that when 150 and 250 C

data were combined, temperature was more important than

Hydrothol concentrations in determining larval weight

gain (p<0.05). Once again, temperature was an important

factor in determining the chronic effect of Hydrothol on

growth of fathead minnow larvae, even though survival was

not significantly affected by temperature. It has been

proposed that survival may be a satisfactory endpoint for

chronic toxicity tests because in many cases it has

proven to be an adequate indicator of long-term impacts

(Mayer et al. 1986). However, at 250 C fathead minnow

larval survival was substantially less sensitive than

growth as an endpoint measurement.

Table 3-7. No effect, lowest effect, and
chronic value concentrations (ug/L) of Hydrothol
fathead minnow larvae at 150 and 250 C.


Parameter Survival Growth Survival Growth
150 C 150 C 250 C 25 C

No Effect 132 200 50

Lowest Effect 265 -265 132

Chronic Value 186 -230 81

Table 3-8. Regression analysis of the relationship
between Hydrothol concentration (X) and growth (Y) at 150
and 250 C.

150 C

250 C

Y = 0.0000032 X + 0.102

R2 = 0.0002a

F = 0.003

Pr > F = 0.9545

Y = -0.0004 X + 0.379

R2 = 0.3940b

F = 15.604

Pr > F = 0.0006

Not significant at p<0.05.

USignificant at p:0.05.


The results of several acute toxicity studies (Table

3-9) with fish species using Hydrothol are available for

comparison with LC50 values calculated from the current

study. Johnson and Finley (1980) determined the 96-h

LC50 of 0.75 mg/L for fathead minnows at 180 C. This

value is close to my 96-h LC50 for the same species.

Rainbow trout (Salmo qairdneri) and golden shiner

(Notemiqonus crysoleucas) were much less sensitive to

Hydrothol as evidenced by their four day LC50s of 1.7

mg/L and 1.6 mg/L, respectively (Mudge et al. 1986,

Finlayson 1980). Other workers reported a Hydrothol no

mortality range of from 3.0-55.0 mg/L (Holmberg and Lee

1976, Liguori et al. 1983, Berry 1984). Since Hydrothol-

191 is applied at a concentration of 1-5 ppmw (part per

million water) and has a half-life of 10 d (Blackburn et

al. 1971, Reinert et al. 1985), it poses a potential

threat to fish. Its concentration can remain above the

96-h LC50 for 10-20 days.

Only limited chronic toxicity data are available

from other studies with endothall products (e.g.

Hydrothol and Aquathol-K) (Liguori et al. 1983, Eller

1973). In a study of the impact of 10 mg/L Hydrothol on

juvenile Chinook salmon (Oncorhynchus tshawytscha),

Liguori et al. (1983) observed marked changes in

branchial tissues. Effects included epithelial

Table 3-9. Comparison of literature LC50 values for
Hydrothol to various freshwater fish.

Test Organism Stage or Temp 96-h Reference
Wet Wt. (C) LC50
(g) (mg/L)

Golden shinera 0.62 18 1.6 1

Rainbow trout 25 15 1.7 2

Rainbow trout 1.2 13 0.56 3

Rainbow trout 1.3 4

Cutthroat trout 1.0 10 0.18 4

Channel catfish 0.3 18 0.49 4

Bluegill 0.5 24 0.94 4

Bluegill 1.2 4

Fathead Minnow 0.6 18 0.75 3

Fathead minnowb larvae 25 0.39 5

Fathead minnowb larvae 15 0.47 5

IFinlayson (1980). 2Mudge et al. (1983). 3johnson and
Finley (1980). 4Pennwalt Corp. (1980). 5Current study.

hyperplasia and lamellar fusion. At lower exposure

levels (<10 mg/L), no histopathologic effects were

detected. The 14-d LC50 was 62.5 mg/L of endothall.

The most sensitive measure of Aquathol-K toxicity

was the seawater test (Liguori et al. 1983). In this

experiment, survival of juvenile Chinook salmon after

placement in seawater was measured following their

exposure to endothall concentrations of 10.1-105.7 mg/L

for 14 days. Transfer to seawater simulated the

migration of this species to the ocean during the

smoltification process, a critical stage in their

development. All fish died within three days of entry

into seawater. Even at sublethal concentrations

endothall exerts an effect that impairs important

physiological processes related to osmoregulation

(Liguori et L. 1983).

Eller (1973) followed histopathological changes in

bluegill exposed to Hydrothol-191 for up to 112 days. He

found significant but transitory changes in gill

epithelium in fish exposed to 0.3 mg/L of the herbicide.

Epithelial hyperplasia and lamellar fusion were noted in

bluegill during the first 14 days of exposure. After

that time, gill damage gradually reversed and gills were

normal by the end of the study. Some abnormalities were

noted in hepatic and testicular cells, but they were not

conclusively related to Hydrothol concentration (Eller


The limited data relating test temperature to the

toxicity of Hydrothol to fish comes from a study by

Walker (1963). He found that 96-h LC50s for bluegill

sunfish (0.05 mg/L), redear sunfish (0.10 mg/L),

largemouth bass (0.14 mg/L) and yellow bullhead (0.31

mg/L) at 23.80 C were reduced by 43%, 44%, 36% and 13%,

respectively, at 18.30 C. Reductions in Hydrothol

toxicity due to test temperature were greater for three

of these four species than was measured for the fathead

minnow in the current study. The only exception was the

yellow bullhead. Interspecific differences in

sensitivity to herbicides is seen throughout the

literature (Johnson and Finley 1980). The greater effect

of test temperature on Hydrothol toxicity to species

tested by Walker (1963) than for the fathead minnow may

be due simply to species differences. No other

explanations are readily apparent.

