Environmental Research Laboratory
Office of Research and Development
RECENT ADVANCES IN FISH TOXICOLOGY
A Symposium
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Corvallis, Oregon 97330
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Healtn Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials Problems are assessed for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service Springfield, Virginia 22161.
-------
EPA-600/3-77-085
July 1977
RECENT ADVANCES IN FISH TOXICOLOGY
A Symposium
edited by
Richard A. Tubb
Department of Fisheries and Wildlife
Oregon State University
Corvallis, Oregon 97331
sponsored by
Corvallis Environmental Research Laboratory
Corvallis, Oregon 97330
in cooperation with
Department of Fisheries and Wildlife
Oregon State University
Corvallis, Oregon 97331
CORVALLIS ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CORVALLIS, OREGON 97330
-------
DISCLAIMER
This report has been reviewed by the CorvaTlis Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental Protection
Agency, nor does mention of trade names or commercial products consti-
tute endorsement or recommendation for use.
-------
FOREWORD
Effective regulatory and enforcement actions by the Environmental
Protection Agency would be virtually impossible without sound scientific
data on pollutants and their impact on environmental stability and human
health. Responsibility for building this data base has been assigned to
EPA's Office of Research and Development and its 15 major field install-
ations, one of which is the Corvallis Environmental Research Laboratory
(CERL).
The primary mission of the Corvallis Laboratory is research on the
effects of environmental pollutants on terrestrial, freshwater, and
marine ecosystems; the behavior, effects and control of pollutants in
lake systems; and the development of predictive models on the movement
of pollutants in the biosphere.
This report is a compilation of reports presented at the Symposium on
Recent Advances in Fish Toxicology, January 13-15, 1977 in Corvallis,
Oregon. The Symposium was cosponsored by The Corvallis Environmental
Research Laboratory and the Department of Fisheries and Vlildlife, Oregon
State University.
A.F. Bartsch
Director, CERL
n
-------
PREFACE
The symposium Recent Advances in Fish Toxicology was held in Corvallis,
Oregon on January 13 and 14, 1977. The Corvallis Environmental Research
Laboratory of the United States Environmental Protection Agency and the
Department of Fisheries and Wildlife of Oregon State University cosponsored
the symposium to encourage the rapid communication of recent findings between
fish toxicologists. New legislation has increased the need for communi-
cation between fish toxicologists, and the 1976 Toxic Substances Act
(PL 94-469) indicates a new era is beginning for water pollution control.
The law now requires the clearance of new chemical products that might
enter waterways before such substances are manufactured and sold.
Prediction of the probable toxic effects to fish and other aquatic organisms
must be based on a developing assessment methodology. Symposium participants
attempted to summarize some of the recent findings in fish toxicology or
pointed out the research that is needed to meet the new legislative mandate.
The symposium is dedicated to Professor Peter Doudoroff who is concluding
a long and active research and teaching career. His pioneer research with
the Public Health Service and Oregon State University helped to define
many of the physical and chemical conditions required for aquatic life.
The results of his research have been applied by many countries to establish
water quality standards. He has been generous in his advice and counsel
and many of the symposium presentations were made by former students and
colleagues.
Richard A. Tubb
Head of Department of
Fisheries and Wildlife
Oregon State University
-------
CONTENTS
Foreword iii
Preface iv
Introductory Remarks ]
A.F. Bartsch
A Multiple Approach to Solving the
Gas Supersaturation Problem 4
R.R. Carton and A.V. Nebeker
Effects of Kepone on Estuarine Organisms 20
D.J. Hansen, D.R. Nimmo, S.C. Schimmel,
G.E. Walsh, and A.J. Wilson, Jr.
Collagen Metabolism in Fish Exposed to Organic Chemicals 31
F.L. Mayer, P.M. Mehrle and R.A. Schoettger
Effects of Short-Term Exposures to Total Residual
Chlorine on the Survival and Behavior of Largemouth
Bass (Micropterus salmoides) 55
G.L. Larson and D.A. Schlesinger
An Approach for Studying the Effects of Mixtures of
Environmental Toxicants on Whole Organism Performances ......... 71
C.F. Muska and L.J. Weber
Relationship Between pH and Acute Toxicity of Free
Cyanide and Dissolved Sulfide Forms to the Fathead Minnow 88
Steven J. Broderius and Lloyd L. Smith, Jr.
The Acute Toxicty of Nitrite to Fishes 118
R.C. Russo and R.V. Thurston
Copper Toxicity: A Question of Form 132
G.A. Chapman and J.K. McCrady
The Role of Cyanide as an Ecological
Stressing Factor to Fish 152
Gerard Leduc
An Assessment of Application Factors in Aquatic Toxicology 183
O.I. Mount
Closing Remarks—An Old Frog Croaks an Appeal for Logic 191
P. Doudoroff
V
-------
INTRODUCTORY REMARKS
A. F. Bartsch, Director
Corvallis Environmental Research Laboratory
U, S. Environmental Protection Agency
200 S.W. 35th
Corvallis, Oregon 97330
I want to begin my remarks with a quotation:
"To a far greater extent than ever before, we live in a man-created and
man-controlled environment. It is within our power to shape our own future,
to guide the evolving patterns of society and determine the nature of the
surroundings in which we and our children will live.
". . .it might be helpful if I were to sketch out for you, in very
broad strokes, the view of the water pollution problem from the national
window of a federal agency charged with rather far-ranging responsibilities
in this field.
"In doing so, I should like to develop four principal points:
"First, that water pollution control is an integral part of the broader
problem of water resource development and use;
"Second, that water pollution control is an inseparable part of the
broader problem of environmental health protection;
"Third, that an impressive amount of productive activity is already
underway in controlling water pollution;
"And fourth, that the problem demands a still stronger effort on the
part of federal, state, and local authorities, industries, and all others
concerned."
These were remarks delivered by Dr. Leroy E. Burney, Surgeon General of
the Public Health Service, in an opening address before the National Confer-
ence on Water Pollution held 16 years ago on December 12.
You may have noticed there was no plea for attention—scientific or
otherwise—to pollution impacts on fish and other aquatic life. In fact,
the existence of aquatic life was not even acknowledged. The same was true
in the 30 recommendations that came from the conference.
-------
At the time of that conference in 1960, Or. Doudoroff had already been
working as a pioneer in the field of water pollution biology for almost 20
years. I am sure, in those early days of the '40's, he was on a first-name
basis with all of his colleagues in the United States. They numbered only a
handful. It was a challenging time because the ground was fertile for
plowing with much to be discovered. It was also a frustrating time because
the place of biology and the role of biologists in water pollution control
was neither well understood nor widely accepted. These conditions have
changed.
Water pollution control did not become a national movement until 1949.
Previously, a few progressive states were moving forward with vigor; others
were standing still. Water pollution control programs, where they existed,
consisted mainly of efforts to stop discharging untreated municipal sewage
in order to protect human health and to terminate public nuisances. Perhaps
some of you here today remember what the Willamette River, flowing through
this city, was like in those days. Industry had no strong motivation to
protect freshwater and marine organisms.
It is fitting that this symposium be dedicated to Dr. Peter Doudoroff.
His past and continuing monumental contributions to the subject of fish
toxicology have not only generated new knowledge that this nation needed but
also served to train researchers and stimulate many others. It is appropri-
ate at this time for a symposium to look at recent advances in fish
toxicology. In this stock-taking, we should be mindful of the broader
framework into which these efforts fit today. That framework has many
aspects; two are especially notable.
One point is the growing frequency and severity of environmental crises
that come largely from human ignorance, indifference and economic greed. We
have just finished the worst year in our history of major oil spills (15
tankers and 200,OOOT). Crises involving mercury, PCB's, asbestos, and kepone
are still fresh memories. The "Legionnaires' Disease" has us baffled.
History will show that we and other nations have been ineffective in fore-
seeing the next environmental crisis.
The other aspect is more encouraging. Today the nonhuman side of
environmental pollution has been acknowledged as important. During the last
three fiscal years, EPA assigned from 83-89% of its research dollars to
problems other than human health—much obviously in biologically-oriented
areas. There is no action impinging on environmental quality that is not
noticed by some powerful, national citizens' organization. And finally,
laws to protect the environment are becoming stronger (unfortunately more
complex) and more far-reaching. In the context of this symposium, one of
the most important pieces of legislation may turn out to be the Toxic
Substances Control Act (PL 94-469) signed into law on October 12, 1976.
The Act is 49 pages long. It is complex. Its far-reaching impact can
be surmised somewhat from its policy statement. Let me quote--
"It is the policy of the United States that—
-------
"(1) adequate data should be developed with respect to the effect of
chemical substances and mixtures on health and the environment and that the
development of such data should be the responsibility of those who manufac-
ture and those who process such chemical substances and mixtures;
"(2) adequate authority should exist to regulate chemical substances and
mixtures which present an unreasonable risk of injury to health or the envi-
ronment, and to take action with respect to chemical substances and mixtures
which are imminent hazards; and
"(3) authority over chemical substances and mixtures should be exercised
in such a manner as not to impede unduly or create unnecessary economic
barriers to technological innovation while fulfilling the primary purpose of
this Act to assure that such innovation and commerce in such chemical sub-
stances and mixtures do not present an unreasonable risk of injury to health
or the environment."
In the past, much of the bioassay development effort focused on lethal
effects of toxic substances. Efforts today, and especially those responding
to the Toxic Substances Control Act, will emphasize sublethal effects. Many
of you may be involved in this activity. If you are, your work will be more
effective and successful because of the foundation that Dr. Doudoroff and his
colleagues have established.
-------
A MULTIPLE APPROACH TO SOLVING THE
GAS SUPERSATURATION PROBLEM
R. R. Garton and A. V. Nebeker
Western Fish Toxicology Station
. S. Environmental Protection Agency
1350 S.E. Goodnight Avenue
Corvallis, Oregon 97330
ABSTRACT
Gas supersaturation of water was first recognized as
a serious problem in the Snake and Columbia rivers of the
Pacific Northwest. To solve the problem, a multiple
approach was used combining laboratory and field studies
to determine sources, effects, persistence, and preven-
tion of the supersaturation. Classical bioassays were
used to determine effect, but additional tests were
needed because of the unique nature of supersaturation.
These tests included assessment of avoidance capability
of fishes, assessment of depth compensation and tempera-
ture effects, and field surveys of aquatic organism
distribution in the affected areas. Data from the
combined approaches were used to set safe levels for
aquatic organisms. In addition, engineering expertise
from other groups was applied in an attempt to prevent
or mitigate the effects of supersaturation.
INTRODUCTION
The purpose of this paper is to demonstrate how a cooperative effort by
a number of agencies was used in an attempt to solve the problem of gas
supersaturation of water in the Columbia River Basin and other rivers and
coastal waters of the U.S. The multiple approach combined classic toxicity
studies, newly-designed special effect studies, field studies, and engineer-
ing expertise. The desire is not so much to present data on any particular
part of the study as to chronicle the overall effort and to show how toxicity
studies are an integral part of the problem definition and solution.
-------
THE PROBLEM
Air supersaturation of water was first noted in hatchery and aquarium
facilities in the early 1900's (Gorham 1901), and was ascribed as causing the
condition in fish known as gas bubble disease. Based on Gorham's conclusions
and the more precise analytical methods of Van Slyke and Neill (1924) for
nitrogen determination, dissolved nitrogen gas was postulated to be the cause
of gas bubble disease and received primary research emphasis. Supersatura-
tion caused by entrainment of air in water spilled over dams first became a
problem on the Columbia River when Bonneville Dam was constructed in 1938,
although it was apparently undetected at that time. As more dams were con-
structed on the river the problem increased; water was supersaturated by each
dam with the result that the entire river could be supersaturated during some
periods of the year (Ebel 1969; Weitkamp and Katz 1973, 1975; Rucker 1972).
Considering only the Snake River Chinook salmon and steel head stocks, Ebel
et al. (1975) forecast a total loss of 2 million adult fish during the period
1976 to 2000 if no remedial actions were taken to reduce the hazards of super-
saturation. In economic terms (1974 dollars), this loss would range between
$47.2 and $126.9 million. These figures are very conservative since they do
not take into account stock from the Columbia River and its tributaries, nor
other species such as sockeye and coho salmon.
ORGANIZATION FOR SOLVING THE PROBLEM
As supersaturation on the Columbia and Snake rivers became more severe
the U. S. Secretary of the Interior and the Tri-State Governors Conference
(Idaho, Washington, and Oregon) formed a Nitrogen Task Force to work out
solutions to the problem. The task force was made up of representatives from
23 public and private agencies. One of their first duties was to organize a
division of labor between various agencies for research on the problem. The
Environmental Protection Agency, represented by the Western Fish Toxicology
Station, was detailed to carry out laboratory studies to determine safe
levels of supersaturation for both adult and juvenile salmonids. This study
was later expanded to include food organisms as well as predators and com-
petitors of these fishes. The National Marine Fisheries Service and the
states of Washington, Oregon, and Idaho were detailed to conduct field studies
to determine effects and persistence of supersaturation in the Columbia River
Basin. They also conducted some laboratory studies on effects.
The U. S. Army Corps of Engineers, in cooperation with the Bonneville
Power Administration, were already operating the dams and regulating the flow
on the lower river; their responsibility for flow manipulation and control of
the power plants continued. In addition the Corps funded fisheries research
through the National Marine Fisheries Service laboratories and conducted
engineering and modeling studies to provide solutions to the problem.
The Northwest Utility Cooperative, together with public and private
research organizations such as Parametrix, Inc., and Battelle Northwest,
provided funding and cooperation for additional research (Weitkamp and Katz
1973, 1975).
-------
DEVELOPMENT OF MEASUREMENT TECHNIQUES
At the beginning of the supersaturation study, measurement techniques
had to be developed for use in both laboratory and field experiments. Pre-
viously two standard techniques had been employed, the standard Winkler
determination for dissolved oxygen (APHA 1971) and the Van Slyke method for
nitrogen determination (Van Slyke and Neil! 1924). The oxygen determination
method was quick and easily used, but did not determine nitrogen or total gas
saturation levels. The Van Slyke method measured 02. N£ and total gas, but
was relatively cumbersome and difficult for field use. A third technique,
the gas chromatograph, was also available for determining nitrogen, oxygen,
or carbon dioxide. It is relatively quick and precise but requires expensive
equipment which is not readily portable (Fickeisen et al. 1975). The develop-
ment of the Weiss saturometer in 1970 significantly changed the method of
analysis, and made field and some laboratory determinations relatively easy
to accomplish. The saturometer, developed by Ray Weiss (1970, unpublished
data, Scripps Inst. Oceanog., La Jolla, Cal.) and adapted for field use by
Robert Rulifson (S. Lambert and R. L. Rulifson, 1972, unpublished report,
U. S. Environmental Protection Agency, Seattle), consists of a metal frame-
work upon which is wound a length of about 100 ft (the length is variable) of
semi-permeable, medical-grade, silastic tubing. The tubing is permeable to
gas when submerged but is not permeable to water; thus, the gas in the water
goes through the wall of the tubing until gas pressure within the water and
the tubing becomes equilibrated. The gas pressure within the tubing is then
measured by a manometer to determine the difference between gas pressure in
the water and gas pressure in the atmosphere. With a known atmospheric pres-
sure one can calculate percent total gas pressure in the water as compared to
that in the atmosphere. This device does not differentiate between nitrogen
pressure and oxygen pressure. However, it can be used in conjunction with
the Winkler method for dissolved oxygen. The saturometer determines total
gas pressure, and by subtraction of oxygen, nitrogen pressures can be
obtained.
FIELD WORK TO DETERMINE EXTENT AND MAJOR PROBLEMS
Intensive field work on the Columbia River was begun in 1966 to determine
causes of supersaturation and its effect on salmon and steel head trout (Ebel
1969). These studies showed that the primary cause of supersaturation was
the spilling of water at dams on the river with a direct correlation between
amount of spill and saturation levels. As the water spilled over the dams it
entrained air, carried it to great depths where it was held under pressure
and dissolved into the water. This water, upon return to lesser depths and
pressure, became supersaturated with both oxygen and nitrogen.
To illustrate the effect of a change in pressure, at 10 C (50 F) one
liter of air-saturated water will hold 36.25 cc of air at a depth of 6.1 m
(20 ft) of water exerting a total pressure of 1208 mm of mercury. It the
water is returned to the surface with a pressure of only 760 mm of mercury
pressure, one liter of air-saturated water will hold only 22.8 cc of air.
Thus, the water will be supersaturated to 159%.
-------
Since passage of water through the turbines did not cause supersatura-
tion at the dams on the Columbia and Snake rivers, there was an initial
tendency on the part of researchers to ignore turbines from other sites as
possible sources of supersaturation. However, studies conducted by Western
Fish Toxicology Station and Bureau of Reclamation personnel in Colorado
(Garton et al. 1973) found supersaturation as high as 130% produced in the
turbine structures at Morrow Point Dam on the Gunnison River. This super-
saturation was the result of insertion of air into the penstock to cushion
the fall of water against the turbine blades. The air dissolved into the
water in the deep draft tube between turbine blades and the outlet to the
river. This same situation was found by MacDonald and Hyatt (1973) on the
Mactaquac River in New Brunswick.
Thermal power plants were also identified as sources of supersaturation;
water at a low temperature can hold more air than water at a higher tempera-
ture and when heated in a thermal power plant it becomes supersaturated.
Depending upon the temperature, water will increase in saturation from 2.0 to
2.8% for every 1 C rise in temperature.
To illustrate the effect of temperature increase, at 3 C (37.4 F) and
at 760 mm of mercury pressure, one liter of air-saturated water will hold
26.9 cc of air. In a power plant with a 16 C (28.8 F) temperature rise (AT),
the effluent temperature would be 19 C (66.2 F). At this effluent tempera-
ture, a liter of air-saturated water would hold only 19.02 cc of air. If air
is not released back to the atmosphere to compensate, the water will be super-
saturated to 142%. A AT of 16 C is high for a power plant but not unreason-
able and, of course, lower AT's also cause supersaturation but in corres-
pondingly lesser amounts.
At the Pilgrim Plant in Massachusetts, thousands of menhaden were killed
by supersaturation when they chose, because of temperature preference, the
warmed but supersaturated discharge canal (Marcello et al. 1975). Since that
time additional sources of supersaturation from thermal power plants have
been identified in both salt and fresh water sites, such as the Green River
in Wyoming (Roy Hamilton, personal communication).
Along with the question of source of supersaturation also comes the
question of determining persistence of the condition. Persistence is largely
determined by the ratio of surface area to volume of the body of water and by
the amount of turbulence at the air-water interface. The Snake and Columbia
rivers, which are deep, slow-moving river-reservoir systems, do not easily
lose dissolved gases. The supersaturated condition may, at times, persist
from upstream dams in Idaho and central Washington all the way to the Pacific
Ocean (Ebel 1969), making supersaturation an especially serious problem.
Garton et al. (1973) found that supersaturation in the Gunnison and Frying
Pan rivers in Colorado was not nearly so persistent. A saturation level of
130% in the turbulent Gunnison River was reduced to 100% in less than eight
miles of flow. The small turbulent Frying Pan River reduced supersaturation
levels from over 115% to 100% in less than three miles of flow. May and
Huston (1975) in Montana found that supersaturation in the Kootenai River
(average peak flows of 65,000 cfs) persisted for 30 miles downstream. High
flows (above 20,000 cfs) kept gas concentrations above 125% saturation as far
-------
as 20 miles downstream from the dam, and at times gas levels had not reache.d
equilibrium 100 miles downstream. The Columbia River below Bonneville Dam at
times remained supersaturated above 110% all the way to Astoria; while the
Snake River, below Hells Canyon Dam was still supersaturated at the mouth of
the Salmon River.
Measurement of effects of supersaturation on fish in the field is often
difficult because of the problems of sampling for dead or injured fish in the
large bodies of water affected. However, Ebel et al. (1975) and others
(Weitkamp and Katz 1975) noted obvious effects on returning fishes at the
ladders on the Snake and Columbia rivers. Gas bubbles were observed in both
adult and juvenile fish, and mark and recapture studies demonstrated high
mortality in downstream migrants. In addition, high supersaturation levels
have reduced return rates of fish released for downstream migration.
Excess spilling of water over the newly-completed dam on the Kootenai
River near Libby, Montana, supersaturated the water with air and killed many
fish in the river downstream (May and Huston 1975). There was a marked
reduction of whitefish numbers, in part from mortality of juveniles due to
gas bubble disease. Large-scale suckers in the first 10-15 miles below the
dam had a high incidence of gas bubble disease, but their numbers remained
high, indicating that they were able to tolerate high gas concentrations,
possibly, in part, due to their preference for deeper water. Complete
mortality of cutthroat trout and mountain whitefish occurred in 1-5 to 3-8
days when held within two feet of the water surface at total gas levels of
131-139% saturation. In volition cages extending to 10 ft the trout still
suffered 55% mortality and the whitefish suffered 67% mortality after 24 days.
"TOXICITY" STUDIES TO DETERMINE EFFECTS AND SAFE LEVELS
Although supersaturation is not a toxic substance in the classic sense,
traditional toxicity studies (bioassays) were used to determine effects of
supersaturation and safe levels for aquatic organisms. Additional experi-
ments were designed along with the toxicity tests to study effects such as
acclimation, avoidance, and recovery from supersaturation.
Before supersaturation could be studied in the laboratory a system had
to be developed to produce supersaturated water in the test tanks. Such
systems were developed simultaneously at the Western Fish Toxicology Station
(Bouck et al. 1976; Nebeker et al. 1976) and by the National Marine Fisheries
Service (Dawley et al. 1976). These systems attempted to simulate pressure
change much like the source of supersaturation at dams. Here atmospheric air
of bottled gases such as nitrogen, oxygen, or carbon dioxide were injected
into water under pressure where they were dissolved to saturation at high
pressures. When the pressure was relieved by release of the water into the
test aquaria or tanks the water became supersaturated with the gas.
The first approach was to determine classic TL50 data for salinonid
fishes of all ages including eggs, embryos, young fish through smolt stage,
and the adult returning to spawn. These studies were conducted in shallow
tanks (.10-60 cm in depth) of varying sizes, with precise temperature control,
where the fish could be exposed to carefully monitored levels of
8
-------
supersaturation and closely observed for effect. Both short-term and long-
term chronic studies were conducted. Longer-term experiments also made pos-
sible detailed pathology studies (Figures 1, 2, 3) to determine and document
the development of the "gas bubble disease" in the fish (Stroud et al, 1975).
Salmonid eggs are seldom exposed to supersaturated conditions in the field.
Because of the importance of supersaturation to hatcheries in the Pacific
Northwest, especially where the water is heated to speed up growth of young
fish, egg through swim-up studies were conducted at the Western Fish Toxi-
cology Station. Nebeker et al. (1977a) determined that 125% supersaturation
was a safe level for salmonid eggs and young sac-fry larvae in shallow tanks
in the laboratory; however, when the swim-up stage developed they died at
gas levels as low as 113% saturation (Figure 4}. Studies conducted by Lorz
and McPherson (1976) suggested that the ability of smolts to migrate and
adapt to sea water may be especially sensitive to some pollutants. Nebeker
et al. (1977c) conducted similar smolt studies using supersaturation as a
toxicant and found that sublethal levels of supersaturation which caused gas
bubble disease had no discernible effect on ability of salmonids to smolt
and to acclimate to salt water.
Because salmonid stocks could be affected by the effects of super-
saturation on their predators, competitors, and food organisms, other tests
in addition to TL50 studies were conducted. Studies conducted at the Western
Fish Toxicology Station (Bouck et al. 1976) with a predator, the largemouth
bass, and juvenile salmonids in the same tank of supersaturated water
demonstrated that bass could tolerate supersaturation levels that killed or
injured young salmonids. Similar studies conducted with squawfish by Meekin
and Turner (.1974) showed that squawfish were more tolerant than salmonids but
ceased to feed at higher saturation levels. Food organism studies with
Daphnia, crayfish, and aquatic insects were conducted at the Western Fish
Toxicology Station utilizing acute, long-term, and full-life cycle studies
(Nebeker 1976). These studies showed that, in general, invertebrates (with
the possible exception of Daphnia magna] were more tolerant to supersaturated
water than fishes.
Because of the special nature of supersaturation, use of the classic
TL50 experiments alone did not provide sufficient data to propose safe
levels for aquatic life. For any given level of gas in the water the percent
saturation is dependent upon both the pressure and the temperature of the
body of water, A rise in water temperature or a decrease in pressure
increases supersaturation. The pressure phenomenon is especially important
in rivers such as the Snake or Columbia because water which is saturated to
130% at the surface, for example, will be saturated to only 100% at a depth
of 10 ft. Fish staying in deeper water escape effects of supersaturation.
Dawley et al. (1976), on a grant from the Environmental Protection Agency,
tested juvenile Chinook and steelhead trout in both shallow and deep tanks
and found that fish tended to move to the slightly deeper levels with
increased levels of supersaturation. Similar results were obtained by Blahm
et al. (1976). However, both studies left a serious question unanswered. It
was not known whether the fish stayed in the deeper areas of the tank to
escape supersaturation or because the configuration of the tank made them
prefer the security of the deeper water during increased stress. M. D.
Knittel and coworkers at the Western Fish Toxicology Station (unpublished
-------
Figure 1. Adult sockeye salmon showing gas blisters in the
mouth and on the left opercle.
10
-------
Figure 2. Gill arch of adult Chinook salmon showing gas blisters
on the gills and rakers.
11
-------
•'•*«• % -
Figure 3. Cross section of adult Chinook salmon flesh showing
cavities (swiss cheese effect) in the muscle.
Figure 4. Steel head trout finger!ing with gas bubble disease.
12
-------
manuscript) held fish in cages (30 inches in diameter by 8 inches deep) at
different levels in 10 ft deep tanks. They determined that fish are able to
compensate for supersaturation by moving to deeper levels in the water but
that the compensation by the fish is not quite as great as one would predict.
From the gas laws we assumed compensation of 5% per 20 inches increase in
depth. However, they found that the compensation was only 3.7% per 20 inches
increase in depth. They also found that the longevity of fish was increased
by submersion to greater depths after exposure to lethal levels of gas super-
saturation.
Even though it was determined that fish compensated for supersaturation
by seeking greater depths there was still the question as to whether fish can
sense supersaturation and whether the fish will voluntarily go to a level
where supersaturation is lower. Stevens et al. (1977 unpublished manuscript)
in their avoidance studies showed that the ability to detect and avoid super-
saturated water seems to vary among species. Salmon were able to differen-
tiate between different percentages of supersaturation in a pie-shaped
avoidance chamber. However, trout were not clearly able to do so consist-
ently. Avoidance of supersaturation might be secondary to behavior such as
increased activity due to aggression and territonality, a need for cover and
lower light intensities, or choice of a particular current velocity. Schiewe
and Weber (1976) found that bubbles in the lateral line of fish exposed to
supersaturated water diminished or completely blocked the ability of the
sensory units to respond to stimuli. The loss of ability to respond to
stimuli decreases the fish's capability to detect objects or locate predators.
Chapman and Nebeker (1977 unpublished manuscript) considered the possibility
of synergism between supersaturation and the heavy metals copper and zinc,
but were unable to detect any such effect. Nebeker et al. (1977b) detected
effects of temperature on fish survival in supersaturated water. With
juvenile steelhead trout, tested from 9 to 18 C at 116% saturation, each one
degree (C) increase decreased time to 50% death by about 30 hours, from 330
hours at 9 C to about 50 hours at 18 C. Increased temperatures significantly
decreased survival time of steelhead and Chinook salmon, but no significant
effect was apparent for sockeye or coho salmon.