Hydrothol is relatively toxic to fish in comparison

with other herbicides such as 2,4-D, dichlobenil, diquat,

and PCP (Table 3-10) that may enter the aquatic

environment. 2,4-D is applied to ponds and lakes for

control of water hyacinth (Ag Consultant 1988) and to

agricultural fields for control of broadleaf weeds (Ware

1978). Its mode of action via a complex mixture of

effects on cell division and nucleic acid metabolism is

somewhat different than the impairment of m-RNA

production caused by Hydrothol (Ware 1978, Ashton and

Table 3-10. Summary of acute toxicity data for selected
herbicides that may enter the aquatic environment in

Herbicide and Animal Temp. 96-h LC50
test organism wt. (g) (C) (mg/L)

fathead minnow 0.5 17 18.0
bluegill 1.4 17 7.5

fathead minnow 0.8 18 6.0
bluegill 1.5 18 8.3

bluegill 1.3 12 245

fathead minnow 1.1 20 0.21
bluegill 0.4 15 0.03
channel catfish 0.3 20 0.07

bluegill 1.3 22 343
channel catfish 0.4 12 >150

fathead minnow 0.6 18 0.75
fathead minnowb 0.4 15 0.39
fathead minnowb 0.4 25 0.47

dData from Johnson and Finley (1980). "Data from the
current study.

Crafts 1981). The 96-h LC50 for 2,4-D was 18.0 mg/L for

fathead minnow and 7.5 mg/L for bluegill (Johnson and

Finley 1980). These levels are far below both the

application rates for aquatic environments or the

expected levels entering water from treated agricultural

areas (Ag Consultant 1988).

Dichlobenil, an inhibitor of CO2 fixation and

oxidative phosphorylation (Ware 1978), is used to

eliminate Chara, Potamogeton spp. and Myriophyllum spp.

in marshes. Its 96-h LC50 for fathead minnows is 6.0

mg/L and 8.3 mg/L for bluegill (Johnson and Finley 1980).

In normal use, dichlobenil is not toxic to fish in

treated areas (Ag Consultant 1988).

Diquat is the most widely used herbicide for control

of broadleaf weeds along ditchbanks and irrigation canals

(Gangstad 1986). In lakes and slow-moving waters, Diquat

use controls coontail (Ceratophyllum demersum),

bladderwort (Utricularia spp.) and pondweed (Potamogeton

spp.). It is a contact herbicide that reduces

photosynthetic activity (Ware 1978). Treatment of canal

banks with 2,4-D at recommended doses results in a water

concentration of only 0.025-0.061 mg/L (Gangstad 1986),

far below the 96-h LC50 of 245 mg/L for bluegill.

PCP (pentachlorophenol), on the other hand, is not

used much anymore primarily because of its extreme

toxicity to biota (Ware 1978). PCP is a non-selective

herbicide and preharvest defoliant. It has multiple

routes of action including plasmolysis and protein

precipitation and is destructive to all cells (Ware

1978). Its toxicity to fish is evident from the 96-h

LC50 of 0.21 mg/L for fathead minnow, 0.03 mg/L for

bluegill and 0.07 mg/L for channel catfish (Johnson and

Finley 1980).

Results of the current study using fathead minnow

larvae indicate that the use of Hydrothol in the aquatic

environment should be limited to properly trained

professionals. It is a highly toxic herbicide for which

there are a number of substitutes. The effect of

temperature on the toxicity of Hydrothol can be

substantial. It should be applied at the lowest water

temperature at which it will control the particular

macrophyte of interest. This temperature will vary based

on the species of plant because Hydrothol is most

effective when applied early in the growing season (Ag

Consultant 1988).



Since zooplankton are an extremely important part of

most aquatic ecosystems and contribute substantially to the

food supply of fish (Horning and Weber 1985, Mount and

Norberg 1984), these organisms have been used extensively in

toxicity tests. Early studies used either Daphnia magna or

Daphnia pulex as test organisms. However, each of these

species had their shortcomings. D. magna has a limited

distribution in aquatic systems and neither animal is easy

to culture in the laboratory.

Ceriodaphnia was chosen for use in a new subchronic

toxicity test for several reasons (Horning and Weber 1985,

Mount and Norberg 1984). Ceriodaphnia reproduce more

rapidly (3 broods in a week) than Daphnia, are ubiquitous,

and are somewhat easier to culture under laboratory

conditions (Horning and Weber 1985). The static renewal

Ceriodaphnia dubia survival and reproduction test (Horning

and Weber 1985) was developed as a substitute for the 21- to

28-day Daphnia chronic toxicity test. Toxicity is based on

survival and reproduction over a 7-day period in the newer

test. Thus, the toxic effects of chronic exposure to a

substance may be more easily and rapidly assessed than

methods using D. magna.

The Florida Department of Environmental Regulation

(FDER) requested the determination of the chronic effects of

Hydrothol-191 on C. dubia using this new test method. Their

concern stemmed from the growing use of Hydrothol in Florida

aquatic systems for control of several species of

macrophytes. Specifically, too little was known about its

long-term impacts on non-target organisms. Since this

herbicide has a half-life of 10 days (Reinert and Rodgers

1987), it can remain at potentially toxic levels in the

environment for 10 days or more. During that time,

zooplankton biomass could be seriously lowered if Hydrothol

affected both survival of adults and their reproductive

capacity. In that case, their role as a food source for

fish would be impaired.

The C. dubia survival and reproduction test was

designed to measure the effects of toxicants on survival of

adults and production of young (Mount and Norberg 1984,

Horning and Weber 1985). Tests were performed at 150 and

250 C to see if temperature at the time of Hydrothol

application would affect its impact on zooplankton. If so,

field use could be limited to times when water temperature

and plant growth activities were compatible.

Materials and Methods

Test Organism

Ceriodaphnia dubia stock obtained from EPA-Newtown,

Ohio was used to start a laboratory culture. The animals

were maintained in 1 L beakers in a 250 C environmental

chamber, with 16 hours of light and 8 hours of dark.

Test Chemical

Several formulations of endothall (7-oxabicyclo [2,2,1]

heptane-2,3-dicarboxylic acid) are used in Florida for

control of aquatic weeds and algae. However, the chronic

toxicity of the alkylamine form of endothall, i.e.

Hydrothol-191 (Pennwalt Corp., Philadelphia, PA), was

assessed in this project based on the response of

Ceriodaphnia dubia during a 7-day test. Hydrothol

concentrations were not measured, but were calculated based

on volume/volume dilutions of the 53% active ingredient (the

alkylamine salt of endothall) indicated on the product


Dilution Water

Moderately hard reconstituted freshwater (Horning and

Weber 1985) inoculated with bacteria-rich aerobically

digested trout chow and aged for one week, was used as the

culture medium (Table 4-1). The addition of bacteria and

aging of the dilution water has been suggested (De Graeve

and Cooney 1987, FDER 1986, Mount and Norberg 1984) to

stabilize water quality and increase ambient food levels.

Table 4-1. Dilution water quality parameters for
Ceriodaphnia dubia Survival and Reproduction Tests.

Parameter Mean S.D.

pH 6.84 0.08

(as mg/L CaCO3) 53.21 1.55

(as mg/L CaCO3) 86.83 1.56

(umhos/cm) 349.3 3.4

Bacteria are a major food source for Ceriodaphnia (Norberg

and Mount 1985). Thus, while cultures were fed daily, the

presence of a high background bacterial population assured

that food density was adequate to support high reproduction.

Aeration was provided by a small air pump set at minimum

output to prevent oxygen depletion by bacterial respiration.

Test Organism Food

Ceriodaphnia were fed a mixture of digested trout chow,

Cerophyll, and yeast (Horning and Weber 1985) provided at a

rate of 3 ml/L of water per day. Most cultures developed a

lush algal growth which was allowed to remain even though

water in the culture chambers was replaced weekly. The

algae provided an extra food source.

Reference Toxicant Tests

At least once a month, a reference toxicant test using

sodium dodecyl sulfate (SDS) was performed to verify that

the in-house Ceriodaphnia cultures were healthy and

nominally sensitive. That is, LC50s for SDS were compared

to those in the literature to ensure that their responses to

the test chemical were not due to an inherent sensitivity.

The SDS was obtained from EPA-Cincinnati specifically for

use as a reference toxicant.

Several toxicity tests were also performed using CuSO4.

The results of tests with CuSO2 proved to provide more

consistent results.

Range-Finding Test

A 48-hour range finding test was performed at the two

test temperatures (150 and 250 C) before definitive testing

began. Hydrothol concentrations ranged from 100-3200 ug/L

based on percent active ingredient (ai) as indicated on the

product label. Dilution and control water were moderately

hard reconstituted freshwater "conditioned" with a bacterial

inoculum and aerated for a week.

Preparation For Chronic Toxicity Tests

Approximately one week prior to the start of a test, 20

brood animals were obtained as neonates and placed in

separate 30 ml plastic cups containing 15 ml of culture

medium. They were fed 0.2 ml of the TCY mixture and 0.2 ml

of an algal mixed culture (Chlamydomonas, Klebsomidium and

Euglena ) each day. Algal supplements have been suggested

for use in C. dubia toxicity tests to promote high fecundity

(Cowgill et al. 1985). Water was changed every other day.

Neonates to be used as test organisms were harvested from

these brood chambers during a 4- hour period on about the

seventh day. They were held 12-24 hours prior to the start

of each test.

Chronic Toxicity Tests

Toxicity tests were performed at 150 and 250 C in a

constant temperature room with a 16L:8D lighting regime. To

begin each test, five toxicant solutions were prepared from

a concentrated stock diluted with moderately hard

reconstituted freshwater. Hydrothol concentrations of 25

ug/L to 400 ug/L were used. Test chambers were filled with

15 ml of toxicant or control water and the neonates were

randomly distributed among them, one to each chamber. Each

treatment consisted of 10 replicate chambers placed in a

plywood rack. Test solutions were prepared and renewed

daily. The presence and number of young were recorded for

each chamber daily before transferring the adult organisms

to fresh test solutions.

Ceriodaphnia were fed 0.2 ml of TCY and 0.2 ml of algal

culture following transfer to clean vessels. Temperature,

pH, alkalinity, hardness and conductivity of the dilution

water were measured each day. Since the dilution water was

saturated with oxygen by aeration, dissolved oxygen

measurements were not made. Each test was terminated after

7 days, and the mean young production per adult was

calculated for each treatment and the control.