One of the results of the toxicity studies was the determination of
total-air-saturation water quality criteria for the protection of fish and
other aquatic life. These criteria were published in the 1976 document,
"Quality Criteria for Water," published by the Environmental Protection Agency
in compliance with Public Law 92-500. In this document, 110% total-air-
saturation was stated as a safe level for salmonid fishes in shallow water.
These criteria were established with full knowledge that fish would be able
to tolerate higher levels of supersaturation at greater depths. However,
sublethal effects of supersaturation were noted at 110%, and lethal levels
are found to be not far above 110% (Nebeker and Brett 1976). Thus, it was
determined that the safe level is near 110%, especially in shallow waters of
hatcheries or fish-rearing areas where depth compensation is not possible,
but that deviations from this may be justified in some specific cases (Table
1, Figure 5).
13
-------
700- II
it
600-
o
500t
< 400'
i—
cr
O
300
O
C\J
O
h-
LlJ
-I
200-
100
A Adult Chinook
n Adult Sockeye
« Adult Coho
• Adult Steelhead
• = Tests with less than
20°/0 mortality
• - Observed time to
20°/o mortality
•4-
•4-
105
110 115 120
PERCENT SATURATION
125
Figure 5. Determination of threshold concentration (114%) for
adult salmonids ( • = 10 fish/test).
14
-------
TABLE 1. COMPARATIVE SENSITIVITY OF JUVENILE AND ADULT SALMONIDS AND BASS TO
AIR-SUPERSATURATED WATER.
Fish Threshold (% sat.)'
Sockeye smolts
Juvenile steelhead
Juvenile sockeye
Adult sockeye
Steelhead smolts
Adult coho
Adult steelhead
Adult chinook
Coho smolts
Juvenile coho
Adult bass
Juvenile bass
113.6 %
113.8
114.0
114.2
114.2
114.4
114.6
114.7
114.8
118.0
126.8
128.0
*based on time to 20% mortality (as determined using methods shown in Fig. 1)
SOLUTIONS
Solutions to the problem in the Pacific Northwest, in the form of struc-
tural modifications of the dams, were determined and are being implemented
primarily by the Corps of Engineers. National Marine Fisheries Service
personnel, funded by the Corps of Engineers, also constructed screening
structures for trapping and hauling downstream migrants around problem areas.
Solutions to the problem can be based either on initial planning to avoid
causing supersaturation or by reduction of supersaturation when it cannot be
avoided due to design of existing facilities. Reduction or elimination of
supersaturation is sometimes a feasible alternative at hatcheries or other
areas where a flow of waters is involved. However, reduction of supersatura-
tion in rivers such as the Columbia is an almost-impossible task and the
problem should be attacked at the start, if possible, by avoidance of super-
saturation production.
Prevention of supersaturation in the Snake and Columbia river system is
approached in two different ways. The first is by manipulation of river
flow to avoid spilling at dams where supersaturation may be produced. The
second is by physical changes in the structure of the dams themselves, such
as the flip lip (Boyer 1974). Another method used to help the fishery
resource is to avoid the supersaturated water completely by collecting fish
with traveling screens at dams, such as Little Goose or Lower Granite on the
Snake River, and trucking the fish in tank trucks to the Lower Columbia River
below Bonneville Dam. This precaution has the advantage of avoiding a large
part of the supersaturated river for downstream migrant fishes; however, it
has the disadvantage of high trucking costs. This solution is currently
under study by the Corps of Engineers and by the National Marine Fisheries
Service to determine feasibility and effect of trucking, or air freighting
15
-------
young fish downstream. Hopefully, continuation of the field studies on young
fishes will determine whether the trucking, the flow manipulations, or the
physical changes are ^tually doing the job in reducing supersaturation effect
on migrating salmonids in the Columbia-Snake river system.
SUMMARY
The supersaturation research project is an example of a case where bio-
assays (classic TL50 and specially designed depth compensation studies) were
used in conjunction with other field and engineering research methods to set
criteria and solve a specific problem. In this case "toxicity" studies were
used as a definite part of the problem-solving technique. Similar studies
have also been conducted on polychlorinated biphenyls, mi rex, and other
pollutants. This approach is likely to continue in the future when a pollu-
tant becomes known as important, and an all-out effort is mobilized to solve
the problem. Solution of the problem depends upon identifying the source and
effects of the pollutant, and determining safe levels through laboratory
studies. The supersaturation study was different from many in that it has
not resulted in enforcement proceedings. In this case it was a cooperative
effort coordinated by the Nitrogen Task Force between state pollution control
agencies and the U. S. Environmental Protection Agency, and between the
fishery resource agencies of the states of Oregon, Washington, and Idaho, and
the National Marine Fisheries Service. The Corps of Engineers, the Bureau of
Reclamation, and the Public Utility Districts shared responsibility for dam
modification and other methods used to decrease the problem or its effect.
In this case there was a cooperative effort to cure mistakes in design which
were carried over from the past when the problem of supersaturation was
well understood.
16
-------
REFERENCES
American Public Health Association et al. 1971. Standard methods for the
examination of water and wastewater. 13th ed. Am. Public Health
Assoc., New York. 874 p.
Blahm, T. H., B. McConnell, and G. R. Snyder. 1976. Gas supersaturation
research, National Marine Fisheries Service, Prescott Facility--!971 to
1974. Pp. 11-19 jn_ D. H. Fickeisen and M. J. Schneider (eds.), Gas
bubble disease. Proceedings of a workshop cosponsored by Battelle
Pacific Northwest Laboratories and U. S. Atomic Energy Commission.
CONF-741033. (Held in Richland, Wash. Oct. 8-9, 1974.) Energy Res.
& Devel. Admin., Technical Information Center, Oak Ridge, Tenn. viii +
123 p.
Bouck, G. R., A. V. Nebeker, and D. G. Stevens. 1976. Mortality, saltwater
adaptation and reproduction of fish during gas supersaturation. Ecol.
Res. Ser. EPA-600/3-76-050. Office of Res. & Devel., U. S. Environ-
mental Protection Agency, Duluth, Minn, ix + 55 p.
Boyer, P. B. 1974. Lower Columbia and Lower Snake rivers; nitrogen (gas)
supersaturation and related data: analysis and interpretation.
Contracts DACW57-74-0146 and DACW57-75-C-0055. North Pacific Division
Corps of Engineers, Portland, Ore. 20 p + appendix [7 p.]
Dawley, E., B. Monk, M. Schiewe, F. Ossiander, and W. Ebel. 1976. Salmonid
bioassay of supersaturated dissolved air in water. Ecol. Res. Ser.
EPA-600/3-76-056. Office of Res. & Devel., U. S. Environmental Protec-
tion Agency, Duluth, Minn, ix + 39 p.
Ebel, W, J. 1969. Supersaturation of nitrogen in the Columbia River and its
effect on salmon and steel head trout. Fishery Bull. 68(1): 1-11.
Ebel, W. J., H. L. Raymond, G. E. Monan, W. E. Farr, and G. K. Tanonaka.
1975. Effect of atmospheric gas supersaturation caused by dams on
salmon and steel head trout of the Snake and Columbia rivers (A review
of the problem and the progress toward a solution, 1974). Northwest
Fisheries Center, National Marine Fisheries Service, Seattle, Wash.
Ill p. Processed.
Fickeisen, D. H., M. J. Schneider, and J. Montgomery. 1975. A comparative
evaluation of the Weiss saturometer. Trans. Am. Fish. Soc. 104(4): 816-
820.
Garton, R. R., H. A. Salman, and F. C. Heller. 1973. Sources of gas super-
saturation in water. Western Association of State Game and Fish Com-
missioners. (Salt Lake City, Utah. July 11-13, 1973.) Western Pro-
ceedings 53: 492-514.
Gorham, F. P. 1901. The gas-bubble disease of fish and its cause. Bull.
U. S. Fish Comm. 19(1899): 33-37.
17
-------
Lorz, H. W., and B. P. McPherson. 1976. Effects of copper or zinc in fresh
water on the adaptation to sea water and ATPase activity, and the
effects of copper on migratory disposition of coho salmon (Onaorhynahus
kisutch). J. Fish. Res. Bd. Canada 33(9): 2023-2030.
MacDonald, J. R., and R. A. Hyatt. 1973. Supersaturation of nitrogen in
water during passage through hydroelectric turbines at Mactaquac Dam.
J, Fish. Res. Bd. Canada 30(9): 1392-1394.
Marcello, R. A., M. H, Krabach, and S. F. Bartlett. 1975. Evaluation of
alternative solutions to gas bubble disease mortality of menhaden at
Pilgrim Nuclear Power Station. YAEC-1087. Environ. Sci. Group, Yankee
Atomic Electric Co., Westboro, Mass, xii + [139] p.
May, B., and J. Huston. 1975. Kootenai River Fisheries Investigations,
Phase 2, Part 1. Final job report (July 1, 1972 - July 30, 1975).
U. S. Army Corps of Engin. Tract DACW 67-73-C-0003. Fish. Div.,
Montana Dept. Fish & Game, Libby. Pp. 1-28.
Meekin, T. K., and B. K. Turner. 1974. Tolerance of salmonid eggs, juve-
niles $ and squawfish to supersaturated nitrogen. Washington Dept.
Fish. Tech. Rept. 12: 78-95,
Nebeker, A. V. 1976. Survival of Daphn-ia, crayfish, and stoneflies in air-
supersaturated water. J. Fish. Res. Bd. Canada 33(6): 1208-1212.
Nebeker, A. V., J. D, Andros, and D. G. Stevens. 1977a. Survival of steel-
head trout embryos and alevins in air-supersaturated water. Trans.
Am. Fish. Soc. 106. (In press.)
Nebeker, A. V., and J. R. Brett. 1976. Effects of air-supersaturated water
on survival of Pacific salmon and steel he,id imolts. Trans. Am. Fish.
Soc. 105(2): 338-342.
flebeker, A. V., A. K. Hauck, and J. Nash. 1977b. Temperature effects on
salmon and steelhead trout in air supersaturated water. J. Fish. Res.
Bd. Canada 34. (Tn press.)
Nebeker, A. V., D. G. Stevens, and R. J. Baker. 1977c. Survival of salmon
smolts in sea water after exposure to air-supersaturated water. J.
Fish. Res. Bd. Canada 34. (In press.)
Nebeker, A. V., D. G. Stevens, and J. R. Brett. 1976. Effects of gas super-
saturated water on freshwater aquatic invertebrates. Pp. 51-65 TT± D.
H. Fickeisen and M. J. Schneider (eds.), Gas bubble disease. Proceed-
ings of a workshop cosponsored by Battelle Pacific Northwest Laborator-
ies and U. S. Atomic Energy Commission. CONF-741033. (Held in Richland,
Wash. Oct. 8-9, 1974.) Energy Res. & Devel. Admin., Technical Informa-
tion Center, Oak Ridge, Tenn. viii + 123 p.
18
-------
Rucker, R. R. 1972. Gas-bubble-disease of salmonids: a critical review.
Bur. Sport Fish. Wild!, Tech. Paper 58. U. S. Dept. of the Interior,
Washington, D. C. 11 p.
Schiewe, M. H., and D. D. Weber. 1976. Effects of gas bubble disease on
lateral line function in juvenile steelhead trout. Pp. 89-92 jn_ D. H.
Fickeisen and M. J. Schneider (eds.), Gas bubble disease. Proceedings
of a workshop cosponsored by Battelle Pacific Northwest Laboratories and
U. S. Atomic Energy Commission. CONF-741033. (Held in Richland, Wash.
Oct. 8-9, 1974.) Energy Res. & Devel. Admin., Technical Information
Center, Oak Ridge, Tenn. viii + 123 p.
Stroud, R. K., G. R. Bouck, and A. V. Nebeker. 1975. Pathology of acute and
chronic exposure of salmonid fishes to supersaturated water. Pp. 435-
449 jji Chemistry and physics of aqueous gas solutions. The Electro-
chemical Society, Princeton, N. J.
U. S. Environmental Protection Agency. 1976. Quality criteria for water.
EPA-440/9-76-023. U. S. Environmental Protection Agency, Washington,
D. C. ix + 501 p.
Van Slyke, D. D., and J. M. Neill. 1924. The determination of gases in
blood and other solutions by vacuum extraction and manometric measure-
ment. I. J. Biol. Chem. LXI(2): 523-573.
Weitkamp, D. E., and M. Katz. 1973. Resource and literature review: dis-
solved gas supersaturation and gas bubble disease. Seattle Marine
Laboratories, Seattle, Wash, i + 60 p.
Weitkamp, D. E., and M. Katz. 1975. Resource and literature review: dis-
solved gas supersaturation and gas bubble disease, 1975. Document
75-0815-042FR, Environ, Sciences Sect., Parametrix, Bellevue, Wash.
70 p.
-------
EFFECTS OF KEPONE® ON ESTUARINE ORGANISMS1
D. J. Hansen, D. R. Nimmo, S. C. Schimmel,
G. E. Walsh, and A. J. Wilson, Jr.
U. S. Environmental Protection Agency
Environmental Research Laboratory
Gulf Breeze, Florida 32561
ABSTRACT
Laboratory toxicity tests were conducted to deter-
mine the effects and accumulations of Kepone in estuarine
algae, mollusks, crustaceans, and fishes. Nominal Kepone
concentrations calculated to decrease algal growth by 50%
in static bioassays lasting seven days were: 350 pgA,
QOQGum sp.; 580 ug/£, Uunaliella tertiolecta', 600
Nitzsckia sp.; and 600 yg/&, Thalassiosira
pseudonana. Measured Kepone concentrations calculated
to cause 50% mortality in flowing-seawater toxicity
tests lasting 96 hours were: 10 ug/£ for the mysid
shrimp (Mysidopsis bakia); 120 pg/£ for the grass shrimp
(Palaemonetes pugio}\ >210 yg/& for the blue crab
(Callinectes sapidus}; 70 yg/£ for the sheepshead minnow
(Cyprinodon variegatus); and 6.6 yg/£ for the spot
(Leiostomus xanthurus). Bioconcentration factors (con-
centration in whole animals divided by concentration
measured in water) in these tests were greatest for
fishes (950 to 1,900) and less for grass shrimp (420 to
930).
Survival, growth, and reproduction of mysids and
sheepshead minnows were decreased in chronic bioassays
lasting 14 to 64 days. Growth of mysids and sheepshead
minnows was reduced by exposure to 0.07 yg/£ and 0.08
ug/& respectively. Bioconcentration factors for sheeps-
head minnows in the chronic bioassay averaged 5,200
K
v-j/Registered trademark, Allied Chemical Corp., 40 Rector St., New York,
10006. Kepone was purchased from Chem Service, West Chester, PA, as
99% pure. Our analyses indicated 88% purity.
Contribution No. 311, Environmental Research Laboratory, Gulf Breeze.
20
-------
(range, 3,100-7,000) for adults exposed for 28 days and
7,200 (3,600-20,000) for juveniles exposed for 36 days.
The chronic toxicity and bioconcentration potential of
Kepone are more important factors than its acute
toxicity in laboratory evaluations of environmental
hazard. Therefore, these factors should be considered
when attempting to assess present impacts and to limit
future impacts of this insecticide on the aquatic
environment.
INTRODUCTION
Kepone (decachlorooctahydro-1,3,4-metheno-2H-cylobuta [cd] pentalene
2-one) is an insecticide that was manufactured and formulated in the United
States to control ants, cockroaches, and insect pests of potatoes and
bananas. Kepone is toxic to birds and mammals, including man (Jaeger 1976),
and acutely toxic to some estuarine organisms (Butler 1963). Recent contam-
ination of water, sediment, and biota in freshwater and estuarine portions of
the James River, Virginia, has stimulated concern about this chemical's
hazard to aquatic biota (Hansen et al. 1976). This concern was based on (1)
the continued occurrence of Kepone in many finfishes and shellfishes in
amounts that forced closure of fishing because of potential human health
hazard, and (2) laboratory studies which showed that Kepone is highly bio-
accumulative and toxic to estuarine organisms, particularly in chronic
exposures. This paper describes the results of these laboratory toxicity
tests with estuarine algae, oysters, crustaceans, and fishes and chronic
tests with a crustacean and a fish.
EXPERIMENTAL PROCEDURES
Acute Toxicity
Algae: The unicellular algae Chlorocoacum sp., Dunaliella, teptioleata,
Nitzschia. sp., and Thalassiosira pseudonana were exposed to Kepone for seven
days to determine its effect on growth (Walsh et al. 1977). Algae were
cultured in 25 or 50 ml of growth media and artificial seawater of 30 °/oo
salinity and a temperature of 20 C (Hollister et al. 1975). Kepone, in 0.1
ml acetone, was added to culture media, and 0.1 ml of acetone was added to
control cultures. Photoperiod consisted of 12 hours dark and 12 hours of
5000 lux illumination. Effect on growth was determined by electrophoto-
metrically measuring optical density. Also, algae grown for 6 days in media
and then exposed to 100 ^g/£ Kepone for 24 hours were analyzed for Kepone
content.
Oysters: The acute toxicity of Kepone to embryos of the eastern oyster
(CrassosLrea virginica) was determined by measuring its effect on development
f>^
-------
of fully-shelled, straight-hinged veligers in a 48-hour static exposure .
Methods used were those of Woelke (1972) and U. S. EPA (1975). Test contain-
ers were l-£ glass jars that contained 900 ma of 20 C, 20 °/oo salinity sea-
water and 25,000 ± 1,000 oyster embryos. All test concentrations were
triplicated. The number of normal and abnormal embryos were counted micro-
scopically in a Sedgewick-Rafter cell at the end of 48 hours of exposure to
Kepone.
Crustaceans and Fishes: The acute toxicity of Kepone to grass shrimp
(Palaemonetes pugio), blue crabs (Callineet&s sapidus], sheepshead minnows
(Cyprinodon variegatus], and spot (Leiostomus xanthurus] was determined in
96-hour flow-through toxicity tests (Schimmel and Wilson 1977). Acclimation
and testing procedures were compatible with those of Standard Methods (APHA
1971). Test animals were caught locally and 20 were placed in each 18£
aquarium. Water flow to each aquarium was 68 £/hour. Stock solutions of
Kepone in acetone were metered into experimental aquaria at the rate of 60
m/hour. Control aquaria received 60 nu of acetone/hour. At the end of the
experiment, surviving animals were chemically analyzed for Kepone content.
The acute toxicity of Kepone to mysids (Mysidopsis bahia) was determined
by using intermittent flows of water from a diluter (Mount and Brungs 1967)
or continuous flow of water from a siphon and Kepone from an infusion pump
(Bahner et al. 1975). Thirty-two 48-hour-old juvenile mysids were placed in
chambers (4 mysids per chamber) in each test aquarium. Chambers consisted of
glass petri dishes to which a 15 cm tall cylinder of 210u mesh nylon screen
was glued. Water in the chambers was renewed by a self-starting siphon which
nearly emptied and then filled each aquarium at about 25 min intervals.
Chronic Toxicity
Mysidopsis bahia: The chronic toxicity of Kepone to this mysid was
determined in 19-day exposures that began with 48-hour-old juveniles. (Nimmo
et al., in press). The time permitted production of several broods for
assessment of reproductive success and survival of progeny. Exposure condi-
tions, apparatus, and number of mysids per concentration were identical to
those of the acute toxicity tests. Three tests were conducted: One to
assess effects on survival and reproduction, and two at lower concentrations
to determine effects on growth. Data from the two growth experiments were
pooled for statistical analysis.
Cyprinodon variegatus : The chronic toxicity of Kepone to sheepshead
minnows was determined in a 64-day flow-through bioassay—exposure of adults
for 28 days followed by a 36-day exposure of their progeny (Hansen et al.
1977). We delivered Kepone, 0.0088 y£ of the solvent triethylene glycol,
and 1.5£ of filtered 30 C seawater (average salinity, 15 °/oo; range, 8-26
°/oo) to each 702, aquariun during each of 440 daily cycles of the dosing
apparatus of Schimmel et al. (1974). Seawater and solvent were delivered to
the control aquarium. Thirty-two adult females and 32 adult males were
This research was performed under an EPA contract by Tom Heitmuller,
Bionomics-EG&G, Inc. Marine Research Laboratory, Pensacola, Florida 32507.
22
-------
exposed to each concentration of Kepone for 28 days. Egg production was
enhanced using injections of 50 I.U. of human choriom'c gonadotrophic hormone
on exposure day 25 and 27 (Schimmel et al. 1974). Eggs were fertilized on
day 28 and placed in chambers (glass petri dishes with 9-cm tall cylinders of
450y nylon mesh). Twenty embryos were used in each chamber. Embryos from
control fish were placed in four chambers in the control aquaria and in four
chambers in each of the six aquaria receiving Kepone. Embryos from fish in
each of the six aquaria receiving Kepone were placed in four chambers in that
aquarium and in four chambers in the control aquarium. Water in the chambers
was exchanged by the action of a self-starting siphon in each aquarium that
caused water levels to fluctuate 5 cm about 40 times per day. In the 36-day
exposure to determine Kepone's effect on survival and growth of progeny,
embryos hatched and fry grew until they were juvenile fish. Kepone content
of adult fish, their eggs, and juvenile fish was determined.
STATISTICAL ANALYSES
Probit analyses of growth and mortality data were used to determine
EC50's and LC50's. Growth data for M. bahia were subjected to analysis of
variance (a = 0.05) and for C. vanegatus, analysis of covariance and Newman-
Kuels tests (a = 0.01) was used.
CHEMICAL ANALYSES
Water from acute and chronic tests with crustaceans and fishes, and
organisms surviving these tests, were analyzed by gas chromatography.
Methods of extraction concentration, cleanup, and quantification were des-
cribed by Schimmel and Wilson (1977).
RESULTS AND DISCUSSION
Acute Toxicity
Algae: Growth of marine unicellular algae was reduced by exposure to
Kepone in static tests (Table 1). Chlorocoociw was the most sensitive of
the four algae tested with a 7-day EC50 of 350 pg/£. The three less sensi-
tive species responded similarly to Kepone with overlapping confidence limits
for EC50's. Algae exposed to 100 yg Kepone/£ of media accumulated the
chemical with Chlorocoacwm containing 0.80 pg/g; D. tertioleata, 0.23 ug/g;
Nitzschta, 0.41 ug/g; and T. pseudonana, 0.52 pg/g. Butler (1963) reported
that when estuarine phytoplankton were exposed to 1,000 ug/fc carbon fixation
was reduced by 95%.
Oysters: The 48-hr EC50 for oyster larvae in static tests was less than
those of algae (Table,.!). The EC50, calculated using nominal water concen-
trations, was 66 ug/£^. Embryos from 56 yg/£ were fully shelled and straight-
hinged but appeared smaller than those from controls. The percentage of nor-
mal embryos in 65 \ig/Si was 32 percent and in 87 yg/£ it was Q%. The concen-
tration of Kepone calculated to reduce shell deposition of juvenile eastern
oysters by 50% in a 96-hour flowing water bioassay was 38 pg/£. in water of
14 C and 11 pg/£ in water of 31 C (Butler 1963)'.
-------
•TABLE 1. ACUTE TOXICITY OF KEPONE TO ESTUARINE ORGANISMS. ALGAL AND MOLLUSK
TOXICITY TESTS WERE STATIC AND ESTIMATED NOMINAL CONCENTRATIONS
REDUCING GROWTH OF ALGAE AND EMBRYONIC DEVELOPMENT OF OYSTERS BY
50% (EC50). TOXICITY TESTS WITH CRUSTACEANS AND FISHES WERE FLOW-
THROUGHS THAT ESTIMATED THE MEASURED CONCENTRATION IN WATER LETHAL
TO 50% (LC50). NINETY-FIVE % CONFIDENCE LIMITS ARE IN PARENTHESES.
Organisms
Temperature, Salinity,
C °/oo
Mollusk
Crosses trea virginiea
Crustaceans
Callineetes sapidus
Mysidopsis bchia
Palaemonetes pugio
Fishes
Cyprinodon uam-egatus
Leiostomus xanthurus
20
19
26
20
1 8
25
21
20
13
16
15
18
Exposure
Duration,
Days
EC50/LC50
4
4
4
4
4
>210
10
120
Algae
ChloroGoooum sp.
Dunaliella teTtiolesta
Nitzschia sp.
Thalassiosipa. pseudonana
20
20
20
20
30
30
30
30
7
7
7
7
350
580
600
600
(270-400)
(510-640)
(530-660)
(500-700)
66 (60-74)
(8.1-12)
(100-170)
70 (56-99)
6.6 (5.3-8.8)
Crustaceans and Fishes: Kepone, at the concentrations tested, was
acutely toxic to mysids (Nimmo et al. 1977), grass shrimp, sheepshead min-
nows, and spot, but not to blue crabs (Schimmel and Wilson 1977) (Table 1).
Spot and mysids were the more sensitive species with 96-hour LC50 values of
6.6 and 10 yg/£. Crabs exposed to as much as 210 yg Kepone/£ suffered no
significant mortality. Symptoms of acute Kepone poisoning in fishes included
lethargy, loss of equilibrium, and darkened coloration on the posterior
portion of the body, occasionally only in one quadrant. Crustaceans became
lethargic before death but exhibited no color change. Butler (1963) reported
48-hour LC50 or EC50 values (based on nominal concentrations) for other
estuarine organisms were: brown shrimp (Penaeus aztecus] , 85 yg/fc; and white
mullet (Mugil aurema) , 55
Kepone was bioconcentrated from water by all four species we exposed for
96 hours. Bioconcentration factors (concentration in tissue divided by
24
-------
measured Kepone in water) for fishes were similar (950 to 1,900). Bioconcen-
tration factors for grass shrimp ranged from 420 to 930 and for blue crabs,
6 to 10.
CHRONIC TOXICITY
Mysidopsie bania: Exposure of this mysid to Kepone for 19 days in the
first experiment decreased its survival and reduced the number of young pro-
duced per female (Table 2) (Nimmo et al. 1977). At the highest concentration
(8.7 pg/fc) all mysids were dead within the first two days. At lesser concen-
trations (1.6 and 4.4 yg/&) mortality continued throughout the test. Eighty-
four % of the mysids survived exposure to 0.39 ug Kepone/£ water and 91%
survived in control aquaria. In addition, natural reproduction was affected.
Average number of young mysids produced per female was 15 in control, 9 in
0.39 yg/i, and 0 in 1.6 yg/£. Mysids that survived throughout the Kepone
exposure appeared smaller than those in control aquaria, therefore, two
additional experiments were conducted to measure Kepone's effect on growth.
TABLE 2. EFFECT OF KEPONE ON THE SURVIVAL OF MYSIDOPSIS BAHIA AND ON AVERAGE
NUMBER OF YOUNG PER FEMALE IN A 19-DAY FLOW-THROUGH TOXICITY TEST.
Average Measured Percentage Number of Young
Kepone Concentration Survival per Female
Control
0.39
1.6
4.4
8.7
91
84
50
3
0
15.3
8.9*
0
--
—
*Statistically significant at a = 0.05 using 2 sample t-test.
In these experiments, the average length (tip of carapace to end of
uropod) of mysids exposed to Kepone was decreased (Nimmo et al. 1977).
Females exposed to 0.072, 0.11, 0.23, or 0.41 vg/a were significantly shorter
than were control mysids; average length was 8.2 mm for exposed versus 8.6 mm
for control female mysids. Unexposed and exposed males, however, were of
similar average lengths, 7.7 to 8.0 mm.
Cyprinodcn variegatus: Kepone was toxic to adult sheepshead minnows
exposed for 28 days (Table 3). Symptoms of poisoning included: scoliosis,
darkening of the body posterior to the dorsal fin, hemorrhaging near the
brain, edema, fin-rot, uncoordinated swimming, and cessation of feeding.
Symptoms were first observed on day 1 in 24 yg/£, 2 in 7.8 pg/£, 3 in 1.9
yg/&, and day 11 in 0.8 ug/£. Mortalities began 5 to 8 days after onset of
symptoms.