Statistical Analysis

Statistical analysis followed standard EPA protocol

based on the original number of adult animals used per test

chamber, i.e. if one died, it was still included in the

calculation of mean brood number and size (Horning and Weber

1985). LC50s for Hydrothol were calculated with the TOX-DAT

Multi-method (Peltier and Weber 1985). This series of

computer programs calculates the LC50 and 95% confidence

intervals by 3 methods: moving-average angle, binomial and


Fisher's Exact Test was used to identify treatments in

which adult survival was significantly different from the

control. No further analysis was performed on such

treatments. However, reproduction data were analyzed for

toxicant levels in which adult survival was not

significantly different from the control using ANOVA and

Dunnett's Procedure. This differentiation between

treatments with and without significant adult mortality was

necessary because average reproductive capacity would have

been affected by the number of live adults. Based on the

results of the Dunnett's Procedure, the No Observed Effect

Concentration (NOEC) and Lowest Observed Effect

Concentration (LOEC) were determined and the Chronic Value

(ChV) was calculated (Horning and Weber 1985). The ChV is

the geometric mean of the NOEC and LOEC.


Reference Toxicant

Results of reference toxicant tests using sodium

dodecyl sulfate (SDS) indicated that the test organisms were

nominally sensitive (Table 4-2). LC50s varied from 2.8-5.5

mg/L. Literature values for SDS 48-hour LC50s are 1.5-8.2

mg/L for Ceriodaphnia dubia (FDER 1986) and 7-13 mg/L for

Table 4-2. Results of the Ceriodaphnia dubia 48-hour
reference toxicant tests using sodium dodecyl sulfate (SDS)
calculated by the moving average angle.

LC50 (mq/L SDS)




95% Confidence Interval




for Daphnia magna, substantially lower Daphnia magna.

Reference toxicity tests with CuSO4 produced a 48-hour LC50

of 92.7 36 ug/L.

Acute Toxicity

A 48-hour range-finding test was used to delineate the

appropriate Hydrothol concentrations for the chronic

toxicity tests (Table 4-3). The Ceriodaphnia dubia 48-hour

LC50 was 0.49 + 0.03 mg/L Hydrothol at 250 C. This agrees

well with the published LC50 for Daphnia sp. at 210 C, 0.36

mg/L (Pennwalt Corp. 1980), but is substantially lower than

values recorded for several algal species. Mudge et al.

(1986) found 1.5 mg/L Hydrothol to be the LC50 for an algal

mix (Cyclotella, Euglena, Fragilaria, Nitzschia and

Pediastrum) after five days of exposure. No other acute

data on plankton responses to Hydrothol are available.

At 150 C, the C. dubia 48-hour LC50 was 1.43 + 0.32

mg/L Hydrothol. This increased tolerance of C. dubia to

Hydrothol compared to results at 250 C may be attributable

to a lower metabolic rate at the lower temperature (Gophen

1976). Since Hydrothol is membrane-active and apparently

affects m-RNA production (Ashton and Crafts 1981), its

impacts may be dampened with decreased temperature because

processes involved in protein synthesis would be slower.

Such a response would be typical of chemical reactions in

general, as well as those mediated by enzymes (Wilson 1972).

Table 4-3. Acute toxicity of Hydrothol to various aquatic


Temp. 2C

LC50 mg/L
(95 % C.I.) Duration of Test (h)

C. dubia*

Daphnia sp.

algal mix

C. dubia









Ceriodaphnia dubia.
a Results of the current experiments.
b Pennwalt Corp. 1980.
c Mudge et al. 1986.





Chronic Toxicity

Survival. The Ceriodaphnia dubia survival and

reproduction test permits the calculation of a 7-day LC50

and uses changes in reproductive capacity over a 7-day

period as a measure of sub-lethal chronic toxicity (Horning

and Weber 1985).

At 250 C, the 7-day LC50 was 190 + 6.2 ug/L (Table 4-

4). This value is lower than the suggested Hydrothol field

application rate of 1-5 mg/L (Pennwalt Corp. 1980) by over

an order of magnitude and points to the potential hazards of

Hydrothol use in aquatic systems. Since its half-life is

approximately 10 days (Blackburn et al. 1971, Reinert et al.

1985), the impact of normal field application on the food

chain could be devastating. While fish may escape the

treated areas providing there are refugia, widespread use of

Hydrothol in a lake could markedly reduce the zooplankton

populations, which are less mobile, for 1-2 weeks after its

application. Consequently, during periods of high fish

reproduction, fry could be adversely affected by low

zooplankton availability.

Based on the results of the 48-hour tests in which C.

dubia had a higher LC50 at 150 C than at 250, it was

expected that the LC50 at seven days would also be higher

for the 150 C test. However, this was not the case. The

LC50 was significantly (p<0.05) lower at 150 C (143 + 4.6

ug/L), than at 250 C (190 + 6.2 ug/L) (Table 4-5). Why this

reversal in sensitivity occurred is not clear. It is

Table 4-4. LC50s from three replicate Ceriodaphnia dubia 7-
day Hydrothol toxicity tests at 150 and 25 C based on the
moving average angle method.

LC50 ug/L at 150 C
(95 % C.I.)

LC50 ug/L at 250 C
(95 % C.I.)










Table 4-5. Reproduction data for replicate tests of
Ceriodaphnia dubia exposed to various concentrations of
Hydrothol at 25 C for 7 days.


Final Survival

Mean (S.D.)

Mean No.













*Indicates a significant difference from control
p 0.05.

opposite to the response of several fish species tested at

18.30 and 23.30 C by Walker (1963), while fathead minnows

(Chapter 3) showed no significant change in sensitivity to

Hydrothol with a 100 C increase in temperature (15-250 C).

Over time, mortality at the two temperatures became equal.

Such results demonstrate the utility of chronic studies in

assessing the impact of a toxicant on aquatic biota.