-------
TABLE 3. EFFECT OF KEPONE ON AND ACCUMULATION OF KEPONE BY ADULT SHEEPSHEAD
MINNOWS EXPOSED FOR 28 DAYS.
Average Measured
Exposure Concentration, pg/£
ND*
0.05
0.16
0.80
1.9
7.8
24.
Percentage
Mortal i ty
5
5
0
22
80
100
100
Whole Body
Concentration,
ND
0.30
0.78
3.0
12.
M9/9
*ND = Kepone not detected in control water (<0.02 yg/2.) nor in control fish
(<0.02
Kepone affected the progeny of 28 day exposed adults. In Kepone-free
water, mortality of embryos from adults exposed to 0.05-0.8 \igfa was similar
to that of embryos from unexposed adults (range, 6-12 percent}. However, in
Kepone-free seawater, 25% of the embryos from fish exposed to 1.9 yg of
Kepone/£ died; abnormal development of 13 of these 20 embryos preceded
mortality.
Kepone in water affected progeny of exposed parents to a greater extent
than progeny of unexposed parents (Table 4). Some embryos exposed to 2.0
ijg/£ developed abnormally and fry had more pronounced symptoms and they
began to die 10 days earlier when parental fish had been exposed to 1.9 vg/i
than was observed in progeny from unexposed parents.
Kepone also affected growth of sheepshead minnows in the 36-day exposure
of progeny (Figure 1). The average standard length of juveniles exposed to
all Kepone concentrations was less than that of unexposed control juveniles.
Lengths decreased in direct proportion to increasing Kepone concentrations in
water and were generally not influenced by parental exposure. A similar
decrease was also noted in weights, but because juveniles exposed to 0.72,
2.0, or 6.6 yg/& were edematous, they weighed more than unexposed juveniles
of similar lengths.
Kepone was bioconcentrated by sheepshead minnow adults and their progeny
exposed to the insecticide in water. Kepone was bioconcentrated in adult
fish in direct proportion to concentration in exposure water (Table 3). Con-
centration factors averaged 5,200 (range, 3,100-7,000). Kepone concentra-
tions in females and their eggs were similar and were 1.3 times greater than
amounts in males. Concentrations of Kepone in juvenile fish, at the end of
the 36-day progeny exposure, increased with increased concentration of Kepone
in water (Table 4). Prior exposure of parental fish apparently did not
26
-------
affect final Kepone concentration in progeny. Concentration factors for
juvenile fish averaged 7,200 (range, 3,600-20,000) and increased with decrease
in concentration of exposure.
TABLE 4. MORTALITY IN PROGENY OF ADULT SHEEPSHEAD MINNOWS THAT WERE EXPOSED
TO KEPONE AND IN PROGENY OF UNEXPOSED, CONTROL FISH. NOMINAL
EXPOSURE FOR THE 28-DAY EXPOSURE OF ADULT FISH AND THE 36-DAY
EXPOSURE OF PROGENY WERE THE SAME. PROGENY EXPOSURE BEGAN WITH
EMBRYOS AND ENDED WITH JUVENILE FISH FROM THE EMBRYOS. RESIDUES
ARE CONCENTRATIONS OF KEPONE (yg/g) IN WHOLE JUVENILES, WET WEIGHT.
Measured Exposure
Concentration
Parental Fish History
Progeny of Unexposed Parents Progeny of Exposed Parents
1
Mortality
Residue
ND = not detectable, <0.02
<0.02 ug/g.
Mortal ity
Residue
Control (ND)
0.08
0.18
0.72
2.0
6.6
33.
10
22
12
28
40
40
100
ND1
1.1
1.4
2.6
7.8
22.
__
10
9
18
18
62
—
--
ND1
1.6
1.0
1.9
8.4
--
__
In our tests, Kepone was acutely toxic to, and accumulated by, estuarine
algae, mollusks, crustaceans, and fishes. Chronic toxicity tests with M.
bahia and C. variegatus revealed that Kepone affected survival, growth, and
reproduction. Effects on growth were observed at 0.001 of the 96-hour LC50.
Ac - 'imulation of Kepone was also greatest in chronic tests. Therefore,
•:.-. );ic tests should be used to assess Kepone's environmental hazard and to
M:; !:>•> decisions necessary to minimize its future impact on the aquatic envi-
• i , >nt.
27
-------
14
E
x
H
UJ
§10
cn
LU
(T
UJ
0.05*
0.16
.CONTROL
• PARENTS UNEXPOSED
• PARENTS EXPOSED
0 0.1 10 10.0
JUVENILE EXPOSURE CONCENTRATION (jig/I)
Figure 1. Average standard length of juvenile sheepshead minnows
exposed as embryos, fry, and juveniles for 36 days to 0,
0.08, 0.18, 0.72, 2.0, or 6.6 iig of Kepone/£ of water.
Parent fish in some instances also were exposed to similar
concentrations of Kepone: 0, 0.05, 0.16, 0.80, or 1.9
Concentration of Kepone in water, yg/fc, for parent fish exposed prior to
placement of their embryos in Kepone-free water.
28
-------
REFERENCES
American Public Health Association et al. 1976. Standard methods for the
examination of water and wastewater. 14th ed. Am. Public Health Assoc.,
Washington, D. C. 1193 p.
Banner, L. H., C. D. Craft, and D. R. Nimmo. 1975. A saltwater flow-through
bioassay method with controlled temperature and salinity. Prog. Fish-
Cult. 37(3): 126-129.
Butler, P. A. 1963. Commercial fisheries investigations. Pp. 11-25 ir\_ J.
L. George (ed.), Pesticide-wildlife studies: a review of Fish and
Wildlife Service investigations during 1961 and 1962. Fish and Wildl.
Serv. Circ. 167. U. S. Dept. Int., Washington, D. C. 109 p.
Hansen, D. J., L. R. Goodman, and A. J. Wilson, Jr. 1977. Kepone^-':
Chronic effects on embryo, fry, juvenile, and adult sheepshead minnows,
(Cyprinodon variegatus], Chesapeake Sci. (In press).
Hansen, D. J., A. J. Wilson, D. R. Nimmo, S. C. Schimmel, L. H. Bahner, and
R. Huggett. 1976. Kepone: hazard to aquatic organisms. Science 193
(4253): 528.
Hollister, T. A., G. E. Walsh, and J. Forester. 1975. Mirex and marine
unicellular algae: accumulation, population growth and oxygen evolution.
Bull. Environ. Contain. Toxicol. 14(6): 753-759.
Jaeger, R. J. 1976. Kepone chronology. Science 193(4248): 94.
Mount, D. I., and W. A. Brungs. 1967. A simplified dosing apparatus for
fish toxicology studies. Water Res. 1(1): 21-29.
Nimmo, D. R., L, H. Bahner, R. A. Rigby, J. M. Sheppard, and A. J. Wilson,
Jr. 1977. Mysidopsis bahia: An estuarine species suitable for life-
cycle bioassays to determine sublethal effects of a pollutant. Jhi
Proceedings Symposium on Aquatic Toxicology and Hazard Evaluation.
(Held in Memphis, Tenn. Oct. 25-26, 1976.) American Society of Testing
Materials. (In press).
Schimmel5 S. C., D. J. Hansen, and J. Forester. 1974. Effects of Aroclor'
1254 on laboratory-reared embryos and fry of sheepshead minnows
(Cyprinodon variegatus). Trans. Am. Fish. Soc. 103(3): 582-586.
Schimmel, S. C., and A. J. Wilson, Jr. 1977. Acute toxicity of Keponev-' to
four estuarine animals. Chesapeake Sci. (In press).
U. S. Environmental Protection Agency, Committee on Methods for Toxicity
Tests with Aquatic Organisms. 1975. Methods for acute toxicity tests
with fish, macroinvertebrates, and amphibians. Ecol. Res. Ser. EPA-
660/3-75-009. Nat!. Environ. Res. Cent., Off. of Res. & Devel., U. S.
Environmental Protection Agency, Corvallis, Ore. v + 61 p.
29
-------
Walsh, G. E., K. Ainsworth, and A. J. Wilson. 1977. Toxicity and uptake of
Kepone in marine unicellular algae. Chesapeake Sci. (In press).
Woelke, C. E. 1972. Development of a receiving water quality bioassay
criterion based on the 48-hour Pacific oyster (Crassostrea gigas) embryo.
Washington Dept. Fish. Tech. Rept. 9: 92 p.
30
-------
COLLAGEN METABOLISM IN FISH
EXPOSED TO ORGANIC CHEMICALS
F. L. Mayer, P. M. Mehrle and R. A. Schoettger
Fish-Pesticide Research Laboratory
Fish and Wildlife Service
U.S. Department of the Interior
Columbia, Missouri 65201
ABSTRACT
One major function of collagen is to serve as the
structural support for bones. Fish grow throughout life
and the vertebrae were assumed to enlarge and elongate
in proportion to growth. The synthesis of vertebral
collagen and hydroloproline was examined as an indicator
of growth, and as a sensitive predict of the chronic
effects of toxaphene, Aroclor 1254, the dimethyl amine
salt of 2,4-D, and di-2-ethylhexyl phthalate. Rainbow
trout (Sal mo eg irdneri), brook trout (Salvelinus fon-
tinalis), fathead minnows (Pimephales promelasJT and
channel catfish (Ictalurus punctatus) were the species
tested in chronic toxicity experiments, and collagen
was reduced by all four chemicals. Interpretation of
collagen synthesis data required information on vitamin
C distribution in liver and bone since the vitamin is
involved in the hydroxylation and detoxification of or-
ganic chemicals in liver and of collagen synthesis in
bone. Toxaphene reduced the vitamin C content of ver-
tebrae in channel catfish, but vitamin C content in the
liver remained constant or showed a slight increase.
The reduction of vitamin C in bone is thought to inhi-
bit collagen formation. Within limits, collagen syn-
thesis can be interpreted as a sensitive indicator
and predictor of fish growth.
31
-------
INTRODUCTION
Chronic toxicity studies of contaminant effects on fish are expensive,
high-risk investigations that require from 10 months to a year to conduct.
Such studies commonly include measurement of the long-term effects of a
contaminant on growth, reproduction, and survival of adults, and growth and
survival of the offspring. Consequently, there is much interest in develop-
ing alternative methodologies that provide similar information with less
effort and expense. Grant and Schoettger (1972) stated that biochemical fac-
tors in fish that can be correlated with toxicant exposures and residues
should provide a useful means of anticipating the subtle, adverse effects of
organic contaminants on fish. However, investigators have used various bio-
chemical indicators of chronic effects without establishing the significance
of such indicators to growth, reproduction, and survival. Biochemical mon-
itoring cannot rely on unsupported assumptions, since the biochemical adaptive
capacity of the fish can lead to broad erroneaous conclusions.
Growth of fish is usually evaluated by measuring weight or length; how-
ever, biochemical changes due to contaminant intoxication would occur before
reductions in growth are observed. Measurement of biochemical changes should
therefore decrease the time required for chronic toxicity determinations.
Initially, we selected vertebral collagen content and the hydroxyproline
concentration in collagen as potential indicators of growth and development in
fish. These biochemical characteristics were incorporated for evaluation into
a general chronic toxicity study of toxaphene that was conducted to establish
water quality criteria for this insecticide (Mayer et al. 1975, 1977; Mehrle
and Mayer 1975). Subsequently, our evaluations of biochemical characteristics
were extended to toxicological studies of Aroclor 1254 (polychlorinated bi-
phenyl), the dimethyl amine salt of 2,4-D (2,4-D DMA), and di-2-ethylhexyl
phthalate (DEHP). Our results are summarized in this report.
METHODS AND MATERIALS
Experimental Desjgn
Rainbow trout (Salmo gjnrdneri), brook trout (Salyelinus fontinalis),
fathead minnows (Pimephales promelas), and channel catfish (Ictalurus
punctatus) were continuously exposed to toxaphene, Aroclor 1254, 2,4-D DMA,
and DEHP in water (Table 1). The exposure systems were proportional diluters
modeled after Mount and Brungs (1967) and modified as recommended by
McAllister et al. (1972). Acetone was used as the carrier solvent for all
chemicals except 2,4-D DMA, for which distilled water was the solvent. Flow-
splitting chambers as desinged by Benoit and Puglisi (1973) were used to
thoroughly mix and divide each chemical concentration for delivery to the
exposure tanks. Artificial daylight was provided by the method of Drummond
and Dawson (1970), and water temperatures were maintained within ± 0.2 C.
Eggs and fish were maintained as recommended by Brauhn and Schoettger
(1975) before and during the studies. Studies on rainbow and brook trout and
fathead minnows were conducted according to the recommended procedures for
32
-------
Table 1. Chemicals tested against fish to determine their effects on the
collagen and hydroxyproline concentrations in vertebrae.
Chemical
Type
Use
Structure
Toxaphene Chlorinated Camphene Insecticide
Cl
FCH2
(CH3)2
x-4-10 (67-69% Cl)
Aroclor1254 Polychiorinated Biphenyl Dielectric Fluid
2, 4-D DMA Phenoxyacetic Acid Herbicide
Di-2-ethylhexyl
Phthalate
Phthalic Acid Ester Plasticizer
x+y = 3-8 (54% C I)
0
c'-O-(CsHi7}
•C-O-(CsHi7)
O
33
-------
chronic tests with brook trout and fathead minnows (U.S. Environmental Pro-
tection Agency 1972 a,b). Test procedures for channel catfish were described
by Mayer et al. (1977).
To determine the interactive effects of organochlorine contaminants
and dietary vitamin C, we continuously exposed 10-month old channel catfish
to a concentration series ranging from 37 to 475 ng/1 of toxaphene. Within
each concentration, the fish were subdivided into three groups, and each group
was fed ad libitum a modification (Mehrle et al. 1977) of the Oregon Test
Diet (National Academy of Science 1973) containing 63, 670, or 5,000 mg/kg
of vitamin C. The amount recommended by the Academy is 100 mg/kg.
The designs of the experiments were completely randomized or randomized
block (Cochran and Cox 1968). Growth and biochemical data were analyzed
statistically by analysis of variance, and treatment means were compared by
using a least significant difference test with the level of significance at
P £0.05 (Snedecor 1965). Linear regression analyses were calculated to de-
termine the relation of vitamin C distribution in liver and vertebrae to
exposure concentrations of toxaphene, and the relation of vertebral collagen
and hydroxyproline to fish weight, weight was presented as percentage of
the weight of control fish for graphical simplification.
Growth Measurements and Biochemical Analyses
The fish were weighed and biochemical determinations were made at times
scheduled for each study. In this paper, however, we have limited the data
presented to those measurements made at the end of the exposures. (The
toxaphene-vitamin C interaction study included growth and biochemical de-
terminations made after channel catfish fingerlings had been exposed to toxa-
phene for 90 and 150 days.) A summary of experimental conditions is pre-
sented in Table 2. Backbones (vertebrae) were dissected from the fish and
collagen, calcium, and phosphorus concentrations were determined; hydroxy-
proline was determined for each isolated collagen fraction. The vertebrae
were dried at 110 C for 2 h in a forced-air oven, split into two fractions,
and weighed. Collagen was isolated from one fraction by the method of
Flanagan and Nichols (1962). The isolated collagen was weighed and subjected
to hydrolysis at 115 C in 5 ml of 6 N HC1 for 16 h. Hydroxyproline was de-
termined in a 2-ml sample (Woessner 1961). The other bone fraction was sub-
jected to hydrolysis at 115 C in 3 ml of 6 N HC1 for 16 h. In this hydro-
lysate, calcium was determined by atomic absorption spectrophotometry and
phosphorus by the Fiske and Subbarow method (1925). In very young fry, only
the whole-body hydroxyproline content was analyzed. Vitamin C was determined
(Hubmann et al. 1969) on each bone and liver sample in the toxaphene-vitamin
C study with channel catfish finger!ings. Protein measurements were per-
formed according to Lowry et al. (1959) on rainbow trout fry. Fathead minnows
and channel catfish exposed to toxaphene were x-rayed to determine changes in
vertebral structures.
34
-------
Table 2. Summary of experimental conditions during continuous exposure of fish to organic chemicals.
Chemical ,
species, and
life stage
Toxaphene
Brook trout
Fry
Fathead minnow
Fry
Fry
Channel catfish
Fry
Fingerlings
Aroclor 1254
Brook trout
Fry
2,4-D DMA
Fathead minnow
Adult
Di-2-ethylhexyl phthalate
Rainbow trout
Fry
Brook trout
Adult
Fathead minnow
Fry
Chemical
concentration
39-502 ng/1
13-173 ng/1
94-727 ng/1
49-630 ng/1
37-475 ng/1
0.43-6.2 yg/1
0.20-2.0 mg/1
5.0-54 yg/1
3.7-52 yg/1
11-100 yg/1
Water
temperature
9
25
25
26
26
12
25
10
9-15
25
Age at
initiation
of exposure
Eyed eggs
40 days
10 days
Oc
10 mo
Eyed eggs
9 mo
Eyed eggs
1.5 yr
10 days
Duration of
fish exposure
(days)
90
98
150
90
90,150
118
60
90
150
127
Toxaphene-vitamin C interaction study.
Exposed 22 days before hatching.
Eggs and fry were produced by exposed parents and remained exposed.
Exposed 10 days before hatching.
-------
RESULTS AND DISCUSSION
Rationale for Monitoring Collagen
Collagen is the major fibrous protein of all vertebrates and most in-
vertebrates (Piez and Likens 1958), and in vertebrates it functions as the
organic matrix of connective tissues and bones. The collagen molecule is
unique in its amino acid content (Harrington and von Hippel 1963); the amino
acids hydroxyproline and proline together make up about one-tenth, and gly-
cine one-third, of all the amino acids in collagen. Hydroxyproline is found
only in two proteins—collagen and elastin. The contribution of elastin to
the total hydroxyproline content is negligible, since the total amount of
elastin is much smaller than that of collagen, and since the hydroxyproline
content of elastin is only one-tenth that in collagen (Green et al. 1968).
The synthesis of collagen, like that of other proteins, occurs on the ribo-
somes in fibroblasts, osteoblasts, and chondroblasts, and the hydroxylation
of proline and lysine occurs after they are incorporated into the polypeptide
protocollagen. The enzyme collagen hydroxylase (peptidyl proline hydroxylase)
begins activity during gastrulation and catalyzes hydroxylation; vitamin C,
-------
Table 3, Weight and backbone composition of fish continuously exposed to
organic chemicals.
Chemical , species,
and concentration
Toxaphene (ng/1)
Brook trout
0
39
68
139
Fathead minnow
Qd
94
205
399
727
Oe
13
25
54
97
173
Channel catfish
0
49
72
129
299
630
Aroclor 1254 (yg/1 )
Brook trout
0
0.43
0.69
1.5
3.1
6.2
2,4-D DMA (mg/1)
Fathead minnow
0
0.2
0.3
0,5
1.0
2.0
Fish weight
(g)
0.81
0.44*
0.59
0.43*
1.28
1.14*
1.01*
1.14*
1.04*
1.02
1.12
0.95
1.01
0.86*
0,79*
1.56
1.48
1.48
1.50
1.00*
1.10*
0.68
0.64
0.65
0.64
0.52
0.74
1.79
1.55
1.68
1.53
1.69
1.65
Collagen
(mg/g)a
300
250*
250*
250*
323
269*
229*
199*
224*
190
220
200
180
140*
150*
270
260
240*
240*
240*
230*
454
437
351*
435
386*
397*
456
470
462
436
410*
373*
Backbone comppsi
tion
Hydro xyproline inorg/org
{rog/g)k constituentsc
19
16*
16*
16*
31
24*
14*
23*
26*
30
29
29
24*
24*
25*
58
53
47*
51*
52*
51*
29
23*
21*
23*
23*
25*
28
28
30
34*
26
34*
0.70
1.24
1.64
1.64
0.49
0.69
1.17
1.23
1.24
0.62
0.56
0.58
0.70
0.86
0.79
0.48
0.57
0.72
0.65
0.59
0.59
0.26
0.23
0.53
0.41
0.60
0.80
0.29
0.29
0.28
0.33
0.49
0.70
-------
Table 3. (continued)
Backbone composition
Chemical , species,
and concentration
DEHP (ug/D
Rainbow trout
0
5
14
54
Brook trout
0
3.7
7.9
13
23
52
Fathead minnow
0
11
15
26
52
100
Fish weight
(g)
0.94
0.92
0.96
1.07
453
434
424
428
422
419
0.93
0.93
0.95
0.92
0.91
0.98
Collagen
(mg/g)a
175
158
123*
144*
445
378*
391*
371*
388*
381*
366
293*
292*
250*
172*
171*
Hydroxyprol
(mg/g)b
52
50
40
50
37
47*
47*
50*
45*
47*
24
30*
30*
35*
26
26
ine inorg/on
constituents'
f
i
_
-
-
0.49
0.56
0.60
0.68
0.64
0.65
0.46
0.57
0.62
0.69
0.92
0.94
Collagen in dry backbone.
Hydroxyproline in dry collagen, except in toxaphene-brook trout study
which was mg/g dry bone.
cCalcium + phosphorus * collagen in dry bone.
dFirst test (Mehrle and Mayer 1975).
eSecond test (Mayer et al. 1977).
Calcium and phosphorus not analyzed.
* \
Values significantly different from the controls (P < 0.05).
38
-------
(Mehrle and Mayer 1975), adult fish weight and vertebral concentrations of
collagen and hydroxyproline were all significantly reduced at all toxaphene
concentrations (94-727 ng/1). In the second fathead minnow study Mayer et al.
(1977) reported that the hydroxyproline concentration in backbone collagen of
adults was significantly reduced by toxaphene concentrations as low as 54
ng/1, whereas weight was significantly reduced only by concentrations of 97
and 173 ng/1; in fathead minnow offspring, however, weights were reduced by
exposures to 54-173 ng/1, and this measurement was more sensitive than hydro-
xyproline as an indicator of toxaphene effects. Growth of channel catfish
fry was not reduced by toxaphene until 30 days after the eggs hatched, but
the hydroxyproline content of eggs from exposed adults was significantly re-
duced. The effects of toxaphene on hydroxyproline occurred in concentrations
ranging from 72 to 630 ng/1, whereas effects on weight were observed only in
the 299 and 630 ng/1 exposures.
The correlation of vertebral collagen and hydroxyproline with fish weight
was relatively high in the fish exposed to toxyphene (Fig. 1). Correlation
coefficients (r) tended to be higher with collagen (r = 0.626-0.911) than
with hydroxyproline (r = 0.179-0.911). The relation of vertebral collagen and
hydroxyproline to growth was lowest among channel catfish, but the correlation
coefficient (0.651) of whole-body hydroxyproline and weight was much higher
at 15 days of age than it was at 90 days (Mayer et al. 1977). The differences
observed in time may have been caused by excessive mortality. Cumulative
mortality continued to increase in the 72- and 129-ng/l concentrations
throughout the 90-day exposure, even though mortality was statistically sig-
nificant only for fish in the 299- and 630-ng/l concentrations. The continued
mortality probably negated further decreases in growth by eliminating the more
susceptible fish. The effect of mortality was more evident in the study of
brook trout exposed to Aroclor 1254, where no significant effects on weight
were observed in fry continuously exposed for 118 days (Table 2). The cor-
relation coefficients were low for both collagen (r = 0.183) and hydroxypro-
line (r = 0.333). However, weight was significantly reduced in the 1.5 to 6.2
yg/1 concentrations after 48 days of exposure and the correlation with whole-
body hydroxyproline content was high (r = 0.824). The differences noted be-
tween 48 and 118 days were again probably due to mortality; no mortality
occurred before 48 days, but by 118 days, 21% of the fish died in the 3.1
yg/1 exposure and 50% died in the 6.2 ug/1 exposure.
No significant effects on growth were found in fish exposed to 2,4-D
DMA or DEHP (Table 3). The collagen content of bone was significantly re-
duced, but the hydroxyproline concentrations in collagen were either not
affected or were significantly increased, whereas both collagen and hydroxy-
proline were decreased in fish exposed to toxaphene and Aroclor 1254.
Correlation coefficients for collagen or hydroxyproline and fish weight
were low (r = 0.044-0.461) in all fish exposed to 2,4-D DMA and DEHP—the
exception being for brook trout exposed to DEHP (r = 0.822 and 0.786 for col-
lagen and hydroxyproline, respectively). The reason for differing biochemi-
cal responses involving collagen and hydroxyproline with different chem-
icals is not clear, but this question is explored later in our discussion
of vitamin C.
-------
Brook Trout
300 r
I 225
V • 190 • t 03X
i - 0 911
Fathead Minnows
Exposure days' —* 150 -~c 9fl
9C 100
Channel Catfish
to "-" ^
a ci
c £
L» - 193 •
r •• 0 626
^—i—
193 • GfelSX
LY - 48 0 -
r - 0 1 .'9
t ••
Weight
(% of control)
Figure 1. Relation between backbone development (vertebral collagen and
hydroxvproline) and weight (expressed as % of controls) of fish
continuously exposed to toxaphene. The vertical lines represent
one standard deviation.
40
-------
The use of collagen and hydroxyproline measurements as indicators or
predictors of effects of environmental contaminants on growth of fish shows
promise, but this approach has not been sufficiently studied. Measurements
of hydroxyproline have an advantage over those of collagen because hydroxy-
proline can be directly determined in eggs and whole fry, whereas collagen
is determined indirectly, except in fish that have backbones large enough
for analysis. The impact of toxicants on collagen and hydroxyproline met-
abolism in fish appears to be greatest during early life. Young fish are
growing rapidly and are generally more sensitive to toxicants that older fish.
Since the present results show that the hydroxyproline and collagen contents
of bone are not always directly related, both constituents should be measured
when possible to facilitate toxicological interpretation. Considerable var-
iation existed in collagen, hydroxyproline, and inorganic constituents of bone
in control fish of the same species in different studies. This variation
poses a serious problem in interpreting these characteristics as contaminant
indicators under field conditions.
Significance of Decreased Collagen jj^ Bone
Bone mineralization is accomplished by a complex mechanism involving the
accumulation of phosphorus salts, and then calcium salts by immature bone
(Nusgens et al. 1972), This mineralization process can occur independently
of the development of the collagen matrix, i.e., the organic substrate of bone
is not believed to be necessary for initiation of mineralization. All the
organic chemicals tested apparently depressed collagen synthesis and lowered
concentrations of vertebral collagen. The resulting effect of the test
chemicals on bone composition was an increase in the ratio of minerals to
collagen (Table 3). Mineralization of bone is a natural process, but the
organic chemicals studied appear to greatly enhance this process, as seen,
for example, in brook trout exposed to toxaphene (Fig. 2).
Calcium metabolism may have been affected, since its concentration in
the vertebrae of all species of fish tested (and with all chemicals) in-
creased more than could be accounted for by the concomitant decrease in
collagen. Phosphorus metabolism did not appear to be affected; its concen-
trations in bone remained relatively constant regardless of toxicant concen-
trations. Exceptions to this were increases in phosphorus similar to those
of calcium in brook trout exposed to toxaphene and in fathead minnows exposed
to 2,4-D DMA. However, further studies on mineral metabolism are needed to
determine whether calcium and phosphorus in bone are specifically affected
by organic toxicants.
The reduction of vertebral collagen has a potential debilitating effect
on fish; increased mineralization and brittleness of vertebrae weakens their
backbones. Many fathead minnows and channel catfish x-rayed after exposure
to toxaphene had broken or deformed backbones (Fig. 3,4). In catfish, por-
tions of vertebrae were missing or compressed, especially in the anterior
and posterior regions. In natural waters, the affected fish would almost
certainly be less capable of competing for available food and habitat or
avoiding predators. Investigations are needed to specifically determine
-------
Cafp/Collagen
p 2.0
- 1.5
- 1.0
->0.5
L-0
Time
(days)
139
Toxaphene Concentration
(ng/l)
Figure 2. Effect of toxaphene on the backbone composition of brook trout fry
exposed for up to 90 days after hatch. The composition is
depicted as the ratio of calcium and phosphorus concentrations to
collagen in dried vertebrae.
42
-------
\.
Figure 3 X-rays and schematics of backbones of 150-day-old fathead minnows:
Aa, control fish; Bb, fish exposed to 94 ng/1 of toxaphene.
/I-3
-------
Figure 4. X-rays and schematics of backbones of 90-day-old channel catfish:
Aa, control fish; Bb, fish exposed to 72 ng/1 of toxaphene.
44
-------
at what level alterations of organic to mineral ratios in vertebrae become
a negative factor in fish health and survival.
Role of Vitamin C_ jn_ Detoxication and Collagen Formation
Vitamin C is a cofactor in the hydroxylation of drugs and chemicals in
the liver of mammals (Axelrod et al. 1954, Levin et al. 1960, Street et al.
1971, Wagstaff and Street 1971}. It is also essential to collagen formation
by way of the hydroxylation of proline and lysine into hydroxyproline and
hydroxylysine (Barnes 1969, Barnes et al. 1970, Mussini et al. 1967,
Peterkofsky 1972). However, vitamin C is an essential and limiting dietary
nutrient in fish because fish are unable to synthesize it (Chatterjee 1973,
Wilson 1973). Inasmuch as these two hydroxylation processes may compete for
available vitamin C in fish, Mayer et al. (1977) investigated the effects of
toxaphene on the distribution of this vitamin in liver and vertebrae of
channel catfish.
Body weight, vertebral collagen, and vitamin C in liver and vertebrae
were determined for fish after 90 and 150 days of exposure to 37-475 ng/1
of toxaphene and fed diets containing 63, 670, or 5,000 mg/kg of vitamin C.
Growth was significantly reduced in fish fed the diet containing 63 mg/kg
of vitamin C and exposed to the three highest concentrations of toxaphene
(Fig. 5,6). Growth was also reduced in fish fed 670 mg/kg of vitamin C and
exposed to 475 ng/1 of toxaphene. No change in collagen concentrations of
vertebrae were observed in fish exposed for 90 days, except for those exposed
to 37 ng/1 of toxaphene and fed the lowest vitamin C diet. However, after
150 days of exposure, all concentrations of toxaphene significantly reduced
vertebral collagen levels in fish fed the diet containing 63 mg/kg of vitamin
C; the three highest toxaphene concentrations reduced collagen levels in fish
fed 670 ng/kg of vitamin C; and only the 475 ng/1 toxaphene concentrations
decreased collagen in fish fed 5,000 mg/kg of vitamin C.
The ratio of vitamin C concentration in liver to that in vertebrae in-
creased most in fish exposed to toxaphene for 90 days and fed the lowest vit-
amin C diet (Fig. 5). The slopes of the regression curves for these ratios
decreased with increasing dietary vitamin C to almost no perceptible effects
in the highest vitamin C diet, A similar trend was observed in ratios for
fish fed the medium and high vitamin C diets after 150 days exposure to
toxaphene (Fig. 6), but effects on the vitamin C content of liver and ver-
tebrae were more pronounced than at 90 days. After 150 days, vitamin C
was low in the vertebrae of all fish, including the controls, fed the diet
containing 63 ng/kg of vitamin C. This response in the controls was probably
due to the chronic effects of the low vitamin C diet itself.
The exposure of fish to an organochlorine contaminant, such as toxaphene,
may markedly reduce the amount of collagen in the vertebrae, possibly because
the use of vitamin C by the liver in hydroxylative detoxication mechanisms is
increased, as indicated by induction of liver enzyme activity (Mayer et al.
1977). Vitamin C in vertebrae was reduced as much as 50%, and this reduction
in bone probably inhibits the formation of hydroxyproline and hydroyxylysine
from proline and lysine, which in turn reduces collagen formation.
-------
V lamln C m Oiet
63 rog/
-------
673 ragfltg
0 37 66 ^0$ 218 475
0 37 68 13S 2<8 475
0 37 6$ I0e 218
Q 37 68 106 218
0 3? 58 IDS 218 475
8 3? M 108 21) 475
250 500
250 SOO
Z50 SOO
Figure 6. Toxaphene-vitanrin C interaction effects on growth, vertebral
collagen, and vitamin C distribution in liver and vertebrae of
channel catfish finger!ings continuously exposed to toxaphene for
150 days. Shaded areas indicate values significantly different
(P < 0.05) from the controls, and vitamin C concentrations in
liver and vertebrae are expressed as ratios.
-------
In addition, the increased use of vitamin C in liver detoxication pro-
cesses may have a direct adverse effect on growth and development of fish.
Dieter (1968) reported that vitamin C stimulates the conversion of folic acid
to the metabolically active folinic acid, and the foTic acids and vitamin B,2
are necessary for growth of many higher animal species, especially during
embryogenesis where tissue development is rapid (Cantarow and Schepartz 1962).
Also, nutrient utilization for nucleic acid synthesis was reduced in animals
deficient in folic acid (Huennekens and Osborn 1959), and Dieter (1968) hypo-
thesized that vitamin C might function during early developmental processes
by indirectly influencing the availability of required metabolic cofactors.
Although vertebral collagen was reduced in fathead minnows continuously
exposed to 2,4-D DMA and DEHP, and in brook trout and rainbow trout exposed
to DEHP, this inhibition of collagen synthesis does not appear to involve the
same mode of action as that of toxaphene and Aroclor 1254. Also, changes in
collagen concentration and fish weight were poorly correlated in fish exposed
to 2,4-D DMA or DEHP (Table 3). In contrast, the hydroxyproline content of
collagen tended to increase in fish exposed to 2,4-D DMA and DEHP, whereas it
decreased in fish exposed to toxaphene and Aroclor 1254. The increase of
hydroxyproline may have been caused by increased hydroxylation of proline in
collagen, incomplete catabolism of collagen, or to some other factor. How-
ever, we did observe an apparent increase in catabolism of total body pro-
teins in rainbow trout exposed to DEHP (Fig. 7). Protein concentrations in
whole fish were significantly decreased and the amount of hydroxyproline in
relation to protein content of whole fish increased at 24 days of exposure.
After 60 days of exposure to DEHP, vertebral collagen decreased from 187 to
125 mg/g and the amoung of hydroxyproline in collagen increased from 33 to 38
mg/g as DEHP concentrations increased. The earlier differences of hydroxy-
proline in collagen had disappeared by 90 days (Table 3). Similar effects on
protein metabolism may occur in fish exposed to 2,4-D DMA, and may in part
explain the difference in responses observed among the various chemicals
tested.
CONCLUSIONS
Biochemical characteristics such as collagen and hydroxyproline concen-
trations in bone can be used (within limits) as indicators or predictors of
growth in fishes exposed to organochlorine contaminants. Measurements of
those variables may shorten chronic toxicity tests. Although growth can be
directly related to collagen and hydroxyproline metabolism in fishes, the
mechanism by which growth is reduced is not known. Other biochemical proc-
cesses requiring vitamin C may also be affected when large amounts of the
vitamin are used by the liver in detoxification of organochlorine contami-
nants through microsomal hydroxylative enzymes. Chemicals such as 2,4-D
DMA and DEHP can also cause a reduction in vertebral collagen without a re-
duction in growth, at least within the limitations of these studies. The
manner in which 2,4-D DMA and DEHP is metabolized and affects fish may be an
important consideration in defining the differences observed between these
chemicals and organochlorine chemicals such as toxaphene and Aroclor 1254.
However, the reduction of vertebral collagen can be a debilitating factor in
48
-------
Time of Exposure After Hatch
5 days 24 days_
300
j- 250
OT
[ i
0)
O
f 1200
c
o
£ 150
;
-
-
\
300 r
250
200
150
0 5 14 54
0 5 14 54
400
3 300
o
D.
C
.S ^ 200
O ^
Q. "~
X
0
1, 100
n
/;-; ;,.'.
'.•;';'.':-:
400
300
200
100
0
• ..•••'
14 54
0
14 54
Di-2-ethylhexyl Pnthalate Concentration
(M9/D
Figure 7. Di-2-ethylhexyl phthalate effects on the protein end riydroxypro-
line content of rainbow trout fry continuously exposed, for 5 and 24
days after hatch. Shaded areas indicate values significantly
different (P < 0.05) from the controls.
-------
itself by increasing the probability of structurally weakened vertebral co-
lumns, and the biochemical processes related to that condition are useful in
toxicological evaluations of organic chemicals on fish growth and development,
ACKNOWLEDGMENT
This research was sponsored in part by the United States Environmental
Protection Agency through Contract No. EPA-IAG-0153(D) and EPA-IAG-141(D).
The Aroclor 1254 data were supplied by W. L. Mauck, and Becky Turk prepared
the illustrations.
50
-------
REFERENCES
Axel rod, J., S. Udenfriend, and B. B. Brodie. 1954. Ascorbic acid in aroma-
tic hydroxylation. III. Effect of ascorbic acid on hydroxylation of
acetanilide, aniline and antipyrine in vivo, J. Pharmacol. Exp.
Therapeut. 3(2):176-181.
Barnes, M. 0. 1969. Ascorbic acid and the biosynthesis of collagen and
elastin. Pages 86-98 j_n_ J. C. Somogyi and E. Kodicek, ed. Nutritional
aspects of the development of bone and connective tissue. S. Karger
AG, Basel, Switzerland.
Barnes, M. J., B. 0. Constable, L. F. Morton, and E. Kocidek. 1970. Studies
i_n_ vivo on the biosynthesis of collagen and elastin in ascorbic acid-
deficient guinea pigs. Biochem. 0. 119(3):575-585
Benoit, D. A., and F. A. Puglisi. 1973. A simplified flow-splitting chamber
and siphon for proprotional diluters. Water Res. 7:1915-1916.
Brauhn, J. L., and R. A. Schoettger. 1975. Acquisition and culture of re-
search fish: Rainbow trout, fathead minnows, channel catfish, and blue-
gills. Ecol. Res. Ser. No. EPA 660/3-75-011. U. S. Environmental
Protection Agency, Corvallis, Oregon. 45 pp.
Cantarow, A., and B. Schepartz. 1962. Biochemistry. W. B. Saunders Co.,
Philadelphia, Pa. 938 pp.
Chatterjee, I. B. 1973. Evolution and the biosynthesis of ascorbic acid.
Science 182:1271-1272.
Cochran, W. G., and G. M. Cox. 1968. Experimental designs. John Wiley
& Sons, Inc., New York. 617 pp.
Dieter, M. P. 1968. The influence of adrenal and testicular steroid hormones
on the intermediary metabolism and development of chicken lymphoid
organs. Ph.D. Diss. Univ. of Missouri, Columbia, Mo. 114 pp.
Drummond, R. A., and W. F. Dawson. 1970. An inexpensive method for simulat-
ing diel patterns of lighting in the laboratory. Trans. Am. Fish. Soc.
99(2):434-435.
Fiske, C. H., and Y. Subbarow. 1925. The colorimetric determination of
phosphorus. J. Biol. Chem. 66:375-400.
Flanegen, B., and G, Nichols. 1962. Metabolic studies of bone in vitro.
IV. Collagen biosynthesis by surviving bone fragments in vitro.
J. Biol. Chem. 237(12):3686-3692.
Grant, B. F., and R. A. Schoettger. 1972. The impact of organochlorine
contaminants on physiologic function in fish. Proc. Tech. Sessions
New York Annu. Meet. Inst. Environ. Sci. 18:245-250.
-------
Green, H. B., B. Goldberg, M. Schwartz, and D. D. Brown. 1968. The
synthesis of collagen during development of Xenophus laevis.
Dev. Biol. 18(4):391-400.
Harrington, W. F., and P. H. von Hipnel. 1961. The structure of collagen and
gelatin. Pages 1-138 jn_C. B. Afinsen, Jr., M. L. Anson, K. Bailey,
and J. T. Edsall, ed. Advances in protein chemistry - Vol. 16.
Academic Press, Inc., New York,
Hubmann, B., D. Monnier, and M. Roth. 1969. A rapid and precise method for
the determination of ascorbic acid; applied to the measurement of
blood plasma. Clin. Chem. Acta 25(1):161-166.
Huennekens, F. M., and M. J. Osborn. 1959. Folic acid coen2ymes and one-
carbon metabolism. Advance. Enzymol. 21:369-446.
Levin, E. Y., B. Levenberg, and S. Kaufman. 1960. The enzymatic conversion
of 3,4-dihydroxyphenylethylamine to norepinephrine. J. Biol. Chem.
235(7):2080-2086.
Lowry, 0. M., N. J. Rosebrough, A. L. Farr, and R. F. Randall. 1951.
Protein measurement with the Folin phenol reagent. 0. Biol. Chem.
193(1):265-275.
Mayer, F. L., P. M. Mehrle, and L. P. Crutcher. 1977. Interactions of toxa-
phene and vitamin C in channel catfish. Proc. Am. Fish. Soc. 2nd Bi-
ennial Fish Health Sect. Workshop. (In press).
Mayer, F. L., P. M. Mehrle, and W. P. Dwyer. 1975. Toxaphene effects on
reproduction, growth, and mortality of brook trout. Ecol. Res. Ser. No.
EPA-600/3-75-013. U. S. Environmental Protection Agency, Duluth, Minn.
51 pp.
Mayer, F. L., P. M. Mehrle, and W. P. Dwyer. 1977. Toxaphene: Chronic
toxicity to fathead minnows and channel catfish. Ecol. Res. Ser.
U. S. Environmental Protection Agency, Duluth, Minn. (In press).
McAllister, W. A., W. L. Mauck, and F. L. Mayer. 1972. A simplified device
for metering chemicals in intermittent-flow bioassays. Trans. Am. Fish.
Soc. 101(3):555-557.
Mehrle, P. M., and F. L. Mayer. 1975. Toxaphene effects of growth and bone
composition of fathead minnows, Pimephales promelas. J. Fish. Res.
Board Ca. 32(5):593-598.
Mehrle, P. M., F. L. Mayer, and W. W. Johnson. 1977. Diet quality in fish
toxicology: Effects on acute and chronic toxicity. Proc. ASTM Symp.
Aquatic Toxicol. Hazard Evaluation. (In press).
Mount, D. I., and W. A. Brungs. 1967. A simplified dosing apparatus for
fish toxicology studies. Water Res. 1(1):21-29.
52
-------
Mussini, E., J. J. Hutton, and S. Udenfriend. 1967. Collagen proline
hydroxylase in wound healing, granuloma formation, scurvy, and growth.
Science 157:927-929.
National Academy of Sciences. 1973. Nutrient requirements of domestic
animals. Nutrient requirements of trout, salmon, and catfish. No. 11
Natl. Acad. Sci., Washington, D.C. 57 pp.
Nusgens, B., A. Chantraine, and C. M. Lapiere. 1972. The protein in the
matrix of bone. Clin. Orthopedics Relat. Res. 88:252-274.
Peterkofsky, B. 1972. The effect of ascorbic acid on collagen polypeptide
synthesis and proline hydroxylation during the growth of cultured fibro-
blasts. Arch. Bioch. and Biophys. 152:318-328.
Piez, K. A., and R. C. Likins. 1960. The nature of collagen. II.
Vertebrate collagens. Calcification of biological systems. Publ. No.
64. Am. Assoc. Advance. Sci., Washington, D.C. 420 pp.
Rollins, J. W., and R. A. Flickinger. 1972. Collagen synthesis in Xenopus
oocytes after injection of nuclear RNA of frog embryos.
Science 178:1204-1205.
Snedecor, G. W. 1965. Statistical methods. Iowa State Univ. Press.,
Ames, la. 534 pp.
Street, J. C., R. C. Baker, D. J. Wagstaff, and F. M. Urry. 1971.
Pesticide interactions in vertebrates: Effects of nutritional and
physiological variables. Proc. IUPAC Int. Congr. Pesticide Chem.
2:281-302,
U.S. Environmental Protection Agency. 1972a. Recommended bioassay procedure
for brook trout Sal veil mis, fontinalis (Mitchill) partial chronic tests.
U. S. Environmental Protection Agency, Environmental Research Laboratory,
Duluth, Minn. 12 pp.
U. S. Environmental Protection Agency. 1972b. Recommended bioassay procedure
for fathead minnow Pimephales promelas Rafinesque chronic tests.
U. S. Environmental Protection Agency, Environmental Research Laboratory,
Duluth, Minn. 13pp.
Wagstaff, D. J., and J. C. Street. 1971. Ascorbic acid deficiency and in-
duction of hepatic microsomal hydroxylative enzymes by organochlorine
pesticides. Toxicol. Appl. Pharmacol. 19:10-19.
Whitehead, R. G., and D. 6. Coward. 1969. Collagen and hydroxyproline
metabolism in malnourished children and rats. Pages 74-85 in J. C.
Somogyi and E. Kodicek, ed. Nutritional aspects of the development
of bone and connective tissue. S. Karger AG, Basel, Switzerland.
-------
Wilson, R. P.» 1973. Absence of ascorbic acid synthesis in channel catfish,
Ictalurus punctatus and blue catfish, Ictalurus furcatus. Comp.
Biochem. Physio!. 466:635-638.
Woessner, J. F. 1961. The determination of hydroxyproline in tissue and
protein samples containing small proportions of this amino acid.
Arch. Biochem. Biophys. 93(2)-.440-447.
54
-------
EFFECTS OF SHORT-TERM EXPOSURES
TO TOTAL RESIDUAL CHLORINE ON THE SURVIVAL AMD BEHAVIOR
OF LARGEMOUTH BASS (Micropterus salmoides)
G. L. Larson and D. A. Schlesinger
Department of Fisheries and Wildlife
Oregon State University
Corvallis, Oregon
97331
A contribution of the Oak Creek Laboratory of Biology. Research supported in
part by the Environmental Protection Agency, Grant No. R-802286.
ABSTRACT
Largemouth bass were subjected to short-term
exposures of total residual chlorine. Two different
time-toxicant concentration curves similar to those
of chlorinated discharges from power generation
plants were used as models for the tests. One
discharge curve (referred to as the square exposure)
was characterized by a rapid rise in toxicant con-
centration to a plateau level, followed by a rapid
decline in concentration after toxicant intro-
duction was terminated. The second discharge curve
(referred to as the spike exposure) was charac-
terized by a rapid rise to a peak toxicant con-
centration, immediately followed by a rapid decline
in toxicant concentration. Acute toxicity tests
included a comparison of the effects of square and
spike exposures, and comparative tests of the
effects of square exposures of varying frequency
and duration. Fish behavior was observed during
acute and sublethal square and spike exposures.
There were no obvious differences in acute
toxicity between the two types of exposures when
mortality (in probits) was plotted against the areas
under the time-concentration curves. The same
results were obtained in tests of one and two 90-
min. exposures,and for one 90-min. exposure and one
150-min. exposure. Thus, measurement of the areas
under the time-concentration curves are a useful
-------
means of studying effects of different kinds and
durations of exposures and different exposure
frequencies.
Bass exhibited several behavioral changes
during the acute toxicity tests. Many behavioral
responses occurred in sublethal tests of square and
spike exposures. The behavioral changes caused
by acute and sublethal exposures probably are
detrimental to the well-being and survival of the
fish in the field.
INTRODUCTION
Intermittent (recurrent) chlorination of cooling waters is a common
method employed to remove organisms from heat exchangers in power generation
plants. Only 10 percent of the power plants in the U.S. chlorinate on a
regularly programmed basis (Brungs 1976). Considerable variation exists
between power plants with regard to the duration, frequency, and amounts
of chlorine introductions. Additional differences between the discharges
from power plants include temperature, water quality (e.g. heavy metal con-
tamination), toxicant concentrations at the points of discharge into
receiving waters, and the forms of the residual chlorine. The heated
effluents from power plants are discharged into rivers, lakes, or estuaries,
but the effluents may pass through channels or ponds before discharge into
receiving waters.
The toxicity of re^Hual chlorine to aquatic organisms under conditions
of continuous exposure dt.js not provide reliable information on the potential
toxicity of intermittent exposures to residual chlorine. Extrapolation of
laboratory results to the field situation is most appropriate when aquatic
animals are exposed to short-term introductions of residual chlorine over
an adequate time.
Based upon literature on the toxicity of residual chlorine to aquatic
organisms kept under continuous exposure to constant toxicant concentrations
in laboratory experiments, considerable concern has developed regarding the
potential toxicity of short-term introductions of residual chlorine to
the organisms. The short-term exposures present a number of special problems
of analysis and comparison of laboratory and field data. Major problems
facing investigators include the effects of the fluctuating toxicant concen-
trations discharged into receiving waters, and effects of the variations in
duration and frequency of the introductions. As a means of dealing with
some of these problems, investigators have often exposed fish to constant
concentrations of residual chlorine in laboratory aquaria for short periods
of time (McLean 1973; Stober and Hanson 1974). Such tests do not mimic the
fluctuating toxicant concentrations to which fish are exposed in the field.
Other investigators have exposed fish to fluctuating toxicant concentrations
and have calculated LC50's on the bases of the mean or peak concentrations
56
-------
of the exposures (Brooks and Seegert 1977; Heath 1977). Basing LC5Q's on
mean concentrations probably is a useful method for comparing different
experiments when the durations and frequencies of the exposures are known.
Acute toxicity data based on peak concentrations, i.e. when the toxicant
attains a peak concentration and then declines in concentration rapidly, are
not comparable, however, unless the time-toxicant relationships are identical.
It appears that comparisons made on the basis of mean exposure concen-
trations to fish for varying periods would be useful in developing an under-
standing of the toxicity to aquatic organisms in intermittent discharges of
chlorine. However, the variable characteristics of intermittent chlorinated
discharges and the array of possible field conditions to which aquatic
organisms could be exposed would undoubtedly lead to nearly endless experi-
mentation.
One way to deal with the complexity of the field conditions is to compare
the exposures in acute toxicity tests on the basis of the areas under the
time-concentration curves. We assumed toxicity was constant for a given area
under the time-concentration curves without regard for the shape of the
curve, the exposure frequency (assuming no recovery between exposures), or
the duration of the exposures. The first objective of our work was to explore
the utility of the area concept as it applied to comparing or predicting the
toxicities of intermittently chlorinated power generation plant discharges.
Secondarily, we observed the behavioral changes of fish during exposures
to short-term introductions of residual chlorine in laboratory aquaria. This
work was initiated because several investigators during fieldstudies noticed
major behavioral changes in fish subjected to short-term exposures to chlorine
(e.g. Basch and Truchan 1976). Erratic behavior might result in physical
damage to fish by reducing the ability to avoid obstacles or increasing the
susceptibility to predation.
METHODS
Largemouth bass were collected from a farm pond near COP/a'Mis, Oregon,
in early June, 1975. Fish were acclimated to the laboratory conditions for
at least one month prior to testing and were fed Oregon Moist Pellets daily.
Feeding was discontinued one day before the tests. The fish were maintained
under the photoperiod regime for this region.
Acute toxicity bioassays were carried out with standard 45-1. glass
aquaria, initially containing 40 1. of water. Dilution water and chlorine
solutions were introduced to the aquaria through two PVC manifolds, each 3.81
cm in diameter. Rapid changes of the chlorine concentrations in the aquaria
were achieved by reducing the water volume to 30 1. (20 cm maximum depth)
and by appropriately manipulating dilution water and toxicant flows.
Chlorine stock solutions were made in a Mariotte bottle by mixing
sodium hypochlorite and well water. The well-water supply was located at
the laboratory. Average water quality characteristics of the well water
-------
were: dissolved oxygen 7.4 mg/1, hardness 128 rng/1, total alkalinity
148 mg/1, pH 7.94, and temperature 24.3 C. Grab samples of chlorine test
solutions were analyzed using a Wallace and Tiernan amperometric titrator.
Free residual chlorine averaged 97.04± 1.23 percent* of the measurable total
residual chlorine (TRC) in the test solutions.
Each experiment was completed within one week. In each acute toxicity
test, six bass of nearly equal size were acclimated for 30 min. to the test
conditions before the toxicant was introduced. Concentrations of chlorine
were measured at 2-10 min. intervals for the duration of each exposure, when
the toxicant was present in the aquarium. In 96-hr, acute toxicity tests
fish were exposed to chlorine and then maintained in fresh water for the
duration of the test. Mortalities were recorded daily; dead fish were
removed. At the end of each test, fish were dried at 70 C. for 7 days and
weighed.
In these studies two types of time-toxicant concentration curves
(referred to as square and spike exposures) were used as models (Figure 1).
These curves represented the extremes of those found in the power plant
effluents at the points of discharge into receiving waters (G. Nelson, EPA,
personal communication). The square exposures were produced by adding
the toxicant to the aquaria at constant rates for predetermined periods.
Chlorine concentrations reached a plateau level 20 min. after initiating
the toxicant flow. With one exception the toxicant flow was terminated at
60 min. and the toxicant was completely flushed from the aquaria after
an additional 30 min. (90 min. total exposure time). Spike exposures were
characterized by a rapid rise to a peak toxicant concentration, followed
immediately by a rapid decline. In bioassays using the spike exposures both
toxicant and dilution water flows were manipulated to achieve the desired
curves. High and low spikes (relative to each other) were used in some
tests, only the low spikes in other tests. The high spike exposure peaked
in concentration 5 min. after initiation of toxicant flow, the low spike
exposure peaked at 22 min. Total exposure times were 51 min. for the high
spikes and 63 min. for the low spike exposures.
Acute toxicity tests at our laboratory have shown that fish weight
can affect the tolerance of coho salmon to residual chlorine (Larson et al.
1977). The bass used in the present work were not of uniform weight. Prelim-
inary tests were conducted in mid-July to determine if body weight affected
the tolerance of the bass to short-term exposures to chlorine. Two weight
groups were tested, one being 3.87± .15, the other 5.93+ .28 g/nsn, di y
weight. Groups of each weight class were exposed to on
-------
HIGH SPIKE
EXPOSURE
LOW SPIKE
EXPOSURE
SQUARE
EXPOSURE
40 60
TIME (min)
100
Figure 1. Examples of the time-concentration relationships of the square
exposure and the high and low spike exposures with total
residual chlorine (TRC).
59
-------
TABLE 1. MEAN DRY WEIGHTS OF BASS USED IN THE ACUTE TOXICITY EXPERIMENTS
Experiment
Mean dry
weight per fish (g) ± 1 S.D.
Body Weight
Group 1
Group 2
Square vs. Spike Exposures
Square (90 min.)
Low spike (63 min.)
High spike (51 min.)
Exposure Frequency
one 90 min.
two 90 min.
Exposure Duration
90 min.
150 min.
3.87 ± .15
5.93 ± .28
8.60 ± .20
8.50 ± .20
8.51 ± .24
6.18 + .19
6.15 ± .32
8.92 ± .36
8.92 ± .53
± 1 Standard deviation
90-min. square exposures
Square exposures
60
-------
Groups of bass were exposed to a high or low spike exposure, or to the
square exposure. The mean plateau concentrations of TRC in square exposures
ranged from 2.35 to 3.32 mg/1. The peak TRC concentrations in spike
exposures ranged from 8.21 to 11.93 (high spike exposures) and 5.73 to 9.06
mg/1 (low spike exposures).
Effects of the three types of exposures were compared on the basis of
areas under the time-concentration curves. Areas under the time-concentration
curves were measured using a compensating planimeter. The curves for all
tests were graphed using the following scale: a 10-min, exposure to 1 mg/1
TRC equalled 5.9 cur.
Exposure-frequency effects were examined by subjecting some groups of
bass to single 90-min. square exposures, while other groups were subjected to
two such exposures. In the two-exposure groups, a 2-hr, recovery period
separated the exposures.