Chronic effects of Hydrothol on reproduction. Chronic

toxicity tests are designed to measure more subtle

(sublethal) responses of organisms to toxicants than are

acute tests. The Ceriodaphnia dubia survival and

reproduction test (Horning and Weber 1985) uses changes in

reproductive rate over a 7-day period as a measure of

sublethal toxicity. The effects of a toxicant on

zooplankton reproduction is more subtle but no less

significant than its lethality. Even if a population of

Ceriodaphnia dubia survives the initial stress of toxicant

input it is still possible that fecundity may decline or

cease. The impact of such an occurrence could seriously

alter trophic level interactions in the ecosystem.

Seven-day reproduction data for C. dubia exposed to

Hydrothol at 250 C indicated that even at concentrations as

low as 25 ug/L (the lowest test concentration), Hydrothol

affected fecundity (Table 4-6). Control animals produced

2.5-2.9 broods of young each and an average of 11.6-23.6

young during the tests. At a Hydrothol concentration of 25

ug/L, fecundity was significantly (p<0.05) reduced to 1.10-

Table 4-6. Summary of the chronic toxicity of Hydrothol at
250 C to Ceriodaphnia dubia based on reproduction.

rHydrotholl 10 uq/L



Rep. 1


Rep. 2


Rep. 3


Mean No.

Mean No.



14.8(9.9) 20.7(4.7)





NOEC (ug/L)

LOEC (ug/L)

ChV (ug/L)

* Denotes significant difference from control at p < 0.05.













1.33 broods and an average of 5.0-5.9 offspring per adult

female. This lowest observed effect concentration (LOEC)

was almost eight times lower than the LC50 at 250 C, 190 +

6.2 ug/L Hydrothol.

At higher toxicant concentrations, reproduction was

reduced even more. Since Hydrothol affects the production

of m-RNA (Ashton and Crafts 1981), its impact on C. dubia

reproduction is not surprising. Normal production and

development of eggs is a process requiring adults to have

healthy metabolic capacity. If protein synthesis is

impaired by the limited availability of m-RNA, enzymes would

become limiting factors.

To determine the NOEC (no observed effect

concentration) used to calculate a chronic value (ChV), I

ran 3 additional 7-day tests with controls and 10 ug/L

Hydrothol test concentrations (Table 4-7). At 250 C, the

chronic value (ChV) for Hydrothol is less than or equal to

15 ug/L based on reproduction.

Since there was no reproduction in seven days in the

tests performed at 150 C, no statistical analysis of

toxicant effects was possible. The fact that no

reproduction occurred at this low temperature is no

surprise. McNaught and Mount (1985) found that the 7-day C.

dubia reproduction test became a 28-day test at 180 C. Even

at 200 C it took nine days for C. dubia to produce three

broods of offspring (Cowgill et al. 1985). At 150 C, the

Table 4-7. Acute toxicities to zooplankton of several
herbicides used to control submergent macrophytes in Florida



48-hour LC50








C. dubia

D. magna

D. magna

D. pulex
D. magna

D. magna

D. pulex
D. magna







Pennwalt Corp. 1980.
Johnson and Finley 1980.
Water hardness 272 ppm CaCO3.
IC50 at 26-h. Crosby and Tucker 1966.

metabolic rate decreases significantly from that at 220 C

(Gophen 1976), and was reflected in a much slower

reproductive rate.


The acute toxicity of Hydrothol to Ceriodaphnia dubia

has been determined based on survival at 48-h. My findings

confirm previous conclusions that Hydrothol is considerably

more toxic to zooplankton than some alternative compounds

(Table 3-4). For example Aquathol-K, an inorganic salt of

endothall, has a 48-h LC50 of 316 mg/L. That level is far

above field use levels (5-10 mg/L) and should not pose a

threat to zooplankton (Pennwalt Corp. 1980). Hydrothol is

often chosen over Aquathol because the former is better for

control of algae and remains effective longer.

Dichlobenil is another effective herbicide that is non-

toxic to aquatic fauna at normal use concentration (Ag

Consultant 1989). It acts to inhibit CO2 fixation and

oxidative phosphorylation in plants. Johnson and Finley

(1980) determined that the 48-h LC50 for D. pulex was 3.7

mg/L and for Simocephalus spp. it was 5.8 mg/L, both at 150


A common herbicide used in water hyacinth (Eichhornia

crassipes) control programs is 2,4-D (Ag Consultant 1989).

It is also used to eliminate broadleaf weeds in sorghum,

sugar cane and alfalfa fields. By a complex mixture of

effects at the cellular level, this herbicide inhibits cell

division and impairs nucleic acid metabolism. At 210 C, the

48-h EC50 of 2,4-D was >100 mg/L for D. magna (Crosby and

Tucker 1966).

Various algae, water hyacinth, coontail (Ceratophyllum

demersum), hydrilla (Hydrilla verticllata), pondweeds

(Potamogeton spp.) and several broadleaf weeds in

agricultural fields can be controlled with the use of

diuron. Diuron is a substituted urea herbicide (Weed

Society of America 1979) that inhibits photosynthesis by

blocking electron transport (Ashton and Crafts 1981). The

48-h LC50 was 2.0 mg/L for Simocephalus and 1.4 mg/L diuron

for D. pulex. These values represent levels exceeding those

produced by proper weed control programs (Ware 1979).

Because of their important role in the aquatic food

web, the response of zooplankton to long-term treatments

with a pesticide is important to know before it is widely

used. However, until recently there were no accepted

methods to assess impacts on zooplankton reproduction. Even

now, only acute toxicity test (24-48 hours) data are

required by EPA for pesticide registration (Zucker 1985a).

Only two other studies have provided information on the

chronic effects of endothall herbicides on zooplankton.