The effect of exposure duration on survival was investigated by sub-
jecting groups of bass to either a 90-min. or a 150-min. square exposure.
For the 150-min. exposures, toxicant flow into the aquaria was terminated at
120 min.
Observations of the changes in behavior of the bass were made throughout
each acute toxicity test; the majority were made during the exposure period
and at 24-hr, intervals for the duration of the 96-hr, tests. Times for the
first occurrence of particular behavioral responses were recorded during most
exposures. After the acute toxicity studies, an experiment was conducted to
determine the behavior of individual bass exposed to sublethal concentrations
of chlorine. The experimental equipment and most of the procedures were
identical to those described above. Two groups of bass were used, one
averaging 14.63 ± .39 g, and the other 7.37 ± .62 g dry weight per fish. In
most cases two fish, one from each group, were kept together in separate
aquaria for at least 5 days prior to testing. The paired fish were then sub-
jected once to either a 90-min. square exposure or a 63-min. low spike
exposure, and then maintained in fresh water for the duration of the 96-hr.
test. Water samples for determining TRC concentrations were obtained by
siphoning test solutions from the aquaria. The range of concentrations in
square exposure tests was 140 to 2422 yg/1 (averages for the 20-min, to
60-min. plateau period), and that in spike exposure tests was 400 to 6980
yg/1 peaks. Behavioral observations were made by an observer sitting 2 m
from the test aquaria during the exposures and for 1 to 2 minutes at 24-hr.
intervals thereafter.
Standard statistical methods (Sokal and Rohlf 1969) were used to perform
regression analyses.
-------
RESULTS AND INTERPRETATIONS
Mortalities usually occurred within 24-48 hours after the exposures.
Dying fish turned over (belly up) or rested upright on the aquaria bottoms,
and had much coagulated mucus adhering to the gills at the ends of the
exposures.
The relationships between mortalities (in probits) and the areas under
the square, low spike, and high spike exposure curves was examined (Figure 2).
There were no obvious differences in toxicity between the three types of
exposures when the areas under the curves were equal. These results suggest
that within the range of experimental toxicant concentrations the shapes of
the time-concentration curves were not as important as the total exposure
areas under the curves.
The LC50 for bass subjected to one 90-min, exposure of chlorine was sub-
stantially greater than that for bass subjected to two 90-min. exposures when
each was expressed as mean plateau concentrations (Figure 3A) or as mean
concentrations for the duration of the exposures (Figure 3B). However, there
were no differences in mortalities between the two types of exposures for a
given exposure area (Figure 3C). Similarly, the LC50 for bass subjected to
one 90-min. exposure was greater than that for one 150-min. exposure based
upon toxicant concentrations (Figures 3D and E), but there was little, if any,
difference between the effects of exposures that were alike on an areal
basis (Figure 3F).
These results suggest that measurement of the areas under the curves is
a useful approach when comparing the toxicity to largemouth bass of different
types of short-term exposures of residual chlorine (mostly free residual).
Furthermore, when expressed on an areal basis, the results of the experiment
with one and two 90-min. exposures indicated that there was insufficient
recovery of the bass during the 2-hr, rest period between the exposures to
reduce the mortality associated with a given area under the time-concen-
tration curve (sum of 2 exposures). With sufficient recovery time fewer
deaths would have occurred at a given exposure area, and the response points
would shift downward (Figure 3C) and to the right. If the fish had attained
complete recovery between the exposures, no deaths would have resulted from
the second exposure.
During the 30-min. acclimation period before each acute toxicity test,
the bass swam slowly and deliberately and seldom coughed. A number of
changes of behavior were observed during the tests, however. The changes
usually occurred in the following sequence: (a) increases of the rates of
swimming, opercular activity, and coughing; (b) reduced swimming activity
near the surface of the water, i.e., positioning just under the water
surface; (c) rapid swimming with thrashing at the water surface, some
jumping; (d) lethargic swimming, frequent collisions with aquarium walls
and other fish; (e) "bobbing," i.e., with dorsal portion of head exposed
at the water surface; (f) resting on tank bottom with heavy, pulsating
opercular activity, and some spurts of irregular swimming; and (g) turning
over (belly up).
62
-------
^ O
O
a- 5
v.
MORTALITY
™" *
— • •
o • a
-
-------
_ 6f (A) ,
(A
3
0
5 5
H
1
< *
O
^
2 90min \
Exposures o
/
o
0
1
i i
<
,a r (B)
1
»
// 9(?/77//7
Exposure
1
1
i i i
i
(4
o
-o
4 4
0
/
i i
44
i
; 4 * 4
o
• o
I I
r- (D)
CT>
'
/ ISO m in
Exposure
5-
o:
o
I
I
I
2 3456
TRC (mg/l)
r- (F)
I
I I
• O
«0 O
I
2 34
mg/l
60 80 100 200 300 400
AREA (cm2)
Figure 3. Relationships between mortality (in probits) and the average plateau concentration
of total residual chlorine (A and D), mean concentration for the entire exposure
(B and E), and area under the time-concentration curve (C and F) for bass subjected
to one or two 90-min. square exposures, or to one 90-min. or one 150-min. square
exposure.
-------
Despite the consistency of the behavioral sequence, individual kinds of
behavior often failed to occur. This absence was particularly evident in
tests of low or high TRC concentrations relative to the 96-hr. LC50. Fish
that did not die usually exhibited normal behavior within 24 hours after
the exposures.
Judging the time to first occurrence of each kind of behavior was sub-
jective but the time appeared to decrease as toxicant concentrations
increased. Only the first occurrence in each group of fish was recorded, and
it was not possible to relate the times to areas under the time-concentration
curves, because the bass exhibited a particular response at different times
when the curves were of different height but had the same area to the time of
the response. Thus, our preliminary results were expressed necessarily as
mean concentrations for the exposures to the times of first occurrence. An
example of the relationship for bobbing behavior is taken from the exper-
iment in which the 90-min. square exposure was repeated (Figure 4). The time
to bobbing decreased as the toxicant concentration increased in the first
exposure. Bobbing occurred earlier in the second exposure than in the first
at nearly equal toxicant concentrations. However, two groups did not exhibit
bobbing in the second exposure. The two groups were exposed to acutely toxic
concentrations in the first exposure, and some of the fish were on the
bottoms of the aquaria at the start of the second exposure. The other fish
in the aquaria were active, but their behavioral changes progressed quickly
through the above sequence during the second exposure and the bobbing
behavior was skipped completely (i.e., was never observed).
The behavioral tests of sublethal concentrations of TRC in square and
spike exposures were conducted to estimate the range of concentrations at
which several of the kinds of behavior mentioned above first occurred. No
deaths occurred in these tests. At the highest sublethal concentrations
the sequence of changes of behavior was consistent with the results of
acute toxicity tests, except that turning over did not occur. The thresholds
of occurrence of the behavioral changes were difficult to determine, but
nearing the surface occurred at a smaller exposure area than did thrashing,
lethargic swimming, on bottom, and bobbing (Figure 5). The results of this
test are important because the five kinds of behavior occurred at sublethal
toxicant concentrations.
DISCUSSION
The results of this study have shown the 96-hr. LC50 for largemouth
bass subjected to short-term exposures to TRC was influenced by fish weight
and by exposure duration and frequency. No obvious differences in mortality
were found between groups of bass subjected to square or spike exposures to
free residual chlorine when the areas under the time-concentration curves
were equal. This aspect needs further investigation, however, since the TRC
discharged from power plants may range from mostly combined residual chlorine
to mostly free residual. Furthermore, the species composition of TRC may
change during passage downstream in rivers from the points of effluent dis-
charge (G. Nelson, EPA, personal communication).
65
-------
0)
O
o
0>
e
50
£ 40
o> 30
o
20
10
o
o
o
o
o
1
0
2 3
mg/l TRC
Figure 4, Relationships between the time to first occurrence of bobbing and
the mean concentration of total resiudal chlorine in the aquaria
until the time of first occurrence during the experiment with one
and two 90-min. square exposures. Symbols: • - one exposure
only; o - first of two successive exposures; and * - the
second exposure.
66
-------
GROUP I
(7.37g /fish)
GROUP 2
(I4.63g /fish)
0 20 40
1 1 l l
NS -O €>••• • •
T-O OOO • • O
L -O OOO O • •
8 -o ooo o o o
0"-o oo«« o •
0 20 40
1 1 1 1
MS -0 » *» » • •
T -o o o » • •
L -o o o o • •
3 -o o o • • •
o'-o o o o • •
AREA OF SQUARE EXPOSURE (cm2 )
60 80 100 0 20 40 60
i i l l l i
• • •• • NS
• • •• • T
• • •• • L
• • •• • 8
• • •• • 0
i 1 1 1 1 l
-O OO«* • • • •
-o ooo» • • • •
-o oooo • • • •
- o ooo o o o o •
- o ooo o o « • •
AREA OF SPIKE EXPOSURE (cm2)
60 80 100 0 20 40 60
l l l 1 1 i
• • NS
• • T
• • L
• • 3
• • 0
1 1 1 1 1 1
-o • • • • •
-o o o o o *
-oooo* •
-oooo* o
-o o o • • o
80 100
1 1 1 1
~ •
~ •
~ •
~ •
80 100
1 1 l i
« «
• •
0 0
• •
Figure 5. Relationships between exposure area and occurrence of five behavioral changes in two
weight groups of bass subjected to square and spike exposures. Symbols: NS - near
the surface; T - thrashing; L - lethargic swimming; B - bobbing; and 0 - on bottom.
-------
The area! approach appears to be a valuable method for evaluation
of the toxicity of chlorine over a wide range of laboratory test conditions,
with varying exposure frequencies and durations. Using this approach, the
recovery of fish between exposures can be compared directly with situations
without recovery in a quantitative manner. This approach may not be valid,
however, when chlorinated power plant discharges are contaminated with com-
pounds (e.g., heavy metal complexes, Dickson et al. 1974) that have metabolic
sites of activity different from those for chlorine in fish.
Very little is known about the behavior of fish subjected to inter-
mittent exposures to chlorine in the field. Basch and Truchan (1976) showed
that alewife avoided discharge plumes during chlorination, but returned when
chlorination was terminated. In other tests, however, they noted that sal-
monids in chlorinated discharge plumes sometimes died or exhibited consid-
erable stress at the water surface. Avian predation of fish floundering at
the water surface has been observed below outfalls of chlorinated discharges
from power plants (Brungs 1976). Studies are required to determine under
what conditions particular fish species are trapped in discharge plumes, and
more information is needed on the cumulative effects of short exposures to
TRC of different species compositions on the behavior and survival of fish,
even those fish able to avoid the chlorinated plumes as in the above example.
Considerable attention should be given to the influence of thermal accli-
mation on fish on the acute toxicity of residual chlorine and on the behavior
of the fish in chlorine solutions at elevated temperatures. The development
of an understanding of the influence of behavioral changes on the survival
and well-being of fish in waters receiving intermittent discharges of
chlorine seems particularly appropriate.
68
-------
ACKNOWLEDGMENTS
We wish to thank Dr. Charles E. Warren for his encouragement and
guidance throughout these studies. Dr. P. Doudoroff reviewed the manuscript
and offered many valuable comments and suggestions. His interest in our work
was sincerely appreciated. Staff members and graduate students at the Oak
Creek Laboratory of Biology kindly gave their time to discuss certain aspects
of the results and helped with laboratory experiments. Their efforts were
appreciated.
-------
REFERENCES
Basch, R. E., and J. G. Truchan. 1976. Toxicity of chlorinated power plant
condenser cooling water to fish. Ecol. Res. Ser. EPA-600/3-76-009.
Office of Res. and Develop., U.S. Environmental Protection Agency,
Duluth, Minn, ix + 105 p.
Brooks, A. S., and G. L. Seegert. 1977. The effects of intermittent
chlorination on the biota of Lake Michigan. Spec. Rept. 31. Center
for Great Lakes Studies, The University of Wisconsin - Milwaukee.
ii + 167 p.
Brungs, W. A. 1976. Effects of wastewater and cooling water chlorination on
aquatic life. Ecol. Res. Ser. EPA-600/3-76-098. Office of Res. and
Develop., U.S. Environmental Protection Agency, Duluth, Minn.
vi + 46 p.
Dickson, K. L., A. C. Hendricks, J. S. Grossman, and J. Cairns, Jr. 1974.
Effects of intermittently chlorinated cooling tower blowdown on fish
and invertebrates. Environ. Sci. Technol. 8{9): 845-849.
Heath, A. G. Toxicity of intermittent chlorination to freshwater fish:
influence of temperature and chlorine form. Hydrobiologia.
(In press).
Larson, G. L., F. E. Hutchins, and L. P. Lamperti. 1977. Laboratory
determination of acute and sublethal toxicities of inorganic
chloramines to early life stages of coho salmon (Oncorhynchus kisutch).
Trans. Am. Fish. Soc. 106. {In press).
McLean, R. I. 1973. Chlorine and temperature stress on estuarine
invertebrates. J. Water Poll. Cont. Fed. 45(5): 837-841.
Sokal, R. R., and F. J. Rohlf. 1969, Biometry. W. H. Freeman and Co.
776 p.
Stober, Q. J., and C. H. Hanson. 1974. Toxicity of chlorine and heat to
pink (Oncorhynchus gorbuscha) and Chinook salmon (0_. tshawytscha).
Trans. Am. Fish. Soc. lOSTS"): 569-576.
70
-------
AN APPROACH FOR STUDYING THE
EFFECTS OF MIXTURES OF ENVIRONMENTAL
TOXICANTS ON WHOLE ORGANISM PERFORMANCES
C. F. Muska and L. J. Weber
Department of Fisheries and Wildlife
Oregon State University
Corvallis, Oregon
97331
ABSTRACT
An extensive methodology has been developed to
evaluate the toxicity of individual environmental pol-
lutants for a variety of test animals; however, an ap-
proach is needed to study the possible interactions of
toxicants found together in the environment. A promising
model has previously been proposed for predicting quantal
(all or none} responses of organisms to mixtures of two
or more toxicants. In our laboratory, toxicity studies
using the common guppy, Poe&Llia retieulata, as a test
organism have demonstrated the utility of this model
for predicting their lethal response to a variety of
toxicant mixtures. The usefulness of this approach to
environmental toxicity problems is evaluated in terms of
its applicability to sublethal studies. The model under
investigation and results from experiments studying the
effects of copper, nickel and their mixture on the gross
growth efficiency, relative growth rate, and food con-
sumption of guppies are discussed.
INTRODUCTION
An extensive methodology has been developed for evaluating the effects
of discrete environmental toxicants on a variety of test organisms; however,
when environmental pollution does occur several toxicants are usually present
simultaneously. The recognition of this situation by environmental toxicolo-
gists and those responsible for assessing the potential hazards of man-made
pollutants has generated considerable interest in developing approaches for
evaluating the effects of mixtures of environmental toxicants. Sprague (1970)
in his series of papers on the measurement of pollutant toxicity to fish re-
viewed some of the approaches and the results of previous studies assessing
the joint toxicity of aquatic pollutants.
71
-------
Several years ago, primarily as a result of conversations with Pete
Doudoroff, Charles Warren, and others at our laboratory, we became interested
in this problem and initiated a program to develop and empirically evaluate
an approach for studying the effects of multiple toxicants on the whole
organism performances of fish.
We recognized as others have (Plackett and Hewlett 1948) that only phar-
mocological studies on the modes of action of toxicants applied separately
and jointly can definitively determine the type of interaction between them.
However, the primary actions (the underlying processes by which toxicants
initiate alterations in some pre-existing physiological or biochemical pro-
cess) of toxicants has been elucidated in only a few cases. Even in these
cases it can probably be expected that the more a presumed action is studied
the more likely it will be found to be an effect, the sequence of biochemical
and physiological events that are initiated by the action of a compound
(Fingl and Woodbury 1965).
Given the difficulty and uncertainty in determining the primary mechan-
isms of action of toxicants, the classical pharmacological approach for
evaluating the toxicity of compounds involves studying the relationship
between the concentration of a toxicant and the effects it produces. The se-
lection of an appropriate effect for evaluating the toxicity of a compound
depends on the objectives of the toxicologist. Lethality is often used as
a starting point for studying the toxic properties of a pollutant. There-
fore, it is not surprising that most studies on the joint toxicity of envir-
onmental toxicants have been on quantal responses (all or none) - primarily
death. However, to insure the success of organisms in nature, it is also
necessary to study the effects of toxic substances on such whole organism
performances as growth, reproduction and behavioral responses.
Plackett and Hewlett (1948) suggested that the mathematical examination
of the concentration mortality curves for individual toxicants may indicate
the types of combined effects that occur when the toxicants are present
simultaneously. As a first step for evaluating the effects of multiple
toxicants on whole organism performances, we based our approach on aspects
of various models originally presented by Bliss (1939) and Plackett and
Hewlett(1948) for quantal response data. Using their approach, Anderson
and Weber (1977) were able in most cases to predict the effects of mixtures
of selected environmental toxicants on the survival of guppies (Poeoilia
reticulata). Based on these results we designed a series of experiments to
evaluate the applicability of the approach to graded (sublethal) responses.
The primary objective of this paper is to discuss the rationale of the
proposed approach for studying both the quantal and graded responses of whole
organisms to mixtures of environmental toxicants. Hypothetical dose response
curves with their associated isobole diagrams are presented to illustrate the
different types of toxicant interaction discussed. The results of prelimi-
nary experiments evaluating the effects of the chlorides of copper, nickel
and their mixture on the growth rate, food consumption> and gross growth
efficiency of juvenile guppies are presented (unpublished data).
72
-------
RATIONALE
Using Bliss's paper (1939) as their point of departure, Plackett and
Hewlett (1952) described rather general biological models for toxicant inter-
actions and deduced mathematical models for each based largely upon statisti-
cal considerations. They proposed general types of toxicant interaction
based on the following two-way classification scheme:
Similar Dissimilar
Non-interactive Simple similar Independent
(concentration addition) (response addition)
Interactive Complex similar Dependent
They defined toxicant mixtures as "similar" or "dissimilar" according to whe-
ther the toxicants acted upon the same or different biological systems and as
"interactive" or "non-interactive" according to whether one toxicant influ-
enced the "biological action" of the other toxicants. "Simple similar" and
"independent action" were regarded as special cases in a continuum of biologi-
cal possibilities and the mathematical models proposed for complex similar
and dependent were generalizations of the models proposed for "simple similar
and independent action" respectively.
Their mathematical models particularly for the quantal responses to mix-
tures of "interactive" toxicants are very complex and require the knowledge
of certain parameters which are normally unattainable when evaluating the
effects of toxicant mixtures on whole organism performances. However,
Hewlett and Plackett's models for "joint action" are useful for elucidating
the limitations of and the assumptions required for the special cases of
"simple similar and independent joint action". As a first approach to
evaluating the effects of toxicant mixtures on the whole organism per-
formances such as survival and growth, the present discussion only considers
the special cases of "non-interactive" toxicant mixtures.
A multitude of terms have been suggested to describe the various types
of combined toxicant effects. Ariens (1972) and Fedeli et al. (1972)
reviewed the various terminologies that have been used. As Sprague (1970)
and Warren (1971) point out, the nomenclature is confusing particularly since
certain terms have been defined in more than one way by different authors.
Furthermore, terminology describing mechanisms of toxicant action is not
appropriate for studies evaluating the effects of toxicant mixtures on whole
organism performances without knowledge of the action of the individual
toxicants. To avoid both ambiguities in terminology and assumptions
implying knowledge of sites and mechanisms of toxicant action, Anderson
-------
(1977) introduced the terms concentration and response addition which are
mathematically analogous to the "simple similar" and "independent action"
defined by Plackett and Hewlett (1952).
Concentration addition is mathematically defined as the additive effect
determined by the summation of the concentrations of the individual consti-
tuents in a mixture after adjusting for differences in their respective po-
tencies. The primary assumption governing this type of addition is that the
toxicants in a mixture act upon similar biological systems and contribute to
a common response in proportion to their respective potencies. Bliss (1939)
and others have assumed that if two toxicants act similarly the variations in
susceptibility of individual organisms to the toxicants are completely corre-
lated. As a consequence the dose response curves for the components and the
mixture are parallel. This has been observed for some toxicant mixtures;
however, Plackett and Hewlett (1952) presented examples of chemically related
insecticides which gave non-parallel lines. They and other toxicologists
(Ariens and Simonis 1961; Casarett 1975) have stated, and we believe right-
fully so, that parallelism and hence complete correlation of individual
susceptibilities is not a necessary prerequisite for this type of addition.
In cases where the dose response curves for the individual toxicants in
a mixture are parallel, a dose response curve for the mixture can be cal-
culated based upon the assumption of concentration addition. With the re-
gression equations for the individual toxicants in the form of
y = a + b log x (where y is the % response to each toxicant and x is its
concentration), the regression equation for a binary mixture can be repre-
sented by (Finney 1971):
ym = a1 + b log (T^ + PT^) + b log Z (1)
where,
ym = % response to the mixture
ai = y intercept of the first toxicant
b = common slope
iq = proportion of the first toxicant in the mixture
iT2 = proportion of the second toxicant in the mixture
p = potency of the second toxicant relative to the first
Z = concentration of the mixture
This equation can be readily adapted to represent mixtures containing more
than two toxicants. It should be noted that equation (1) for concentration
addition is similar in principle to the toxic unit method used by Lloyd
(1961), Brown (1968) and others. Whereas the toxic unit method measures
the toxicity of mixtures only at particular levels of response (LC10, LC50,
etc.), equation (1) incorporates the entire dose response curve.
Response addition is the additive effect determined by the summation of
the responses of the organism to each toxicant in a mixture. This form of
addition is based on the assumption that the toxic constituents of a mixture
74
-------
act upon different biological systems within the organism. Each organism
in a population is assumed to have a tolerance for each of the toxicants in
a mixture and will only show a response to a toxicant if the concentration
exceeds its tolerance. Consequently, the responses to a binary mixture are
additive only if the concentrations of both toxicants are above their
respective tolerance thresholds. However, for quanta! responses the
tolerances to the toxicants in a mixture may vary from one individual to
another in a population; therefore, the response of the test animals depends
also upon the correlation between the susceptibilities of the individual
organisms to the discrete toxicants. For example, in order to predict the
proportion of organisms killed by a binary mixture, it is necessary to know
not only the proportion that would be killed by each toxicant alone but also
to what degree the susceptibility of organisms to one toxicant is correlated
with their susceptibility to the other toxicant.
Plackett and Hewlett (1948) recognized this statistical concept and de-
veloped mathematical models that accounted for the correlation of individual
tolerances ranging from total negative to total positive correlation. If the
correlation is completely negative (r = -1) so that the organisms most sus-
ceptible to one toxicant (A) are least susceptible to the other (B), then the
proportion of individuals responding to the the mixture (P ) can be
represented by: m
Pm=PA + PB 1f (W-1* (2a)
where PA and Pg are the respective proportion of organisms responding to
the individual toxicants A and B. With no correlation {r = 0) in suscepti-
bility the relationship is expressed by:
WM1-^ (2b)
In the limiting case of complete and positive correlation (r = 1), individuals
very susceptible to toxicant A in comparison with the population will be
correspondingly very susceptible to toxicant B. In this situation the pro-
portion of animals responding to the mixture is equal to the response to the
most toxic constituent in the mixture. Mathematically this is represented by:
1f
PB-PA
For response addition no significance can be placed on the slope of the dose
response curves because the toxicants in a mixture are acting primarily upon
different biological systems with varying degrees of susceptibility between
organisms. Even if the regression equations for the constituents in a mix-
ture are parallel for toxicants acting in this manner, the dose response
curve for the mixture will not be linear (Finney 1971). This will be illus-
trated later for two hypothetical toxicants whose dose response curves
-------
are parallel. Although the mathematicl equations (2 a,b,c) representing
response addition are relatively simple, the statistical consequences of
this type of addition are more complicated than those of concentration
addition (Finney 1971).
Terms such as supra- and infra-addition are used to describe toxicant in-
teractions which are greater or less than those predicted on the basis of
either concentration or response addition.
Quanta! Response Studies
Hypothetical Dose Response Curves
To graphically illustrate the relationship between concentration and re-
sponse addition, hypothetical dose response curves for two toxicants (A and B)
are plotted in Figure 1 expressing percent response in probits as a function
of the logarithm of total concentration. In this example the dose response
curves for the discrete toxicants are parallel with A being 100 times more tox-
ic than B, We could have also chosen non-parallel curves; however, for these
cases equation (1) for concentration addition is not appropriate. Hewlett and
Plackett (1959) have developed a more generalized model (from which equation
(1) can be deduced) which does not depend on the assumption of parallel dose
response curves.
Dose response curves for mixtures of toxicant A and B are obtained when
the total concentration is varied and the ratio of the concentrations for the
individual toxicants is kept constant. Using the equations {1 and 2 a,b,c)for
concentration (C.A.) and response addition (R.A.), dose response curves were
calculated for different mixtures containing fixed proportions of toxicants
A:B (1:10, 1:100, 1:1000). In Figure 1, the responses to the mixtures are
shown graphically in relation to the dose response curves of toxicants A and
B.
Several observations can be made from the relationships between the dose
response curves in Figure 1. As should be expected, the relative toxicity of
the mixture depends on the ratio of its constituents. In Figure 1, a 1:10
mixture is more toxic than the other mixtures depicted because of the greater
proportion of the more toxic component - toxicant A. At certain ratios,
regardless of the correlation of susceptibility (r), the relative potencies
of the mixtures acting in either a concentration or a responsive additive
manner are very similar. This is observed in Figure 1 for fixed proportions of
1:10 and 1:1000. Furthermore, for any one ratio the relative potency of the
dose response curves for concentration and response addition (r = 1, 0, -1)
depends on the level of response. Focussing on the dose response curves for
mixtures in the ratio of 1:100, it can be noted that at low levels of response
(i.e., at the probit of 2 which corresponds to approximately a 0% response)
the mixtures acting in a concentration additive manner are considerably more
toxic than those acting by response addition regardless of the degree of
correlation (r). This is due to a fundamental difference in the two types
of addition. At threshold or below threshold concentrations of toxicants A
and B, a mixture acting in a concentration additive manner can elicit a
76
-------
o
o
o
9.0 -
8.0-
7.0-
6.0-
5 5.0
o
cc
°- 4.0
3.0
2.0
1.0
TOXICANT A
( Y = 9 + 4X)
50% RESPONSE
<•
(t:
TOXICANT B
(Y = I + 4X)
-2.0
-1.5
-1.0
-0.5 0.0 0.5
LOG TOTAL CONCENTRATION (X)
1.0
1.5
2.0
Figure 1. Hypothetical dose response curves for toxicant A (1:0), toxicant B (0:1) and their mixture
containing the fixed proportions (1:10, 1:100, 1:1000). See text for explanation.
-------
measurable effect because both toxicants are acting upon similar biological
systems. Therefore, their concentrations can sum to produce a concen-
tration for the mixture which is above the threshold level. However,
the responses to toxicants acting upon different biological systems
(response addition) are only additive if each toxicant in a binary mixture
is present in concentrations above their respective threshold levels. For
similar reasons, as the concentrations for the toxicants in a 1:100 mixture
increase, the dose response curves for response addition {except in the
special limiting case where r = 1) become progressively more toxic relative
to the dose response curve for concentration addition. It is even possible
that at high levels of response (in this example, for responses greater than
84% probit of 6.0) mixtures acting in a response additive manner with neg-
ative correlation of susceptibility (r = - 1) can be more toxic than those
acting on the basis of concentration addition.
These factors — the type of interaction, the ratio of the toxicants in
a mixture, and the level of response — must also be considered along with
the toxic properties of the individual toxicants in assessing the relative
toxicity of a mixture. The failure to recognize these factors can poten-
tially lead to erroneous conclusions concerning the nature of the inter-
action of multiple toxicants.
Isobole Diagram
It is difficult to visualize the relationships between the dose response
curves in Figure 1 primarily due to the number of curves presented. However,
the relationships between the hypothetical curves in Figure 1 can be readily
conceptualized with isobole diagrams, a technique introduced by Loewe (1928,
1953). Isoboles are lines of equivalent response. They are constructed by
plotting on a two-dimensional diagram the concentrations of a binary mixture
of toxicants that produce a quantitatively defined response, i.e. a 10%, 50%
or 90% lethal response. It should be noted that en isobole diagram can be
constructed for any level of response and that the relationship between the
isoboles may vary depending upon the response level selected.
The isobole diagram for the 50% level of response of the hypothetical
dose response curves in Figure 1 is present in Figure 2. The x and y axes
in this diagram represent the concentrations of toxicant B and A respectively.
The radiating dashed lines or mixing rays correspond to a series of mixtures
(A:B) of fixed proportions. If the 50% response is produced by combinations of
the two toxicants represented by points inside the square area, the toxicants
are additive. Antagonistic interactions are represented by combinations of
concentrations falling outside the square.
The isoboles for concentration and response addition are determined from
the concentrations of the two toxicants which correspond to the points of in-
tersection between the 50% response line (Figure 1) and the respective hypo-
thetical dose response curves. These concentrations are plotted in Figure 2
on the appropriate mixing ray. The lines connecting these points define
the course of the isobole. Concentration addition is represented by the
diagonal isobole. For quantal data, response addition is defined by the
78
-------
CO
IMO
|:50
|: 100
.10
RESPONSE, ADDITION (r=J)_
:£00
I: ICOO
2.0 4.0 6.0 8.0
CONCENTRATION OF TOXICANT B
10.0
Figure 2. Isobole diagram for quanta! response data. Isoboles for
concentration and response addition were determined from
hypothetical dose response curves in Figure 1.
79
-------
curved isoboles for complete negative (r = -1) and for no correlation (r = 0)
in susceptibility. The upper and right boundaries of the square cor-
respond to the limiting case of response addition with complete positive
correlation (r = 1).
The term "no interaction" had been used by other authors (Sprague 1970;
Warren 1971) to describe the response additive isobole in Figure 2 corres-
ponding to complete positive correlation of susceptibilities. We recognize
that the equation (2c) used to determine this isobole is not additive in a
strictly mathematical sense. For example, in lethality studies, organisms
whose tolerances to the individual toxicants are positively correlated
(r = 1) die in response to the most toxic constituent in the mixture; there-
fore there is no addition of responses. However, in experimental situations,
it is unlikely that complete positive or for that matter complete negative
correlation will often be observed. Consequently we have chosen to represent
complete positive correlation as a limiting case of response addition to be
consistent in our terminology and more importantly to emphasize that the
isobole for response addition will for most toxicant mixtures fall between
the extreme cases of r = -1 to r = 1 depending upon the degree of correlation.
For reasons similar to the one presented by Warren (1971), we have chosen
to use the terms supra- and infra-addition to describe interactions that are
greater or less than expected on the basis of either concentration or response
addition. It is important that these terms be used in reference to a parti-
cular type of addition. For example, an isobole falling between the isoboles
for concentration and response addition (r = -1) could be designated as both
infra- and supra-additive depending on the nature of the interaction. This
potentially confusing situation is avoided by using the terms in the manner
we have suggested.
The term antagonism in Figure 2 refers to a physiological or functional
antagonism. In the present discussion, we do not consider toxicants which can
chemically or physically react in the external medium of an organism to form
an inactive or less toxic product (chemical antagonism). Some investigators
have used the term antagonism to describe interactions that are less toxic than
strict additivity (concentration addition) but whose mixture still has a com-
bined effect greater that either constituent applied alone. We prefer to use
the term infra-addition to describe these cases and to reserve antagonism for
those cases where the presence of one toxicant necessitates that a higher
concentration of another toxicant be present to obtain the defined level of
response.
Graded Response Studies
A consideration of the nature of the dose response curves for quantal
and graded responses shows that the effects they express are quite different.
Quantal dose response curves express the incidence of an all-or-none effect
(usually death) when varying concentrations are applied to a group of organ-
isms. The curve is derived by observing the number of organism; which respond or
80
-------
fail to respond at various concentrations. Consequently, the slopes of these
curves primarily express the individual variation of the population to a par-
ticular toxicant. Graded dose response curves characterize the relationship
between the concentration of a toxicant and the magnitude of the effect under
consideration. The dose response curve can be derived by measuring on a
continuous scale the average response of a group of organisms at each concen-
tration.
As Clark (1937) and others have pointed out, it is possible to represent
any graded response as a quanta! response provided that the response of each
individual organism can be measured. However, this procedure if adopted is at
the expense of some "loss of information" (Gaddum 1953). Quanta! response dia
reveals only the number of organisms that respond or fail to respond at some
particular concentration. On the other hand, graded response data not only
tells us whether or not a group of organisms respond but also how much they
respond.
The mathematical equations (2 a,b,c) for the response addition are not
appropriate for graded effects for two reasons. First, there is a dif-
ference in the way the two types of data are measured. For quanta! responses
the proportion of organisms responding to any concentration is determined by
the ratio of number of organisms showing the response to the total number
subjected to the concentration. For graded responses the mean response to each
dose is measured but in general the maximum possible response is not known.
In cases where the maximal effect is not known, no proportional response can
be calculated. This is particularly true for growth experiments where an
organism's response can potentially range from growth enhancement to negative
growth depending on the concentration of a particular toxicant. Secondly,
the statistical concept of correlation between the susceptibilities of the
organisms to the discrete toxicants in a mixture is not appropriate for
graded responses measured in the manner described earlier. Graded response
data represent the average response of a group of organisms. Therefore, the
response of each individual organism to the toxicants is not known. To be
sure the tolerances of the individuals in the group will vary for the dif-
ferent toxicants in a mixture; however, this factor will not alter the
relative toxicity of the mixture because the range of tolerances of the
population is theoretically represented in the sample of organisms from this
population.
For graded response data, we have represented the combined response to a
mixture of toxicant? acting in a response additive manner as simply the sum
of the intensities of response which each component toxicant produces when
administered alone. A similar relationship was defined by Loewe (1953). Con-
centration addition can be predicted for a toxicant mixture using equation
(.1) if the component toxicants exhibit parallel dose response curves. Figure
3 represents an isobole diagram for a graded response. The isoboles for
concentration and response addition were determined with the appropriate
mathematical equations discussed above,,
81
-------
1:0
O
X
O
h-
.10
.08 -
:IOO
cr
h-
z
UJ
O
2:
O
CJ
8.0
CONCENTRATION OF TOXICANT B
10.0
0:1
Figure 3. Isobole diagram for graded response data.
82
-------
The relatively simple types of isoboles represented in Figure 2 and 3
should only be expected for relatively simple in vitro systems or in
situations where there is a clear cut relationship between dose and effect.
Given the complexity and interdependency of physiological systems, it is
reasonable to suppose a priori that the special types of additivity as repre-
sented by strict concentration and response addition will be approximated
only occassionally in the responses of whole organisms to mixtures of envir-
onmental toxicants. Furthermore, as mentioned earlier, the relative tox-
icity of a mixture depends on several factors which include the level of
response (i.e., 10%, 50%, 90% response), the ratio of the toxicants in a
mixture (i.e. 1:10, 1:100, 1:1000) and the nature of the response itself.
It should be noted that the type of addition can only be described in
relation to the response under consideration. With the same mixture of
toxicants, different types of toxicant interaction might be expected for
different responses (i.e., survival, growth, reproduction). However, these
special types of toxicant interaction do provide a frame of reference for
evaluating the effects of toxicant mixtures on whole organism performances.
Isobole diagrams are useful for visualizing the relationship between dif-
ferent types of toxicant interactions and for delineating the various factors
which can influence the relative toxicity of multiple toxicants. However, in
practice, isoboles are difficult to derive requiring a series of dose response
curves for the mixture at different ratios of the component toxicants. Further-
more, there is no statistical criteria which might be used to distinguish
between one form of interaction and another (Plackett and Hewlett 1952). Fol-
lowing the procedures of Anderson and Weber (1977) we empirically studied the
interaction of copper and nickel by deriving a dose response curve for the
mixture at one fixed proportion. The dose response curve determined for the
mixture was statistically compared to curves predicted on either the basis of
concentration or response addition. This approach, utilized by Anderson and
Weber (1977) for lethality studies, was adopted in the present study in order
to test its applicability to graded response data.
EXPERIMENTAL STUDIES
Lethality Studies
Anderson and Weber (1977) conducted a series of 96 hour bioassays,
studying the effects of copper, nickel and their mixture on the survival of
male guppies. Statistical tests suggested that the individual dose response
curves derived for copper and nickel were parallel. Based upon this obser-
vation, it was assumed that the mixture would be concentration additive. To
test this prediction, Anderson performed experiments exposing test organisms
to a series of mixtures of the two toxicants at a fixed proportion. A sta-
tistical comparison of the observed dose response curve to the regression
equation calculated on the basis of equation (1) indicated that the
assumption of concentration addition adequately described the joint toxicity
of the mixture. Using a similar experimental procedure he demonstrated that
a mixture of copper and zinc was supra-additive relative to concentration
-------
addition. Further studies showed that separate binary mixtures of dieldrin
and potassium cyanide and potassium pentachlorophenate and potassium cyanide
were response additive.
Growth Studies
Growth was selected as the graded response for this study because it re-
presents a performance of the integrated activities of the whole organism and
as such is often a sensitive indicator of the suitability of the environment
(Warren 1971). Two of the ways environmental toxicants can affect the growth
of an organism are: (1) alter its ability to assimilate and convert food
material into body tissue, and/or (2) change its rate of food consumption. To
determine the manner in which toxicants affect the growth of an organism, both
processes were investigated separately. The methodological and statistical
procedures along with the complete results of this study will be published at
a later date; however, the results of a preliminary analysis of this data
are discussed.
Juvenile guppies were fed daily a restricted ration of tubificid worms to
determine the effect of the toxicants on the gross growth efficiency and re-
lative growth rate (as defined by Warren 1971) of the fish. The effect of the
individual toxicants and their mixture on food consumption was investigated
by feeding groups of fish an unrestricted ration and measuring the amount of
worms consumed.
Statistical tests comparing the slopes of the individual dose response
curves for copper and nickel derived for each response suggested that they
were parallel. On the basis of the mathematical model for concentration addi-
tion, equations for the predicted dose response curves were calculated and sta-
tistically compared to the regression equations experimentally determined for
the mixture. The results indicate that the effects of the toxicant mixture on
the gross growth efficiency of the fish subjected to both the restricted and
unrestricted feeding regimes are predictable on the basis of concentration
addition. However, the dose response curves for the mixture representing the
effects of the toxicants on the food consumption of the fish was supra-
additive relative to the dose response curve predicted on the basis of concen-
tration addition. Because of the relationship between growth, gross growth
efficiency, and food consumption, the effects of the mixture on the relative
growth rate are similar to the ones observed for gross growth efficiency at
the restricted ration (concentration addition) and for food consumption at the
unrestricted ration (supra-addition).
CONCLUSIONS
The results indicate that the assumption of concentration addition ade-
quately predicts the effects of a copper-nickel mixture on both the survival
and gross growth efficiency of guppies. The dose response curves for the mix-
ture representing the effects of the toxicants on the food consumption of the
fish was supra-additive relative to the dose response curve predicted on the
basis of concentration addition. An explanation for the differences in
these two responses to the mixture is beyond the scope of the present study.
84
-------
However, it is possible that the effects of the toxicants on the metabolic
processes involved in the conversion of food material into body tissue
might be somewhat different than their effects on the biological processes
regulating the consumption of food.
In our studies we found that the mathematical model for concentration
addition predicted the responses of guppies to both lethal and sublethal con-
centrations of a copper and nickel mixture. However, it should not be in-
ferred from these results that the type of joint toxicity observed when or-
ganisms are subjected to high, rapidly lethal concentrations of mixtures will
necessarily occur in cases where animals are subjected to low concentrations
of the same toxicants. Furthermore, the nature of toxicant interaction can
only be meaningfully described in relation to the particular response under
consideration. For example we found that mixtures of copper and nickel were
concentration additive in experiments evaluating their effects on the gross
growth efficiency of the guppies; however, in the food consumption studies,
the same mixture at similar concentrations produced a more toxic response than
was predicted on the assumption of concentration addition.
To insure the success of a species in nature, it is necessary to evaluate
the effects of potentially hazardous toxicant mixtures on the performances
of whole organisms. The proposed approach provides a methodology for assessing
the toxicity of mixtures of environmental toxicants at this level of biological
organization. However, to offer explanations as to why mixtures of environ-
ment toxicants interact in a particular manner requires knowledge of the
effects of combined toxicants on underlying biochemical processes and physi-
ogical functions. Such studies will be useful for evaluating the assumptions
of the proposed approach and in suggesting other possible types of toxicant
interaction.
ACKNOWLEDGEMENTS
This research was supported by NIH Grant ES-00210 and a traineeship
from NIH-PHS Grant GM07148.
-------
REFERENCES
Anderson, P, D. and L. J. Weber. 1977. The toxicity to aquatic populations
of mixtures containing certain heavy metals. Proceedings of the
International Conference on Heavy Metals. 2:933-953. (In press).
Ariens, E. J. and A. M. Simonis. 1961. Analysis of the action of drugs and
drug combinations. Pages 286-311 in H. de Jonge, editor. Quanti-
tative Methods in Pharmacology. North-Holland Publishing Company,
Amsterdam. 391 pp.
Ariens, E. J. 1972. Adverse drug interactions — interaction of drugs on
the pharmacodynamic level. Proceedings of the European Society for the
Study of Drug Toxicity. 13:137-163.
Bliss, C. I. 1939. The toxicity of poisons applied jointly. Ann. Appl.
Biol. 26(3):585-615.
Brown, U. M. 1968. The calculation of the acute toxicity of mixtures of
poisons to rainbow trout. Water Research. 2(10):723-733.
Casarett, L. J. 1975. Toxicological evaluation. Pages 11-25 -in L. J.
Casarett and J. Doull, editors. Toxicology -- the Basic Science of
Poisons. MacMillan Publishing Company, Inc., New York. 768 pp.
Clark, A. J. 1937. General pharmacology. In W. Heubner and J. Schuller,
editors. Heffler's Handbuch der Experimentellen Pharmakologie<,
Vol. 4. Verlag von Julius Springer, Berlin. 228 pp.
Fedeli, L., L. Meneghini, M. Sangiovanni, F. Scrollini and E. Gori.
1972. Quantitative evaluation of joint drug action. Proceedings of
the European Society for the Study of Drug Toxicity. 13:231-245,
Fingl, E., and D. M. Woodbury. 1965. General principles. Pages 1-36 in
L. S. Goodman and A. Gilman, editors. The Pharmacological Basis of
Therapeutics. 3rd ed. The MacMillan Company, New York. 1785 pp.
Finney, D. J. 1971. Probit Analysis. 3rd ed. Cambridge University Press,
Cambridge. 333 pp.
Gaddum, J. H. 1953. Bioassays and mathematics. Pharmacological Reviews.
5(1):87-134.
Hewlett, P. S., and R. L. Plackett. 1959. A unified theory for quantal
responses to mixtures of drugs: non-interactive action.
Biometrics 15(4) :591-610.
Lloyd, R. 1961. The toxicity of mixtures of zinc and copper sulphates to
rainbow trout (Salmo gairdnevii Richardson) Ann. Appl. Biol.
49(3):535-538.
86
-------
Loewe, S. 1928. Die quantitativen Problems der Pharmakologie. Ergeb.
Physio!., biol. Chem., exp. Pharmakol. 27:47-187.
Loewe, S. 1953, The problem of synergism and antagonism of combined drugs.
Arzneimittel - Forsch. 3:285-290.
Plackett, R. L. and P. S. Hewlett. 1948. Statistical aspects of the
independent joint action of poisons, particularly insecticides.
I. The toxicity of mixtures of poisons. Ann. Appl. Biol.
35(3):347-358.
Plackett, R. L. and P. S. Hewlett. 1952. Quantal responses to mixtures
of poisons. J. Royal Statistical Soc. B14(2):141-163.
Sprague, J. B. 1970. Measurement of pollutant toxicity to fish.
II. Utilizing and applying bioassay results. Water Research.
4(l):3-32.
Warren, C. E. 1971. Biology and Water Pollution Control. W. B. Saunders
Company, Philadelphia. 434 pp.
-------
RELATIONSHIP BETWEEN pH AND ACUTE TOXICITY
OF FREE CYANIDE AND DISSOLVED SULFIDE FORMS
TO THE FATHEAD MINNOW
Steven J. Broderius and Lloyd L. Smith, Jr.
Department of Entomology, Fisheries and Wildlife
University of Minnesota
St. Paul, Minnesota 55108
CREDIT
The authors wish to acknowledge the assistance of David T. Lind in
conducting the bioassays and Dr. Peter Doudoroff for his critical review
of the manuscript.
This research was supported by the University of Minnesota Agricultural
Experiment Station and by the U.S. Environmental Protection Agency under
Grant Numbers R800992 and R802914.
INTRODUCTION
There are several fish surfaces where exchange of gases and ions between
blood and water can occur, but the gill epithelium is recognized as the
primary site. Generally ions have less toxicity than the more lipid-soluble
un-ionized molecules. Large hydrated ions are less toxic because of the
difficulty in penetrating strongly charged membranes. Ions are repulsed by the
charged protein surfaces of the membranes or adsorbed to the membranes. The
acute toxicity to fish of solutions containing free cyanide (i.e., HCN plus
CN-) of dissolved sulfide (i.e., H2$, HS~ and S2~) is mainly attributed to
the toxic action of molecular HCN of H2S. The toxicity is related directly
to the concentration of these gases in solution and inversely to pH, and is
largely independent of the concentrations of the CN" or HS~ and S^- anions,
which are considerably less toxic than the molecular forms (Doudoroff and
Katz 1950; Bonn and Foil is 1967; Doudoroff 1976).
Two different gill permeability theories have been proposed to explain
the observed changes in toxicity of ammonium salt solutions to aquatic
organisms with change of pH. Lloyd and Herbert (1960) suggested that only
un-ionized NH3 is effective and that it is not the pH value of the test
solution that is important in determining the toxicity of ammonium salts
to fish, but it is the pH value of the solution at the gill surface which
controls the concentration of NH3 at the penetration site. The gill surface
pH supposedly depends on the amount of respiratory C02 excreted, which lowers
-------
the pH value of the solution in contact with the gills. A second theory
(Tabata 1962) stated that the ionized fraction of ammonia penetrates mem-
branes and has a measurable toxicity considerably less than the more
rapidly penetrating molecular form. Both theories for ammonium salts
appear generally applicable to the toxicity of weak bases and weak acids.
The Theories presumably can be used to assess the toxicity of other poisons,
affected by pH changes within the range tolerated by fish;
When HCN or h^S gas are dissolved in water, ionization equilibria are
established that can be represented by the equations:
HCN(aq) ^ H+ + CN~ or H2S J1 H+ + HS~ i2 2H+ + S2' (1)
The second equilibrium constant for dissolved sulfide is small in comparison
with the first and can be omitted in equilibrium calculations, since the
sulfide ion (S^-) is negligible when the test pH is less than about 11. The
K, and K-J constants are such that in most natural waters molecular HCN or
the hydrosulfide ion (HS~) can be expected to be the predominant free cya-
nide of dissolved sulfide forms.
The change in tolerance limits for the fathead minnow (Pmephales
promelas Rafinesque) and dissolved sulfide forms were studied as a function
of pH, because the toxicity of weak acids and bases is known to be pH
dependent. It was anticipated that experimental results could largely be
explained by one of the gill permeability theories.
MATERIALS AND METHODS
TEST WATER AND FISH
The experimental well water used in all bioassays had a total hardness
of 220 mg/1 as CaCCh. A comprehensive analysis of the water was reported
by Smith et al . (1976).
Juvenile fathead minnows were used as test organisms to study the
toxicity of solutions containing cyanide or sulfide at various pH values.
The fathead minnow was chosen as an experimental organism because it can be
cultured and maintained in a laboratory, is handled with ease, and has a wide
distribution in chemically diverse natural waters, including those of acid
bog lakes and lake waters of high pH. The fathead minnows used in all the
bioassays were cultured in the laboratory from a stock originally obtained
from the U.S. Environmental Protection Agency's Environmental Research Lab-
oratory in Duluth, Minnesota. The minnows were reared in the laboratory
under a constant photoperiod in 30-liter glass aquaria receiving a continuous
supply of well water at 25 C and with a pH of approximately 7.9. We believed
that the inbred laboratory strain of fish would have a uniform sensitivity
to the toxicants not too different from that of other stocks tested at dif-
ferent times, and that possible adverse effects of disease stress and/or
-------
treatment could be avoided by not using wild stocks.
Six lots of fish were tested during separate 15-week periods in each
of the cyanide and sulfide series. The fish were all approximately 13-
weeks old at testing, had mean total lengths of 30.8 and 30.1 mm, and had
mean wet weights for survivors of 0.289 and 0.293 g in the cyanide and sulfide
bioassays, respectively. The fish were fed Oregon Moist and Glencoe pelleted
food twice daily until one day before exposure to the toxicants.
TESTING CONDITIONS
The 96-hr toxicity bioassays were performed in three identical diluter
and test chamber units, each including one control and four treatment
chambers. The experimental glass chambers measured 50 x 25 x 20 cm high
and contained 20 liters of test solution. The intermittent water delivery
and toxicant introduction systems were modifications of those described by
Brungs and Mount (1970) and Mount and Warner (1965), respectively. Flow
through each chamber was at the rate of 500 nfl/min, affording 99% replace-
ment in about 3 hours.
The pH of the test water was controlled by dispensing a sulfuric acid
or sodium hydroxide solution with a "dipping bird" into the head reservoirs.
The temperature of the test water was thermostatically controlled at 20 C.
The test water was aerated in the head reservoirs to maintain dissolved
oxygen concentrations in the test chambers near 7.5 mg/1. Each test chamber
was illuminated for 12 hours each day with a 40-watt incandescent bulb
placed 10 inches above the chamber.
Stock solutions of sodium cyanide or sodium sulfide were prepared with
reagent grade chemicals and deionized water. One pellet of sodium hydroxide
was added to each liter of stock solution to raise the pH, thus retarding
escape of HCN or I^S from the "dipping bird" reservoirs. The solutions
were dispensed to the reservoirs from Mariotte bottles. Three days before
initiation of the bioassays, 10 or 20 fish acclimated to 20 C for one week
were randomly distributed among the 12 treatment and 3 control chambers.
Sulfuric acid or sodium hydroxide was then slowly added to the head reser-
voirs to attain the desired pH. The fish were acclimated to the specified
pH for at least two days before Introduction of the toxicants.
CHEMICAL ANALYSIS
During each bioassay, water temperature, dissolved oxygen (DO), and pH
in each test chamber were measured daily. Alkalinity was determined daily in
each control chamber by potentiometric titration with a standard 0.02
N ^504 solution to the successive bicarbonate and carbonic acid equivalence
points, indicated by inflections of the titration curve. Dissolved oxygen
was measured with a galvanic-type membrane electrode meter precal ibrated by
the Winkler method, and pH with a Corning Model 112 glass electrode meter
standardized with two primary buffers (APHA 1975). The free carbon dioxide
90
-------
concentrations in test solutions were derived by the nongraphic method
(APHA 1975) from the pH, temperature, bicarbonate alkalinity, and total
filtrable residue of 247 mg/1.
Free cyanide concentrations in each chamber were determined daily by the
pyridine-pyrazolone colon metric method (APHA 1971). Dissolved sulfide
concentrations, which were determined to be essentially the same as total
sulfide concentrations, were measured in treatment chambers at least twice
daily. Water samples taken from the center of each test chamber were
stabilized with zinc acetate and analyzed for sulfide by the methylene blue
colorimetric procedure (APHA 1971). The concentration of molecular HCN or
H2S was calculated for each free cyanide or dissolved sulfide determination
using the daily pH and temperature measurement and the pK equilibrium con-
stants i.e., -log K) calculated by using the equations pKucN = 3-658 + 1662/T
(Broderius, unpublished data) and pKH2S = 3.122 + 1132/T (Broderius and Smith
1977), where T is temperature in degrees Kelvin. The ratio [HCN]/[free
cyanide] or [H2]/[dissolved sulfide] is taken to be equal to I/O + lpPH - P«a)>
assuming [S2~] to be negligible. When the molecular HCN or H2S and free
cyanide or dissolved sulfide concentrations are expressed in the same units
and as the molecular form, then free cyanide or dissolved sulfide times the
appropriate factor (i.e., the ratio computed as shown above) will equal molecu-
lar HCN or H2S.
STATISTICAL ANALYSIS
Estimates of the concentration of cyanide or sulfide most likely to cause
50% mortality (LC50) after 96 hours of exposure were made from lines fitted
mathematically by the BMD03S log-probit analysis computer program (Dixon
1973). The data from toxicity tests conducted at a specific pH aim were
composited for probit analysis. The 96% confidence intervals for LC50
values were computed according to formulas proposed by Litchfield and
Wilcoxon (1949) and Finney (1971). (Chi)2 tests were applied to each group of
data to determine variability and acceptability. When heterogeneous data
were indicated, the appropriate adjustment in the 95% confidence intervals
for LC50 values were made.
RESULTS AND DISCUSSION
The relationship between test pH and acute toxicity of free cyanide
and dissolved sulfide forms to the fathead minnow at 20 C was determined for
pH values ranging from about 6.8 to 9.3 and 6.5 to 8.7, respectively. The
test conditions and log-probit analysis of composite test results grouped
according to pH are summarized in Table 1-3 and Figure 1. It is apparent that
the 96-hour median lethal concentrations (LC50) of free cyanide and molecular
HCN were little different and fairly constant within the pH range 6.8 to 8.3.
Beyond this pH to pH 9.3, the values diverged markedly, with the free cyanide
values increasing and the HCN values decreasing.
Except for some increase with rise of pH from about 6.5 to about 7.1
-------
u
160
120
80
40
^.O 6.5
FREE CYANIDE
CN ~
-• HCN
800
V)
CJ
X
600
D
400
200
• • DISSOLVED SULFIDE
* A HS-
•• H2S
Figure 1. Relationship between test pH and 96-hr LC50 cyanide or sulfide
concentrations for the fathead minnow at 20° C.