Serns (1975) followed the response of zooplankton to a

5 mg/L Aquathol-K exposure from June through October. Plant

control was effective, resulting in the increased presence

of Chara, but no significant change in the structure or

composition of the zooplankton community was noted.

Cladocerans and copepods exhibited their usual seasonal

changes in density.

Results from a field study of the efficacy and impacts

of Aquathol-K and Hydout, a pelletized amine formulation

used to control Hydrilla, found little effect on zooplankton

populations (Westerdahl 1983). A movement of zooplankton

into the water column as plant height decreased and an

increase in naupliar size 49 days after treatment were

noted. However, zooplankton community structure and

composition remained constant throughout the post-treatment


Results from the aforementioned studies contradict

those of my laboratory study with Hydrothol-191. I found a

significant reduction in reproduction by Ceriodaphnia dubia

in concentrations as low as 0.016 mg/L. There are several

factors that may explain these differences. First of all,

neither the study by Serns (1963) nor the work by Westerdahl

(1983) used the same formulation as I did. Serns (1963)

tested Aquathol-K, while Westerdahl (1983) used both

Aquathol-K and Hydout. The toxicity of Aquathol-K to

aquatic organisms is several orders of magnitude lower than

that of Hydrothol-191 (Pennwalt Corp. 1980, Johnson and

Finley 1980). Hydout is an amine formulation, as is

Hydrothol-191, but the former is a slow-release granular

product, while Hydrothol 191 is a liquid. This difference

may affect the amount of herbicide in solution at any time.

Second, in the laboratory study I renewed the test

solutions each day, thereby maintaining a constant exposure

level. The studies by both Serns (1963) and Westerdahl

(1983) were conducted outdoors using one application of the

herbicide. Therefore, concentrations of endothall began to

decrease immediately due to microbial degradation and

biotransformation (Reinert and Rodgers 1987).

Finally, adsorption of a herbicide may remove

significant amounts from the pool of biologically active

compound in waters containing macrophytes, algae and

sediments. Both of the field studies were performed in the

presence of natural flora and sediment (Serns 1963,

Westerdahl 1983). Thus, effective concentrations of

endothall were likely reduced compared to those in the

laboratory test vessels. The latter contained only the test

organism and solution.

This study showed that temperature had a measurable

effect on the toxicity of Hydrothol to C. dubia. However,

the relationship between temperature and survival after 2-d

exposures was inverse to that noted in 7-d tests. At 48-h,

the LC50 at 150 C was 1.43 mg/L Hydrothol, while at 250 C it

was 0.49 mg/L. With 7-d exposures, the LC50s decreased at

both test temperatures, but the survival rate was lower at

150 C than at the higher temperature. The reasons for this

contradiction are unclear. A lower metabolic rate may have

initially protected the animals in 150 C tests from the

impact of Hydrothol on m-RNA production. However, it

appears that with longer exposure low temperature compounded

the toxicity of Hydrothol. Because this herbicide is

applied in late spring or early summer, zooplankton in

Florida should not be concurrently exposed to both low

temperature and Hydrothol toxicity.



Recently, there has been growing concern over the

loss of freshwater mussel species (Unionidae) and their

declining densities in areas perturbed by pollution and

installation of dams. These molluscs, historically

abundant in most North American waters, inhabit both

lakes and streams. The area with the greatest number of

species and individuals was the Mississippi River and its

tributaries, notably the Cumberland, Tennessee and Ohio

rivers (Burch 1973).

The unusual mode of reproduction of unionid molluscs

makes their culture in the laboratory more difficult than

other groups that have free-living veliger larvae. The

life cycle of unionid mussels includes a parasitic larva

(glochidium) that normally attaches to fish gills or fins

during early development. To propagate these molluscs in

vitro, a suitable culture medium is necessary to provide

the nourishment usually obtained from the host fish.

With the hope of replenishing declining natural

populations, the Tennessee Valley Authority began funding

research to develop methods for in vitro propagation of

these freshwater molluscs in the early 1980s (1982,

1984). The goal was to eliminate the need for fish hosts

during the larval stage so that laboratory culture of

mussels would be practical. In turn, such artificial

propagation would produce a large number of juvenile

mussels for use in restoration of lost natural


As a result, a culture medium containing vitamins,

glucose, amino acids, antibiotics and fish plasma in

place of live fish was developed (Isom and Hudson 1982,

1984; Isom 1986). The transformation in culture of

glochidia to juveniles takes 9-30 days (230 + 30 C)

depending on the species, culture temperature and

glochidia maturity at the start of incubation (Isom and

Hudson 1982). While the Hudson and Isom method (1982) is

far better than that of Ellis and Ellis (1926) which

relied on the use of fish hosts for encystment of

glochidia during transformation, further simplification

is desirable. The old method (Isom and Hudson 1982,

1984) is laborious, still requires the use of fish plasma

which may not be readily available nor of consistent

quality, and a CO2 incubator.

A simplified method for mussel culture is necessary

before juveniles can be available in the numbers needed

for replenishment of declining wild stocks or for other

purposes, e.g. toxicity tests. The objectives of the

work described here were: (1) to use standard tissue

culture media and plasma available from commercial

suppliers to propagate unionid mussels in vitro as a

means of simplifying the culture technique, (2) to

determine if the use of non-bicarbonate organic buffers,

i.e. N-2-Hydroxyethylpiperazine-N'-2-ethanesulfonic acid

(HEPES) or 3-[N-morpholino] propanesulfonic acid (MOPS),

would circumvent the need for a CO2 incubator to maintain

pH, and (3) to test the efficacy of these methods in the

culture of several species of mussels.

Materials and Methods

Test Organisms

Glochidia of Anodonta imbecilis, the feeble mussel,

was used in culture experiments. Since longterm

propagation and culture of unionid mussels has not been

achieved to date, gravid females must be collected when

they are naturally available. In northern Florida,

females carrying mature glochidia can be found from April

through June. Most of the mussels used in the

development of these culture techniques were collected

from the Suwannee River, Florida. Several specimens of

A. imbecilis were obtained from Dr. Paul Yokely, of the

University of Northern Alabama.

Anodonta imbecilis was chosen because it had been

successfully cultured by Isom and Hudson (1982), it is a

widely distributed mussel and has a relatively long

reproductive period (two to three months depending on the

location). These characteristics are important when

choosing an organism as a potential bioassay animal, one

of the proposed uses of juvenile mussels produced by

these in vitro techniques.

Plasma substitutes

Culture techniques developed by Isom (1986) and

Isom and Hudson (1982) were used as the starting point

for simplification. Their culture medium, modified from

Ellis and Ellis (1926) and Eagle (1959), uses vacuum

sterilized fish plasma as a nutrient source during

culturing in place of the fish themselves. It contains a

mixture of amino acids, salts, glucose, vitamins,

antibiotics (carbenicillin, rifampin, gentamycin,

amphotericin B) and phenol red as a pH indicator. A

typical 15 X 60 mm culture dish would contain 2 ml of

medium, 1 ml of serum and 0.5 ml of the

antibiotic/antimycotic agents as described in Isom

(1986). Glochidia are removed from the gills of a female

mussel, washed and added (500-1000) to the culture medium

under a sterile hood. The plates are placed in a CO2

incubator (5% CO2) at 230 + 30 C. Isom and Hudson (1982)

found 230 C to be the incubation temperature that allowed

transformation but kept bacterial and fungal growth to a

minimum. Cultures are monitored daily under a microscope

to follow the process of organogenesis. When the foot

becomes active and other parts are developed, juveniles

are said to be transformed. They are then placed in

water where they can begin siphoning water for oxygen and


While standard tissue culture methods require the

use of plasma or serum (Ham and McKeehan 1979) because

they contain essential proteins, growth factors and

hormones that enhance cell division, there has been no

indication of their specific role in glochidia

transformation. Isom and Hudson (1982) determined that

fish plasma was an absolute necessity for successful

transformation of glochidia for all mussel species they

cultured. However, verification of their findings was

desirable since the simplification of procedures afforded

by the substitution of a more readily available protein

source would be substantial. Therefore, two

modifications of the culture medium studied were first,

the substitution of other protein sources for fish plasma

at 5% w/v and second, the use of other sera (33 % v/v)

readily available from tissue culture supply houses in

place of fish plasma.

Protein sources were acetone precipitates of trout

liver, salmon liver and rabbit pancreas, and bovine

casein (Sigma Chemical Co.). For each protein, 3 g of

powdered extract were mixed in 60 ml of distilled water

for three minutes by vortexing. The resulting slurry was

centrifuged at 1500 g for 5 minutes to remove undissolved

materials. One ml of the supernatant was used per three

ml of culture medium. Alternate sera used were bovine,

neonatal calf and horse. These sera were used at the

same final culture concentration as was fish plasma (1

ml/2ml growth medium). Glochidia were also cultured in

medium with no protein source or plasma. Culture medium

with fish plasma was used as the control.

Per cent transformation of glochidia in each culture

medium was used as the measure of success of the

modification. ANOVA and Duncan's Procedure were used to

analyze results from a total of 6 trials with 2 plates

per treatment. Three microscope fields (40X) were

counted per treatment to determine the number of

glochidia that had transformed vs those that had not.

The untransformed glochidia included those that did not

begin to develop at all due to lack of maturity, those

that did not complete transformation and those that were

non-viable after 24 h. In cases where glochidia

transformed but the juveniles were lethargic and survived

only 24 hours, such a response was taken as an indication

of morbidity and the medium judged unsuccessful in

producing juveniles for field or laboratory use.


Once the necessity for plasma was tested, the use of

organic buffers in place of the CO2 incubator as a pH-

stat was investigated. In the Isom and Hudson (1982)

method, culture pH is maintained in the optimal range

(7.3-7.4) by a HCO3-CO3 buffer system based on NaHCO3 and

CO2. The CO2 atmosphere is provided by a CO2 incubator.

MOPS (3-[N-morpholino] propanesulfonic acid) and HEPES

(hydroxyethylpiperazine-N'-2-ethanesulfonic acid), non-

bicarbonate organic buffers, are widely used in tissue

culture methods for many cell lines (Ham and McKeehan


To test their efficacy in pH maintenance in a non-

CO2 environment, either MOPS or HEPES were added to the

complete medium (0.22 %) in addition to NaHCO3, and the

pH was adjusted to 7.3-7.4 by titration with NaOH. Five

hundred to a thousand glochidia were cultured at 230 + 3

C in pairs of culture dishes containing standard medium

with bicarbonate, or medium fortified with MOPS or HEPES

(0.22% w/v) in addition to NaHCO3. Incubation

temperature was set at 23 + 30 C based on results of

Isom and Hudson's work. One dish was then placed in an

incubator with 5% CO2 at 100% relative humidity, while

its duplicate was incubated in ambient air at 24 + 30 C.

Again, per cent transformation was the parameter used for

statistical analysis.