92
-------
Table 1. Mean test conditions for composite free cyanide and dissolved
sulfide bioassays: standard deviations in parentheses
Alkalinity,
mg/1 CaC00
Test
pH
6.830(0.028)
7.559(0.041)
8.286(0.034)
8.672(0.067)
8.974(0.068)
9.262(0.091)
6.462(0.070)
7.101(0.030)
7.698(0.036)
8.151(0.030)
8.430(0.018)
8.693(0.014)
Temperature,
C
20.1(0.12)
20.0(0.02)
20.0(0.01)
20.1(0.09)
20.0(0.12)
20.2(0.09)
20.0(0.05)
20.0(0.04)
20.0(0.20)
20.0(0.11)
20.0(0.05)
20.1(0.10)
DO.
mg/1
Cyanide
7.65(0.06)
7.56(0.09)
7.66(0.06)
7.59(0.09)
7.71(0.10)
7.63(0.12)
Sulfide
7.62(0.08)
7.66(0.09)
7.65(0.07)
7.49(0.24)
7.55(0.11)
7.49(0.11)
Bicar-
bonate
Series
96
193
240
229
214
206
Series
63
128
198
229
234
234
o
Total
96
193
240
244
243
250
63
128
198
229
238
243
Free C02 in
test solution*
mg/1
28.0
11.0
2.5
1.0
0.47
0.24
44.0
20.5
7.9
3.2
1.7
1.0
mm Hg
12.4
4.9
1.1
0.44
0.21
0.11
19.5
9.1
3.5
1.4
0.75
0.44
Free C02 evaluated by nomographic method (APHA 1975). Assuming
K = H2COs/Pc02 and 1°9 K = -1.41 at 20 C and one atmosphere, then
1 mg/1 Her C02 = 0.444 mm Hg C02 tension (Stumm and Morgan, p. 148,
1970),
-------
Table 2. Analysis by the log-probit method of the results of fathead minnow
bioassays of cyanide at 20 C, with tests grouped according to pH
Log-probit regression analysis* 95% confidence
Test Treat- 96-hr LC50, limits for
pH ments a 3 yg/1 as HCN LC50, yg/1
HCN
6.830 6 -23.80 13.76 124 106 - 144
7.559 7 -23.47 13.81 115 102 - 130
8.286 6 -27.37 15.52 122 115 - 128
8.672 9 -33.46 18.88 109 105 - 113
8.974 17 -16.52 10.68 104 96 - 112
9.262 10 -28.73 17.55 83 80 - 87
Free Cyanide
6.830
7.559
8.286
8.672
8.974
9.262
6
7
6
9
17
10
-23.82
-23.60
-28.00
-34.59
-16.44
-28.57
13.76
13.82
15.54
18.64
9.83
15.29
124
117
133
133
152
157
107 -
102 -
126 -
128 -
140 -
142 -
144
135
140
138
165
173
For equation Y-j = a + 6 (log Xi) when Y^ is the maximum likelihood probit
value and X-j is log cyanide concentration as yg/1 HCN (Dixon 1973).
94
-------
Table 3. Analysis by the log-probit method of the results of fathead minnow
bioassays of sulfide at 20 C, with tests grouped according to pH
Log-probit regression analysis*
95% confidence
Test Treat- 96-hr LC50, limits for
pH ment a B yg/1 as h^S LC50, yg/1
H£
6.462 11 - 9.85 8.78 49.2 45.1 - 53.7
7.101 8 -14.12 10.88 57.2 52.6 - 62.1
7.698 8 -18.26 14.70 38.2 32.8 - 44.5
8.151 6 -31.70 25.98 25.8 24.7 - 27.0
8.430 6 -30.67 28.19 18.4 17.5 - 19.4
8.693 9 -11.07 13.68 14.9 13.0 - 17.2
Dissolved Sulfide
6.462
7.101
7.698
8.151
8.430
8.693
11
8
8
6
6
9
-10.64
-18.57
-39.54
-52.98
-52.31
-40.23
8.65
11.09
18.73
22.10
20.94
15.56
64.1
133
239
420
546
806
58.7
123
219
394
483
726
- 70.0
- 145
- 260
- 448
- 616
- 895
For equation Y^ = a + 3 (log Xj) when Y-j is the maximum likelihood probit
value and X-j is log sulfide concentration as yg/1 H£S (Dixon 1973).
95
-------
the 96-hr LC50 concentrations of molecular l^S decreased as the test pH
increased. These values, in ug/liter H2$ ranged from 57.2 at pH 7.1 to 14.9
at pH 8.7. Within this pH range, a 0.1 unit increase in pH was calculated
by linear regression (r of -0.994) to result in a 2.7 ug/liter decrease in
the 96-hr LC50 value of molecular H2S. However, as pH increased, the con-
centration of dissolved sulfide in equally toxic solutions increased
logarithmically.
The anomalous result for the sulfide experiments performed at a pH
of about 6.5 may be due to an interaction between sulfide and the relatively
high CO? concentration in the test solutions. It is also possible that the
LC50 values for H2S are relatively constant over the pH range of about 6.5
to 7.1 and at a constant free C02 level.
LLOYD AND HERBERT'S THEORY
According to the theory of Lloyd and Herbert (I960), the toxicity of
cyanide and sulfide solutions to the fathead minnow is increased by
depression of pH at the gill surface due to respiratory excretion of ($2.
This change in toxicity results from the conversion of CN~ of HS~ anions'to
molecular HCN or H2S in the solution in contact with the gills. The magnitude
of this effect depends upon the concentration of free C02 present in solution
and the shifting of the C02-bicarbonate-carbonate chemical equilibrium.
When the concentration of free C02 in the water is very low (high pH), the
addition of respiratory C0£ considerably reduces the pH value at the gill
surface. As the level of free C02 rises in the bulk of the solution, the pH
change becomes less. The pH at the gill surface can, theoretically, be
calculated from the bicarbonate alkalinity, temperature, free C02 concen-
tration in the test solution, and the free C02 excreted by the gills of the
fish by use of the standard nomographic method (APHA 1975). The increase
in concentration of excreted C02 in the respiratory water (as mg C02/liter)
was estimated by Lloyd and Herbert (1960) by means of the equation:
mol. wt. C0?
Increase in OX, = DO x RO x mol> wt< ^ x m (2)
where DO = dissolved oxygen concentration of water in mg/liter
RQ = respiratory quotient of the fish
P = percentage of oxygen removed from the respiratory water
by the fish
Kutty (1968) has determined that the respiratory quotient (RQ) is
essentially unity when freshwater fish are spontaneously active in nearly
air-saturated water. Since the C02 is excreted along the surface of the
lamellae, a pH gradient may be present in the gills. Lloyd and Herbert
proposed that the average pH be taken to be that which is produced when half
of the total amount of C02 excreted is at equilibrium with the carbonate
96
-------
system. An analysis of the cyanide and sulfide bioassay data by their
procedure is summarized in Table 4. The apparent change in HCN and H?S
toxicity with change of test pH could mathematically best be reconciled with
this theory by assuming a respiratory quotient of 1.0 and that the fathead
minnow removed from the respiratory water about 10 and 55% of the dissolved
oxygen in the cyanide and sulfide bioassays, respectively. If a respiratory
quotient of less than 1.0 is assumed, as proposed by Lloyd and Herbert (1960),
then a greater percentage of the dissolved oxygen available at the gills would
need to be removed for conformance with the theory.
The above proposed explanation of the relation to pH of the toxicity
to fish of weak acids and bases is interesting, but for a number of reasons
it may not be appropriate. First, it was assumed by Lloyd and Herbert in
their calculations that the utilization of oxygen initially present in the
water passing over the gills of rainbow trout is about 80%. Recent studies
have shown that the percentage of oxygen removed is variable among individu-
als of the same and different species and for a given fish under different
conditions and at different times. According to a review by Shelton (1970),
almost all of the reported utilization values are less than 80% under
nearly ideal environmental conditions and the utilization is usually less
than 50% when there is hypoxic stress (Holeton and Randall 1967; Davis and
Cameron 1971). In view of the stress occurring in an acute toxicity bioassay,
it is not unreasonable to assume that the utilization might even be consider-
ably less than the 55% value necessary for agreement with the pH drop at
the gill theory in the case of the sulfide bioassays. In attempting to
explain the 96-hr LC50 bioassay results (Table 4) obtained in tests performed
under nearly identical environmental conditions, the percentage utilization
of dissolved oxygen at the gills must be 5.5 times as great for fathead
minnows tested in sulfide solutions as for those tested in the cyanide solu-
tions. This discrepancy is an additional reason why the theory is not viewed
as an appropriate explanation of the relation to pH of the toxicity of weak
acids to fish.
The respiratory quotient of fish may not remain constant throughout the
duration of an acute toxicity test. According to Kutty (1968, 1972) and
Kutty et al. (1971), the in vivo RQ can increase upon marked reduction of
DO as a result of the accumulation of lactic acid in the tissues and conse-
quent release of (XL from the bicarbonate reserve, which take place when
metabolism is partially anaerobic. Because of the nature of the toxic action
of cyanide and sulfide, a similar response may be expected of fish dying from
these poisons. If the RQ does increase, the percentage of available oxygen
removed at the gills could decrease and still be accompanied by an increase
in CO- at the gills.
A second weakness of the theory of Lloyd and Herbert is apparent when
one examines the manner and amount in which CO,, is excreted at the gills.
According to Dejours et al. (1968), when the respiratory quotient is near
unity, the changes in C02 tension of the water passing over the gills of
teleost fish are small. In their experiments with goldfish at 25 C, this
change was less than 1.0 mm Hg. The change in tension of highly soluble ^
in well-aerated, high-carbonate water is small because of the high rate of
ventilation which is necessary to extract poorly soluble oxygen. The
97
-------
Table 4. Average estimated 96-hr LC50 concentrations of molecular HCN or H2S at the fathead minnow gill
surface assuming a dissolved oxygen utilization of 10 and 55 percent for the cyanide and
sulfide bioassays at 20 C, respectively, and an RQ of 1.0.
CO
Test
PH
6.830
7.559
8.286
8.672
8.974
9.262
Average free
C02 at gill,1
mg/1
28.53
11.52
3.03
1.52
1.00
0.76
Average pH
at gill2
6.82
7.52
8.19
8.50
8.68
8.81
One-half
pH decrease
at gill
Cyanide Series
0.01
0.04
0.10
0.17
0.29
0.45
lonization
factor for
gill pH3
0.997
0.985
0.932
0.871
0.816
0.767
Free cyanide
or dissolved
sulfide 96-hr
LC50, yg/1
124
117
133
133
152
157
HCN or H2S
at gill to
give 96-hr
LC50,yg/l
124
115
124
116
124
120
Mean 120
SD 4.2
Sulfide Series
6.462
7.101
7.698
8.151
8.430
8.693
46.88
23.40
10.79
6.13
4.55
3.83
6.41
7.04
7.57
7.87
8.01
8.10
0.05
0.06
0.13
0.28
0.42
0.59
0.789
0.467
0.205
0.115
0.0858
0.0709
64
133.
239
420
546
806
50.5
62.3
49.0
48.2
46.8
57.1
Mean 52.3
SD 6.1
Free C02 in test solution (Table 1) plus one-half increase in CO? at gill surface (Lloyd and Herbert
1960).
Calculated from average free C0£ at gill, using nomograph for evaluation of C02 (APHA 1975).
3 Factor = 1/(1 + 1QPH ' PK*) when pKHCN + 9.328 and pKH2S = 6.983 at 20 C.
-------
carbonate-bicarbonate system absorbs a considerable amount of C02 produced
by respiring fish, and the titration alkalinity increases as a result of
NH4+ excretion in conjunction with active ion exchange for Na+ (Dejours
et al., 1968). Holeton and Randall (1967) determined that the P^2 of the
blood in the vental aorta was greater by about 1.0 mm Hg than that in the dor-
sal aorta of rainbow trout both in well-aerated water and in a hypoxic
environment. Rahn (1966) stated that the change in C02 tension of the water
at the gills could not be much greater than 5 mm Hg at 20 C, and it could be
that great only if nearly all the 63 in the water passing over the gills was
extracted.
The mechanism for the excretion of metabolic C02 by freshwater fish
includes the catalytic conversion to bicarbonate of some of the CC>2 in the
blood by carbonic anhydrase in the gill epithelium (Randall 1970). There-
fore, along with the free C02 entering the water, bicarbonate passes across
the gill epithelium by an active exchange mechanism which involves chloride.
Stumm and Morgan (1970) indicated that the hydration/dehydration reaction
of C02(aq) + HpO t H2C03 proceeds slowly in water, the establishment of the
hydration equilibrium at pH values near 7 requiring a finite time on the order
of many seconds. The formation of C02 from the bicarbonate actively excreted
by the gill epithelium is also slow. Water is generally considered to pass
over the gill epithelium in less than two seconds, and the hydration of CC>2
and formation of C02 from bicarbonate in water is on the order of many seconds.
Therefore, the major portion of the rise in PCQ? and ultimate pH shift at
equilibrium should occur after the water has left the respiratory surface.
To explain the cyanide and sulfide toxicity bioassay data by the theory
of Lloyd and Herbert, the total increases in C02 at the gill surface when
oxygen utilization is 10 and 55%, respectively, would need to average about
1.0 and 5.7 mg/1 or 0.44 and 2.5 mm Hg, respectively. These increases,
although physiologically possible, are greater for the sulfide bioassays than
those that have been reported in the literature (Randall 1970). The
accompanying maximum total pH change necessary for agreement with the theory
would need to be 0.9 and 1.2 units for the cyanide and sulfide bioassays at
the highest test pHs, respectively. These changes appear to be extreme, since
Holeton and Randall (1967) measured a 0.2 pH unit difference between water
samples from the buccal and opercular chambers of the rainbow trout. Because
of all the above considerations, it is concluded that the theory of Lloyd
and Herbert is not an appropriate explanation of the relation of toxicity
to pH observed in the reported cyanide and sulfide bioassays.
TABATA'S THEORY
It is apparent from examination of the data in Tables 2 and 3 and
Figure 1 that with an increase in test pH more free cyanide or dissolved
sulfide becomes necessary to produce the acute response. However, the
increase in concentration needed is not large enough to maintain a constant
96-hr LC50 concentration of molecular NCN or H2$. If the toxicity of free
cyanide and dissolved sulfide is attributable to the molecular component
only, the slopes of curves relating the proportion of weak acid present in
the molecular form and the toxicity to test pH should be parallel.
-------
Inspection of these relationships plotted in Figure 2 shows that this is not
the case. The discrepancy between the curves occurs mainly in the alka-
line region where the ratio of weak acid anions to total acid increases
rapidly with rise of pH. Therefore, it appears that the CM" and HS" anions
may have a toxicity equal to at least a fraction of the toxicity of the
neutral molecules. The theory of Tabata (1962), which assumed that not only
the molecular forms but also the ionized fraction can penetrate membranes
and have a measurable toxicity, may thus be appropriate for explaining the
toxicity of cyanide and sulfide solutions to fish.
The relationship expressing Tabata's theory where the total toxicity
is equal to the sum of the toxicities due to the molecular and ionic forms
can be represented by the equation:
= Tm [molecular form] + T. [ionic form] (3)
where C is the total concentration of molecular plus ionic forms, 1/LC50 is
an expression of total toxicity, and Tm and T-j are the molar toxicities of
the molecular and ionic forms, respectively. The ratio between the toxicity
of the molecular and ionic forms (Tm/T-j) can be derived from the LC50
determined at one pH and that (LC50') determined at another pH (pH1) in the
manner shown in Appendix A. This relationship for weak acids was defined
by Tabata as follows:
K
[H*l
[l + K 1
1 [H+1 J
f l + K
L m+ii
LC501 -
] LC50
K
ru i
1 n I
- i1 +
[i 4- K i reft
K ,
FH+1 ' LC50'
]
^-::- :- ; :': " (4)
where K is the acidic ionization constant. This equation was used to ana-
lyze the cyanide and sulfide bioassay data; the calculated T^/T^ values are
presented in Table 5. It was anticipated that the ratios would be fairly
constant, that is, independent of the pH. This relationship was essentially
the case for the cyanide bioassays, the overall mean ratio being 2.3. The
ratios calculated for the sulfide bioassays were somewhat variable but fairly
constant when the pH values were between 7.7 and 8.4. The calculated Tm/T.j
values within this pH range average about 15, when the value 6.49 is
omitted. Therefore, the effective toxicity to the fathead minnow of the HCN
and H2$ molecules in solution is apparently about 2.3 and 15 times that of
CN~ and HS~ anions, respectively.
The above equation (4) can be modified to estimate a new LC50' for
a new pH1 from the given pH, LC50, and Tm/Tj values (Appendix A). This
relationship for weak acids was defined by Tabata as follows:
1QO
-------
1.00
0.80
0.60
0.40-
0.20-
O.IO<—
6.0
Hi PROPORTION HCN
TOXICITY
e.s
7.0
7.8
8.0
6.9
9.0
11.00
0.80
0.60
0.40
020
s
i
x
8
o.K>
9.9
1.00
0.80
o.eo
0.40
in
w 0-2°
<
Ul 0.10
w O.08 •
O °'°*
F
9 0.04
0.02
0.01
Hi PROPORTION H2S
HI TOXICITY
e.s
7.0
7.9
8.0
8.9
9.0
TEST PH
,1.00
•0.80
0.80
0.40
0.20
•010
0.08 »
0.06 3
0,4 J
X
0.02 O
0.01
9.9
Figure 2.
Toxicity of free cyanide or dissolved sulfide to the fathead
minnow and the proportion of each present in the molecular form
(as HCN or H2S) in relation to test pH at 20° C.
-------
Table 5.
Calculated T /T. ratio
different leVeli of pH
T. ratios and predicted LC50 values for cyanide and sulfide bioassays at
o
ro
Test
no.
pH
1
VTi
2
ratio
for
3
Cyanide
1
2
3
4
5
6
1
2
3
4
5
6
6
7
8
8
8
9
6
7
7
8
8
8
.830
.559
.286
.672
.974
.262
.462
.101
.698
.151
.430
.693
_
0
5
1
2
1
.
-7
17
16
16
22
_
.18
.75
.60
.51
.82
.
.22
.3
.8
.8
.9
.
-
-1
3
4
2
6
10
11
16
_
-
.40
.45
.42
.26
„
—
.49
.2
.6
.9
1
2
1
Sul
16
16
25
--
--
.01
.14
.63
fide
—
—
.3
.6
.5
test number
4 5
Series
..
— __
— __
__ __
6.08
1.95 1.23
Series
—
__ _-
— _-
17.0
35.2 77.3
Free cyanide or
dissolved sulfide
96-hr LC50, yg/12
Determined
124
117
133
133
152
157
64
133
239
420
546
806
Predicted
120
120
125
133
145
162
66
111
239
412
526
615
1
Ratio applies to comparison of the test indicated by the test number below with the test at a higher
pH whose test number appears in the first column (equation 4). The K^CN values at 20 C are
4.700 x 10~10 and 1.041 x 10~7, respectively.
Values based on Tm/T.j of 2.3 or 15, and mean of tests 1
bioassays, respectively.
and 2 or test 3 for cyanide and sulfide
-------
[ 1 +
f l +
K
[H+T J
K
PUT J
i n J
rv
L T.
r m + •
[ T.
K
fH+]
K
run
In j
1
j
• J
LC501 = Lii_i_J ^ LILJ . LC50 (5)
The predicted LC501 values (Table 5) for the cyanide bioassays are estimates
based on the mean pH and LC50 values for tests 1 and 2 and a Tm/T.j ratio of
2.3. The predicted LC50' values for the sulfide bioassays are estimates
based on the LC50 experimentally determined at pH 7.7 and a Tm/Tj of 15. As
seen in Table 5 and Figure 3, the predicted and determined 96-hr LC50 values
generally show good agreement, but at the highest pH value of the sulfide
series (pH 8.7) the determined value for dissolved sulfide is decidedly lar-
ger than the predicted one.
The percentages of total toxicity attributable at different pH values
to the molecular and ionic forms can be calculated from the expressions
fm/{Pm + P^W}] [100] and [P1-/{Pi + Pm(Tm/T1)}[100]respect1vely,
where Pm and P^ represent the proportions of free cyanide or dissolved sul-
fide in the molecular and ionic forms, respectively. These toxicity rela-
tionships, as presented in Figure 4, were calculated assuming T^T-,- ratios
for the cyanide and sulfide bioassays of 2.3 and 15, respectively. The
anions contribute a larger proportion of the total toxicity with increasing
pH, At pH levels below 9.5, the contribution of the HS~ ion is greater than
that of CN~, even though CN~ is not as much less toxic than HCN as the HS"
ion is less toxic than HpS. An appreciable deviation from the theoretical
relationship, indicated by line A in Figure 4, was noted only in the sulfide
bioassay series when the pH was relatively high.
In order to explain more fully the variable relation between the
toxicity of solutions of the weak acids and the concentrations of the molecu-
lar forms, one should consider the concentrations and forms of the toxicants
not only in the test solutions but also in the body fluids. Carbon dioxide
is known to diffuse readily through tissues, and the C02 tension (Pc02^
in blood vessels efferent to the gills of fishes approximates that of the
water in the buccal cavity (Stevens and Randall 1967). An increase in C02
content of the water thus produces an increased CO? content of the blood.
Plasma proteins and hemoglobin buffer the H+ deriving from the dissociation
of carbonic acid, which is formed when C02 is hydrated. The buffering
capacity is limited; and, as Albers (1970) has stated, the linear relation-
ship between fish blood pH and log Prr^ varies in slope, depending on the
buffering capacity. It can be determined from Figure 11 of his review that,
in the carp (Cyprinus carpio), as the blood Pc02 is increased from about 2
to 12 mm Hg, the pH decreased from about 7.9 to 7.4. Ferguson and Black
(1941) determined that, at 15 C, as the blood PCQ? of the carp was increased
from 2 to 20 mm Hg, the plasma pH of oxygenated blood decreased from 7.91
to 7.23. A comparable decrease for the rainbow trout was from 7.66 to 7.15.
In a study by Hunn (1972) on the effect of thanite on the blood chemistry of
the carp, a decrease in 62 and glucose utilization and accumulation of lactic
acid resulting from blockage of the electron transfer chain by cyanide
-------
160
o
er
UJ
120
UJ
O
o
80-
40
O
6 JO
6.5
7.5
8.0
8.5
9.0
9.5
8OO
CO
CM
X
-------
100
80
60
tr
2 40
§ 20
HCN
TM/T[ = 2.3
20
40
60
80
m
3
m
i
m
6.5
7.0
7.5
8.0
8.9
9.0
IOO
O
UJ
60
40
20
6.0
6.5
H2S
HS'
A
20
40
6O
60
00
O
ft
O
O
s
7.0
7.8 8.0
TEST PH
S.5
9.0
9.5
Figure 4. Relations between pH and the percentages of total toxicity
attributable to the molecular and ionic forms of cyanide and
sulfide in bioassays with the fathead minnow at 20° C. Deviation
of experimental data from the theoretical relationship is
indicated by Line A.
105
-------
poisoning were noted. The increase of lactic acid levels was reflected in a
decrease of blood pH from 7.85 (level found in controls) to 7.2 in exposed
fish. Therefore, the blood pH of fathead minnows exposed to cyanide and
sulfide test solutions probably decreased with increasing ambient C02
tensions, decreasing test pH, and accumulation of lactic acid in the blood
due to poisoning. Changes in plasma pH parallel fairly closely changes in
cellular pH, and, if there is a difference, intracellular pH is usually
lower than that of the blood.
It is generally recognized that molecular forms penetrate membranes
more readily than charged ions do, and that blood levels should increase
in fish concurrently with increases in ambient concentrations. Assuming the
molecular forms are the major internal toxicants, then an explanation for
the observed relationship between test pH and cyanide or sulfide toxicity must
include consideration not only of penetration of the gill epithelium mainly
by the molecular forms, but also of variations in internal ionization with
changes of blood and intracellular pH associated with changes of internal
C02 tension and lactic acid concentration. Warren and Schenker (1962) pro-
posed that there is a marked difference in effect on ammonia toxicity to
mice between equivalent plasma pH changes produced by either strong acids
or bases and free C02« Infused strong acids or bases will penetrate tissue
barriers poorly, thus causing a change in the extracellular fluid-
intracellular fluid pH gradient with the redistributed NHj being trapped as
NH4+ on the side of lower pH. With C02» the pH gradient is less marked
since the pH is changed on both sides of the membrane almost simultaneously
because C02 crosses membranes with ease. Consequently, NH3 will tend to
redistribute less extensively, The distribution between ammonia levels in
fish and in their aqueous environment was proposed by Warren and Schenker
(1962) as a possible explanation of Lloyd and Herbert's (1960) data which
demonstrated an increased apparent toxicity of un-ionized ammonia with
reduction in test pH from increased ambient free C02, However, because
fish blood is buffered against large pH changes, at high ambient C02
tensions and low pH a gradient would favor NH3 diffusing from the blood into
and trapped as Nfy* in the more acidic test medium. This effect should
decrease the toxicity of un-ionized ammonia with increases in ambient free
C02. Warren and Schenker1 s theory does not adequately explain our results,
since it is anticipated that by lowering the test pH with strong acid the
apparent toxicity of the weak acid molecular forms should increase because
the diffusion gradient would favor their penetration of the gill epithelium.
Instead, the toxicity of the molecular forms decreased with decreasing test
pH. Warren and Schenker apparently failed to realize that when strong acid
is added to water with a high bicarbonate alkalinity the free C02 concen-
tration is substantially increased,
A change in the permeability of the gills to molecular HCN or H2$ may
also contribute to the apparent change in toxicity, but an extent of this
change sufficient to entirely account for the decrease in molecular form
LC50 values, especially in the sulfide series of tests, is most unlikely.
106
-------
Cyanide and sulfide each has a specific inhibitory effect upon certain
enzymatic processes. According to Hewitt and Nicholas (1963), cyanide may
react with metal enzymes as HCN or as CM", but the general consensus is that
various types of inhibition mainly involve the HCN molecule. Sulfide,
including the anionic species, can inhibit certain enzymes by formation
of complexes with essential metals contained in the enzymes. Since the
molecular forms can move across internal tissue barriers more readily than
charged ions do, it seems reasonable to suppose that the effectiveness of
cyanide or of sulfide as an internal poison becomes greater with an increase
in the proportion existing in the molecular form. Changes in blood pH will
thus affect the toxicity by changing the concentration of the freely pene-
trating molecular form.
If the CM" and HS" anions penetrate the gills along with the molecular
forms, the toxicity of cyanide and sulfide solutions to fish should not
be entirely determined by the ambient molecular levels. Internal conditions
should also be taken into consideration. At the pH of blood of the fish in
all the cyanide test solutions, high percentages of the free cyanide which
penetrated the gills must have existed as HCN. Thus, when the calculated
Tfj/T-j ratio of 2.3 was used, the predicted 96-hr LC50 values for free cya-
nide bioassays showed good agreement with the determined values (Table 5).
For sulfide, predicted values based on a T^T-,- ratio of 15 and tests at
pH 7.7, at which the blood pH is supposed to be near that of the test
solution, good agreement between determined and predicted 96-hr LC50 values
was observed, except at pH greater than 8.4. The apparent anomaly may be
explained as having been due to the pH of the blood having been slightly
higher in fish tested in the more alkaline solutions than in those tested
at pH 7.7 and 7.9 mg/1 free CC^. In the bioassays at high pH, more of the
sulfide presumably had to penetrate, largely as HS" ion, in order to pro-
duce a toxic effect equal to that produced at pH 7.7. Therefore, a greater
amount of dissolved sulfide than predicted was required to produce the toxic
response at the high pH. At a high enough pH, the presence of an internal
H2$ concentration higher than the level in the ambient medium is conceivable.
The concentration gradient across the gill surface then would allow some
H£$ to move by diffusion from the blood into the water, thus additionally
reducing the toxicity of the dissolved sulfide. It is thus concluded that
the acute toxicity to fathead minnows of free cyanide and dissolved sulfide
solutions does not depend entirely on the concentration of ambient molecular
HCN or H2$, but that the CN" and HS~ anions penetrate the gill epithelium
less readily than do the molecular forms and contribute to the toxicities
of these solutions increasingly as the pH increased.
The equations expressing the theory proposed by Tabata, with K as the
basic ionization constant and [H+] replaced by [OH"1, can be used to analyze
the ammonia bioassay data of Lloyd and Herbert (I960). The calculated Tm/T-j
ratios presented in Table 6 average 82 when the results for the tests at
pH 7.0 with a very high CC>2 level are omitted. Therefore, the effective
toxicity to the rainbow trout of the NH3 molecule in solution is apparently
about 82 times that of the NH4+ anion. The predicted LC50 values in Table 6
were estimated on the basis of the LC50 determined at pH 7.8 and a Tm/T-j
ratio of 82. Only at pH 7.0 was there a marked difference between the
107
-------
Table 6. Calculated Tm/T.j ratios and predicted LC50 values for ammonia bioassays with rainbow trout
at different pH levels and assumed mean temperature of 19 C (data from Lloyd and Herbert
(1960)
Determined
Test
no.