Use of Commercial Media

A third series of experiments was designed to see if

standard tissue culture media could be used in place of

the medium developed by Isom and Hudson (1982). Their

medium must be made from many separate reagents that are

components of commercial media, e.g. Medium 199 (M199)

and Dulbecco's Modified Eagle's Medium with high glucose

(DME). The advantages of using commercial media are that

they are: (1) easier to use, (2) readily available, and

(3) manufactured under consistent conditions with quality

control that may not be possible in all research


Glochidia were cultured in the Isom and Hudson

(1982) medium, M199 or DME (with added antibiotics), and

horse serum (1 ml/2ml medium). DME and M199 were

hydrated in distilled water, adjusted to pH 7.3-7.4 with

NaOH and filter-sterilized prior to their use, just as

was the Isom and Hudson medium. Per cent transformation

was compared among the three media as a measure of media

suitability for mussel culture.

Species Cultured

Finally, Anodonta imbecilis, Lampsilis teres and

Villosa lienosa were cultured using M199, DME or Isom and

Hudson's (1982) medium and horse serum. As mentioned

before, A. imbecilis has been cultured in vitro for

several years by Hudson and Isom (1982). Hudson and Isom

(1982) have also had success culturing V. lienosa and L.

teres using fish plasma and their own culture medium.

These species are less widely distributed than A.

imbecilis but are common in northern Florida and were

collected in the Suwannee River (Burch 1973). The

usefulness of simpler culture methods would be greatly

enhanced if a number of species could be transformed

using them. Transformation of Villosa lienosa and

Lampsilis teres was also attempted using horse serum and

the commercial media.


In the first group of tests, transformation success

ranged from 0% with casein to a mean of 95.5% for

neonatal calf serum based on six trials (Table 5-1).

While they did develop, juvenile mussels transformed in

the salmon and trout media were not as healthy (inactive,

lethargic) as those from the neonatal calf and horse

media although the transformation success for these four

groups were not significantly different based on ANOVA

and Duncan's procedure (p : 0.05).

Since in vitro propagation of mussels is designed to

provide stock either for replenishment of declining wild

populations or for toxicity testing, survivability past

transformation is important. Therefore, salmon and trout

acetone precipitates of liver were judged inadequate

serum substitutes. While transformation success was as

good in neonatal calf serum (95.5 + 1.9%) as it was in

horse serum (94.7 + 4.0 %), horse serum was selected over

neonatal calf serum because the latter is more expensive.

In all cases, the use of sterile serum eliminated or

markedly decreased bacterial growth common

Table 5-1. Per cent transformation of Anodonta imbecilis
glochidia in media with various protein sources or sera
for 6 trials with 2 plates counted per treatment.

Serum or Protein Source

Neonatal calf serum

Mean %


95.5a (1.87)

Horse serum 94.7a (3.98)

Salmon liver 91.5a (5.39)

Trout liver 83.0ab (13.83)

Fish plasma 81.8b (7.47)

Rabbit pancreas

67.5b (20.17)

au"Treatments with the same letters were not
significantly different from each other (p < 0.01).

in cultures with fish plasma. This was a major problem

in earlier work (Isom and Hudson 1982).

Results from the second group of experiments testing

transformation success for Anodonta imbecilis cultures

incubated either in a CO2 (5%) atmosphere or ambient air

indicated that there was significantly (p0.05) more

transformation in C02-incubated cultures (Table 5-2).

This was true whether fish plasma, horse serum or

neonatal calf serum was used. Neither Villosa lienosa

nor Lampsilis teres transformed to the juvenile stage in

any of the media in numbers that were useful.

It was hypothesized that increased pH, apparent from

color changes of the phenol red indicator, might be the

cause of the lower success of glochidia cultured in

ambient air. So, L-15 of Liebovitz (Ham and McKeehan

1979), a medium designed for incubation of cell cultures

without a CO2 environment, was used in numerous culture

tests. None of the cultures were successful. It appears

that mussel glochidia require CO2 to transform.

Another simplification in the in vitro culture of

freshwater mussels was achieved by the substitution of

commercial powdered tissue culture media for the

idiosyncratic medium of Isom and Hudson (1982). M199 and

DME contain many of the same components found in Isom and

Hudson's medium plus 4-5 times the glucose (Table 5-3),

but can be made by simply hydrating a powder and

Table 5-2. Transformation success for Anodonta imbecilis
glochidia incubated in Isom and Hudson's (1984) basal
salt medium in CO2 (5%), in ambient air and different
buffer systems. Values are based on three trials using
two plates per treatment.

Mean %
Transformation (s.d.)

Growth medium

Fish plasma, NaHCO3
Horse serum, NaHCO3
Neonatal calf, NaHCO3

Fish plasma, MOPS1
Horse serum, MOPS
Neonatal calf, MOPS

Fish plasma, HEPES2
Horse serum, HEPES
Neonatal calf, HEPES

CO2 incubated

86.8a (6.4)
76.0ab (9.6)
60.0bcd (40.1)

78.5ab (15.6)
48.2cde (8.7)
76.7ab (9.5)

73.2ab (25.0)
77.7ab (14.1)
687abc (10.4)

Ambient air

42.Ode (2.0)
35.7e (8.5)
8.5fg (11.4)

74.1ab (22.6)
24.5ef (12.8)

75.7ab (23.6)
43.2cde (20.7)

Qa-Treatments with the same letters were not
significantly different from each (p< 0.01).
1 MOPS=3-[N-morpholino] propanesulfonic acid.

Table 5-3. Components of growth media from Isom and
Hudson (1982), DMEa and M199a. Concentrations are in


Isom and Hudson

choline chloride
folic acid
nicotinic acid
pantothenic acid
calcium pantothenate
p-amino benzoic acid
Ferric nitrate