1
2
3
4
PH
8.20
7.80
7.37
7.00
Free
C02,
mg/1
3.2
7.7
21.5
48.0
Tm/Ti
1
56
81
337
i ratio for test number *
2 3
—
109
815 -1201
500 min LC50.2
mg/1
NHq
0.84
0.62
0.42
0.49
as N
Total
ammonia
15.2
27.2
48.8
133
Predicted
total
ammonia LC503
mg/1 as N
14.1
27.2
45.6
59.6
o
00
Ratio applies to comparison of the test indicated by the test number below with the test at a higher
pH whose test number appears in the first column. In the formula for computation of Tn/T-j, [H+] is
replaced by [OH-] and Kb = 1.695 x 10~5 at 19 C.
Assumed factor for proportion of total ammonia as molecular NH$ of 1/(1 + 10PKa " pH) with pKa at
19 C equal to 9.432 (Robinson and Stokes 1955).
3
LC50' calculations based on T^T^ of 82, test 2, and Kb, with [H+] replaced by [OH~] in the formula.
-------
determined and predicted total ammonia LC50 values. At this low test pH
associated with a high CC^ tension, the blood pH was most likely lower than
that of the fish tested at pH 7.8. The percentage of total ammonia present
in the molecular form is thus decreased, and if NH^ is the major internal
toxic form, the observed LC50 at pH 7.0 should be greater than the pre-
dicted value. Therefore, the correct explanation of ammonia toxicity may
be similar to that proposed for cyanide and sulfide, whose ionic forms
are believed to penetrate the gill and to have a measurable toxicity con-
siderably less than that of the respective molecular forms. However, because
of involvement of ammonia in active exchange at the gill, a complete expla-
nation may be more complicated, awaiting further physiological and toxico-
logical investigation.
109
-------
REFERENCES
Albers, C. 1970. Acid-base balance. Pp. 173-208 in W. S. Hoar and D. J.
Randall (eds.), Fish physiology. Vol. IV. The nervous system,
circulation, and respiration. Academic Press, New York, xvi + 532 p.
American Public Health Association et al. 1971. Standard methods for the
examination of water and wastewater. 13th ed. Am. Public Health
Assoc., New York, xxxv + 874 p.
American Public Health Association et al. 1975. Standard methods for
the examination of water and wastewater. 14th ed. Am. Public Health
Assoc., Washington, D.C. xxxix + 1193 p.
Bonn, E. W., and B. 0. Follis. 1967. Effects of hydrogen sulfide on
channel catfish, letalurus punetatus. Trans. Am. Fish. Soc.
96(1): 31-36.
Broderius, S., and L. L. Smith, Jr. 1977. Direct determination and
calculation of aqueous hydrogen sulfide. Anal. Chem. 49(3):424-428.
Brungs, W. A., and D. I. Mount. 1970. A water delivery system for small
fishholding tanks. Trans. Am. Fish. Soc. 99(4):799-802.
Davis, J. C., and J. N. Cameron. 1971. Water flow and gas exchange at
the gills of rainbow trout, Salmo gairdneri. J. Exp. Biol. 54(1): 1-18.
Dejours, P., J. Armand, and G. Verriest. 1968. Carbon dioxide dissociation
curves of water and gas exchange of water-breathers. Respir. Physiol.
5(l):23-33.
Dixon, W. J. (ed.). 1973. BMD; biotnedical computer programs. 3rd ed.
University of California Press, Berkeley. 733 p.
Doudoroff, P. 1976. Toxicity to fish of cyanides and related compounds;
a review. Ecol. Res. Ser. EPA-600/3-76-038. Environmental Research
Laboratory, Office of Research and Development, U.S. Environmental
Protection Agency, Duluth, Minn, vi + 155 p.
Doudoroff, P., and M. Katz. 1950. Critical review of literature on the
toxicity of industrial wastes and their components to fish. I.
Alkalies, acids, and inorganic gases. Sewage Ind. Wastes 22(11):
1432-1458.
Ferguson, J. K. W., and E. C. Black. 1941. The transport of COg in the
blood of certain freshwater fishes. Biol. Bull. 80(2): 139-152.
Finney, D. J. 1971. Probit analysis. 3rd ed. Cambridge University
Press, London. 333 p.
110
-------
Hewitt, E. J., and D. J. D. Nicholas. 1963. Cations and anions: inhibitions
and interactions in metabolism and in enzyme activity. Pp. 311-436 in
R. M. Hochster and J. H. Quastel (eds.), Metabolic inhibitors; a
comprehensive treatise. Vol. 2. Academic Press, London. 753 p.
Holeton, G. F.} and D. J. Randall. 1967. The effect of hypoxia upon the
partial pressure of gases in the blood and water afferent and efferent
to the gills of rainbow trout. J. Exp. Biol. 46(2}: 317-327.
Hunn, J. B. 1972. The effects of exposure to thanite on the blood chemistry
of carp. Prog. Fish-Cult. 34(2): 81-84.
Kutty, M. N. 1968. Respiratory quotients in goldfish and rainbow trout.
J. Fish. Res. Bd. Canada 25(8): 1689-1728.
Kutty, M. N. 1972. Respiratory quotient and ammonia excretion in Tilapia
mossambiea. Mar. Biol. 16(2): 126-133.
Kutty, M. N., N. V. Karuppannan, M. Narayanan, and M. Peer Mohamed. 1971.
Maros-Schulek technique for measurement of carbon dioxide production
in fish and respiratory quotient in Tilapia mossarrbioa, J. Fish, Res.
Bd. Canada 28(9): 1342-1344.
Litchfield, J. T., Jr., and F. Wilcoxon. 1949. A simplified method of
evaluating dose-effect experiments. J. Pharmacol. Exp. Ther.
96(2): 99-113.
Lloyd, R., and D. W. M. Herbert. 1960. The influence of carbon dioxide on
the toxicity of un-ionized ammonia to rainbow trout (Salmo gairdnepii
Richardson). Ann. Appl. Biol. 48(2): 399-404.
Mount, D. I., and R. E. Warner. 1965. A serial-dilution apparatus for
continuous delivery of various concentrations of materials in water.
Publ. 999-WP-23. U. S. Public Health Serv., Cincinnati, Ohio. 16 p.
Rahn, H. 1966. Aquatic gas exchange: theory. Resp. Physiol. 1(1): 1-12.
Randall, D. J. 1970. Gas exchange in fish. Pp. 253-292 in W. S. Hoar and
D. J. Randall (eds.), Fish physiology. Vol. IV. The nervous system,
circulation, and respiration. Academic Press, New York, xiv + 532 p.
Robinson, R. A., and R. H. Stokes. 1955. Electrolyte solutions. (Appendix
12 I, p. 496.) Butterworths Scientific Publ., London. 512 p.
She!ton, G. 1970. The regulation of breathing. Pp. 293-359 in W. S. Hoar
and D. J. Randall (eds.), Fish physiology. Vol. IV. The nervous
system, circulation, and respiration. Academic Press, New York.
xiv + 532 p.
Ill
-------
Smith, L. L., Jr., D. M. Oseid, G. L. Kimball, and S. M. El-Kandelgy. 1976.
Toxicity of hydrogen sulfide to various life history stages of bluegill
(Lepomis macroehirus). Trans. Am. Fish. Soc. 105(3):442-449,
Stevens, E. D., and D. J. Randall. 1967. Changes of gas concentrations in
blood and water during moderate swimming activity in rainbow trout.
J. Exp. Biol. 46(2): 329-337.
Stumm, W., and J. J. Morgan. 1970. Aquatic chemistry; an introduction
emphasizing chemical equilibria in natural waters. John Wiley & Sons,
New York. 583 p.
Tabata, K. 1962. Toxicity of ammonia to aquatic animals with reference to
the effect of pH and carbon dioxide. Bull. Tokai Reg. Fish. Res.
Lab. 34:67-74.
Warren, K. S., and S. Schenker. 1962. Differential effect of fixed acid and
carbon dioxide on ammonia toxicity. Am. J. Physio!. 203(5):903-906.
112
-------
APPENDIX A
DERIVATION OF EQUATIONS PROPOSED BY TABATA
Assuming a weak acid HA dissociates to give H+ and A", then
= and at one pHj m , Jk. and at pH'. J*Z Ka (1)
-
[HA] [HA] [H+] [HA] , [H+] ,
To account for the relationship between pH variations and weak acid
toxicity, Tabata proposed the following formula whaere the toxicity of 1 mole
of molecular form (Tm) and that of 1 mole of ionized form (Tj) is given by:
= [HA] • Tm + [A'] • Ti (2)
The left hand side of the equation (2) expresses the total toxicity of
the weak acid while the first term on the right hand side stands for the
toxicity due to the molecular form; the second term expresses the toxicity
attributable to the ionic form. C is the proportionality constant.
The ratio of the toxicity of the molecular form to that of the ionized
form (T,p/Ti) can be obtained from equations (1) and (2) by determining
the LC50 at one pH, here referred to simple as pH, and the LC50' at another,
here designated by the symbol pH1.
At pH C = [HA] + [A~] and at
(3)
pH' C1 = [HA]1 + {A-]1
113
-------
Therefore: = [HA] • T + [A~] • T. and
n
- [HA]' -IA']' -T
LC501 m
By equating the two expressions above, the relationship for T /T. can be
defined as follows:
HA + A" = [HA]1 + [A-]'
LC50([HA] • Tm + [A-] • T.) LC50'([HA]' • Tm + [A~]'- T.)
LC50'([HA]' . Tm + [A']1 • 1.)
LC50([HA] . Tm + [A"] • T.) [HA] + [A~]
([HA] + [A-])[LC50'([HA]' . Tm + [A']1 'I.)] =
([HA]' + [A-]')[LC50([HA] • Tm + [A~] • T.
([HA] + [A-]) ([HA]1 • LC501 • T +[A-]' • LC501 • TI) =
([HA]1 + [A']')([HA] • LC50 - Tm + [A~]• LC50 - TI
([HA] + [A-])([HA]1 • LC501 • Tj + ([HA] + [A-])([A-]' • LC501 • 1^
([HA]1 + [A-]')([HA] • LC50 • Tm) + ([HA]1 + [A-]')([A-] • LC50 - T^
([HA] + [A'])([A']1 - LC501 • T^)- ([HA]' + [A"]'}([A-] • LC50 -T^) =
([HA]1 + [A-]')([HA] • LC50 - Tj - ([HA] + [A-])([HA]' • LC50' - TJ
([HA][A']1 • LC501 • T.) + ([A'][A']1 • LC50' • T.) -
([HA]'[A-] • LC50 - T.) - ([A-/[A'] ' LC50 • T.) =
([HA]1 [HA] • LC50 - Tj + ([A']'[HA]- LC50 «Tm) -
([HA][HA]1 • LC501 • Tm) - ([A"][HA]1 • LC501 • T )
114
-------
T.[([HA][A"]'-LC50') + ([A"][A"]'-LC501} -
([HA]'[A"]'LC50) - ([A~]'[A~]-LC50)] =
Tm[([HA]'[HA]'LC50) + ([A~]'[HA]'LCBO) -
([HA][HA]'-LC50') - ([A"][HA]'-LC501}]
([HA][A~]'-LC50') + ([A'JCA'J'-LCSO') -
JH = (fHAl'fA>LC50) - ([A"TrA"]-LC50)
T. ([HA]'[HA]«LC50 + ([A~]'[HA]-LC50) -
([HA][HA]'-LC50') - ([A"][HA]'-LC501)
lU ([A"][A"]')(LC50' - LC50) + [HA][A~]'-LC50' - [HA]' [A"]-LC50
T. ([HA]'[HA])(LC50 - LC501) + [A"]'[HA]'LC50 - [A"][HA]'-LC501
1A HA ]' (|_C50' . LC50) + -^1^- -LC50' - - -LC50
^m. = [HA]' [HA] [HA]' [HA]
Ti LC50 - LC501 + &^- -LC50 - -^-L -LC501
[HA]' [HA]
K, K K K
T —T- ' —T- (LC501 - LC50) + —!- ' LC501 - ~ • LC50
+ +'+'
T.
l
Tm
T.
i
K
LC50 - LC501 + — f-
[H+]'
K K K
a «i pqn '+ a ^
, LUOU T • ,
[H ]' [H ][H ]'
K
i rt;n + a. T-
K
•LC50 - —~ - LC501
[H]
K K K
* i r^n1 a • i P^O ,, a j r^n
[H ] [H][H]'
K
. i c^n _ i rein1 _ .._.£.._ • i rt;n'
—-
[H ]' [H+]
115
-------
m _
'I + a. }|_C50' - -f—U + -|- HC5°
[H ] [H] [HI'
T,
(1
—|- ) LC50 - (1 -t—|—}LC50(
[H]' [H+]
(5)
The above equation (5) can be modified to estimate a new LC501 for a
new pH1 from the given pH» LC50, and "L/T. value.
[H 31 [H ]
a «
)LC50'.T. --f— (1 •»•— !— )LC50-T. =
T [H ] [H ]' 1
,
LH J
L" J
K K K Ka K,K,
—| LCSO'.T. + . a a. .LC50'T. 1- -LC50.T. -2-2- -LCSO-T
[H ]' 1 [H ]'[n n [H1"] 1 [H""
K. K
LC50*T + —|~ -LCSO'T - LC50'«T |-
m ru+ii w m rU+,
-LCSO'-T,
m
Ka K K K
•LCSO'-T, + . a a . 'LCSO'.T. + LC501«T + — 1- *LC50'«T
[H]1
LC50-T +
m
LC50.T
-LC50-T. +
LC50*T.
116
-------
Ka KA
LC501 ( —^- -T. + T^-
KK
LC50 (T + —f— .T + —£- -T. + T^-ST T.)
m +, m + T + +, i
CL +-4- ' T + —|- -I, +
K K
LC501 =
LC50
m
T K
m . a
/_m . a t IN , a ^_
LC501 = —J—tiLT i—OiJ—
K K
a a
LC50
i peril -
LlyOU
(]
K.
'f [HH
K
[HH
T I/
i \ / m h a i
"]' 1 [H+] '
> \ / m , a
'] T1 [H+]'
I
. i rso
)
(6)
117
-------
THE ACUTE TOXICITY OF NITRITE TO FISHES
R. C. Russo and R. V. Thurston
Fisheries Bioassay Laboratory
Montana State University
Bozeman, Montana 59715
INTRODUCTION
Only in the past few years has nitrite received much attention in
toxicity studies by fisheries biologists, perhaps because it is generally
present in only trace amounts in most natural freshwater systems. It is an
intermediate product in the conversion of ammonia to nitrate by the nitrifi-
cation process. In this process nitrosomonads convert ammonia to nitrite,
and nitrobacters convert nitrite to nitrate. In a relatively stable situa-
tion, the first conversion, i.e. that of ammonia to nitrite, is the rate-
limiting step in the total process. However, if something occurs to disrupt
the stability of the process, such as a malfunction at a sewage treatment
plant, or extremely low ambient temperatures, then nitrite may be discharged
into, or produced in, the receiving water at a level which may be toxic to
fishes. Water reuse systems using the nitrification process may also mal-
function, resulting in increased nitrite levels in the treated water.
Although these increased nitrite levels may be a short-lived phenomenon,
nonetheless the highly toxic nature of nitrite to fishes warrants considera-
tion.
Anthonisen and coworkers (1976) have demonstrated that a nitrite buildup
can occur due to inhibition of the nitrification process by nitrous acid and
un-ionized ammonia. Un-ionized ammonia inhibits nitrobacters at much lower
concentrations than those at which it inhibits nitrosomonads (0.1 -1.0 vs.
10 - 150 mg/1). Nitrous acid inhibits both nitrobacters and nitrosomonads
at concentrations between 0.22 and 2.8 mg/1. When nitrite oxidation is
inhibited, incomplete nitrification is observed, resulting in nitrite accumu-
lation (Anthonisen et al. 1976).
Nitrite concentrations of 30 mg/1 N02-N and higher have been reported by
Klingler (1957) in receiving waters for effluents from metal, dye, and
celluloid industries. McCoy (1972) has reported levels up to 73 mg/1 NO^-N
in Wisconsin lakes and streams. In a reasonably clean cold water stream in
Montana, we have occasionally found levels around 0.1 mg/1 N02~N below a
sewage treatment plant (Russo and Thurston 1974).
118
-------
Until recently, only a small amount of information had been published on
the toxicity of nitrite to fishes, and most of that literature dealt with
static bioassays of 48 hours or less. However, in the last two years a
greater number of papers have appeared. The available literature information
is summarized in Table 1. There is a wide variation in the toxicity results
reported. This range is probably attributable both to differences in water
chemistry among the different investigators' experiments and to some genuine
differences in susceptibility among fish species. McCoy (1972) tested 13
species, presumably using the same dilution water in all cases, and observed
a wide variation in susceptibilities. There is also some indication that
younger fish may be somewhat less susceptible to nitrite than are older fish
of the same species (Russo et al. 1974, Smith and Williams 1974).
It is known that nitrite oxidizes hemoglobin (Hb) to methemoglobin
(MetHb) (Bodansky 1951, Jaffe 1964, Kiese 1974). Thus, one way that nitrite
is toxic to fishes is through formation of excessive amounts of MetHb which,
unlike Hb, is incapable of transporting oxygen; MetHb in sufficiently high
concentrations in fish blood can cause death. The percentage of total hemo-
globin which is MetHb under normal conditions has been reported by Cameron
(1971) to be 2.9% for wild rainbow trout (Salmo gairdneri) and 17% for
hatchery-reared rainbow trout. Shterman (1970) has reported MetHb levels
(as percent of total hemoglobin) in rainbow trout to be 2.7 - 3.9%; Brown
and McLeay (1975) have reported levels of 0.9%, and Smith and Russo (1975)
have reported levels of 3.6%. These values have been detected in the species
when not under stress from environmental nitrite.
Smith and Williams (1974) observed elevated levels of MetHb in rainbow
trout and Chinook salmon (Oncorhynekus tshauytscha} exposed to nitrite. We
have measured MetHb levels in 30-g rainbow trout exposed for 1-8 days to
nitrite concentrations from 0.1 to 0.78 mg/1 NO^-M and found an increase in
MetHb concentrations even at the lowest nitrite exposure: 14.3% MetHb (of
total Hb) vs. 3.6$ for controls (Smith and Russo 1975). Our results were
comparable to those of Brown and McLeay (1975) who also found a significant
rise in MetHb in rainbow trout exposed to 0.015 mg/1 NOp-N for four days;
mortalities were observed at 0.2 and 0.3 mg/1, at which 80% of the fish
blood Hb was in the MetHb form. Brown and McLeay also observed a decrease in
total Hb at NO^-N concentrations above 0.1 mg/1. In contrast to the above
results, Cameron (1971) observed no change in MetHb content of rainbow trout
blood after two days' exposure to 2 mg/1 N02' (0.61 mg/1 N02-N).
This paper reports on some additional acute toxicity studies we have
conducted with rainbow trout, including an investigation of the effect of
chloride ion on nitrite toxicity. Some data on the toxicity of nitrite to
fathead minnows (Pimephales promelas) and mottled sculpins (Coitus bairdi]
are also presented.
MATERIALS AND METHODS
The bioassays were conducted either in plastic tanks containing 64
liters of water with a replacement time of 5-6 hours and using proportional
diluters (M9unt and Brungs 1967) for toxicant delivery, or in fiberglass
tanks containing 350 liters of water with a replacement time of 1.3 hours and
119
-------
TABLE 1. SUMMARY OF LITERATURE DATA ON NITRITE TOXICITY TO FISHES.
Reference
Gillette et al.
(1952)
Wallen et al.
(1957)
Klingler
(1957)
McCoy
(1972)c
.
ro
o
Smith and
Williams
(1974)
Uestin
(1974)
Fish and Size
creek chub (Semotilus a.
atromaaulatus) , 3-4 inches
mosquitofish (Gambueia
affinie), adult female
minnow (Phoxinus laevis) ,
5-8 on
logperch (Percina caprodes)
brook stickleback (Culaea
inconstans)
carp (Cyprinus aarpio)
black bullhead (ictalurus
melae)
common white sucker (Cato-
etomus aormeraoni)
quillback (Carpiodes eyprinus)
rainbow trout (Salmo gaird-
neri) 100 g
4.5 g
Chinook salmon (Onoorhynohus
tehauytesha) 32 g
Chinook salmon (Oncorhynchus
tahawytecha) 1.50-10.55 g
Type of
Bioassay
static
static
»
partial
static
II
static
11
"
11
"
«
-
"
flow-
through
»
"
partial
static
H
Temp.
(°C)
15-21
21-24
••
18-21
«
NR
NR
NR
NR
HR
NR
NR
NR
10
"
"
13.6-15.6
II
pH
8.3
7.1-7.5
"
NRb
NR
NR
NR
NR
NR
NR
NR
NR
NR
7.9
"
"
6.8-7.2d
"
Type of Water
(mg/liter)
Hardness
98.0
Alkalinity
<100
Turbidity
120-140
»
NR
NR
NR
NR
NR
NR
HR
NR
NR
NR
Hardness
200
"
"
NR
NR
N02-N
(mg/liter)
80-400a
1.6a
1.5a
10a
2030a
5
5
40
100
40
100
100
100
0.55
1.60
0.50
0.88a
0.73a
Results Reported
critical range
24-hr LC50
48- and 96-hr LC50
fatal in 14 days
fatal in 1.5 hr
mortality in <3 hr
mortality in 3-5 hr
no mortality in 48 hr
mortality in 45 hr
no mortality in 48 hr
mortality in 24 hr
survived for 48 hr
survived for 36 hr
55% mortality in 24 hr
50% mortality in 24 hr
40% mortality in 24 hr
96- hr LC50
7-day LC50
(continued)
-------
TABLE 1. Continued.
Reference
Colt (1974), Colt
and Tchobanoglous
(1976)
Russo, et al .
(1974)
Brown and
McLeay
(1975)
Konikoff
(1973, 1975)
Thurston, et al.
(To be submitted)
Fish
and Size
channel catfish (letalurus
vunctatus) , finger lings
rainbow trout
jairdneri)
rainbow trout
gairdneri)
(Salrrv
2 g
12-14 g
235 g
12 g
(Salm>
9 1
channel catfish (ictalwms
punctatus) 40 g
cutthroat trout (Salmo
clarki) 1 g
3 g
1-3 g
Type of
Bioassay
static
flow-
through
ii
H
"
flow-
through
static
flow-
through
H
"
Temp.
(°C)
30
10.8
11.6-12.6
9.5
12.4
12
21
12.4
11.3,12.1
11.8-12.4
Type of Water
pH (mg/liter)
8.6-8.8 Alkalinity
220
7.9 Alkalinity
176
Hardness
199
H H
M H
H H
6.4-6.7 Alkalinity
2-8
Hardness
3-9
7.4-7.8 Alkalinity
60-70
7.85,7.88 Alkalinity
178
Hardness
200
7.88,7.80
7.80-7.88
NOp-N
(mg/liter)
13a
0.39
0.19-0.27
0.20
0.14-0.15
0.23
7.5a
0.38,0.37
0.56,0.48
0.4
Results Reported
96-hr LC50
96-hr LC50
96-hr LC50
96-hr LC50
Asymptotic LC50 (8-19 days)
96-hr LC50
96-hr TLm
36-day LC50
96-hr LC50
Asymptotic LC50
Calculated from NaNO^ or W)^ data reported by author.
NR = not reported by author.
°The following species were also tested, and survived less than 12-24'hr in 20-40 mg/liter N02-N: Johnny darter (Etheoetoma nitrun),
bluegill (Lepomie maerochims), pumpkinseed (Lepomie gibbosns), spotfin shiner (Notropie spilopterus), sand shiner (Notropie etramineus),
hog sucker (Hypentelium nigrieana), stonecat (Notwnts flovus); all fishes tested were small fingerlings or minnows.
D. T. Uestin, personal communication.
-------
using metering pumps for toxicant delivery. Reagent grade NaNQ? and NaCl
were used. Five test tanks plus a control tank were used in all cases, with
either 10 or 20 fish in each tank. The rainbow trout used were hatchery-
reared fish obtained from the Bozeman (Montana) Fish Cultural Development
Center, U. S. Fish and Wildlife Service. They were reared to test size in
water from the same ground spring source as that which was subsequently used
as the test dilution water. The fathead minnows were obtained from the
Miles City (Montana) Hatchery, U. S. Fish and Wildlife Service. The sculpins
were collected from Rocky Creek (Gallatin County), Montana. Fish were
acclimated to the test tanks for at least two days prior to toxicant introduc-
tion and were not fed during acclimation or testing. Fish which died during
the test were individually weighed and measured within 0-8 hours. Survivors
were measured at the termination of the test.
Nitrite concentrations were determined by the method described by EPA
(1974). Dissolved oxygen was measured either using the azide modification
of the iodometric method (American Public Health Association 1976) with
phenylarsine oxide substituted for sodium thiosulfate, or with a Yellow
Springs Instrument Co. Model 54-RC meter. Temperature was measured with a
certified thermometer, and pH with a Beckman Phasar-I meter. All other
chemical analyses were performed according to the procedures of the American
Public Health Association (1976). All colorimetric measurements were made on
a Varian 635 ultraviolet-visible spectrophotometer.
Averages and ranges of values for the tank water over all tests were:
dissolved oxygen 8.9 (7.9 - 10.0) mg/1; alkalinity 177 (171 - 191) mg/1 CaCOs;
hardness 199 (188 - 207) mg/1 CaC03; NH^-N 0.00 (0.00 - 0.07) mg/1; N03-N
0.08 - 6.85 mg/1; Cl~ 0.35 (0.00 - 0.74) mg/1. The range of N02-N concentra-
tions for all tests, and the Cl" concentrations for those tests in which Cl~
was a test variable, are reported in Tables 2 and 3. Temperature and pH
values for each test are also given in Tables 2 and 3.
Median lethal concentration (LC50) values and their 95% confidence limits
were calculated from the experimental data using the trimmed Spearman-Karber
method (Hamilton et al. 1977).
RESULTS AND DISCUSSION
The results of five 96-hr bioassays on rainbow trout are presented in
Table 2 and Figure 1. The 53- and 60-g fish were from the same lot; the 21-,
24- and 188-g fish were from a second lot. Although the two lots of fish
were tested approximately 10 months apart, there is good agreement among all
five tests, with 96-hr LC50 values ranging from 0.19 to 0.28 mg/1 N02-N. In
an earlier report on nitrite toxicity (Russo et al. 1974), four other 96-hr
bioassays on rainbow trout (size range 12 - 235 g) gave LC50 values of 0.19 -
0.27 mg/1 N02-N. Although the fish sizes in that earlier paper covered a
wider range than those reported here, the LC50 values were within the range
reported in the present study. The average 96-hr LC50 for all nine tests on
rainbow trout within the size range 12 - 235 g is 0.24 mg/1 N02-N (range
0.19 - 0.28). The highest concentrations tested where no mortalities occurred
ranged between 0.06 and 0.13 mg/1 N02-N. However, earlier work (Russo et al.
1974) on 2-g and sac fry rainbow trout showed that these smaller fish were
122
-------
TABLE 2. RESULTS OF ACUTE NITRITE BIOASSAYS ON RAINBOW TROUT (SALMO GAIRDNERl), FATHEAD MINNOW (PIMEPHALES PROMELAS),
AND MOTTLED SCULPIN (COTTUS BAIRDl).
ro
co
Test
No.
323
326
243
244
423
346
349
315
318
319
348
Avg
Wt (g)
20.6
24.3
53.1
60.5
188
2.3
2.3
1.8
2.0
2.3
1.6
Fish Size
Length (cm)
11.8
12.3
15.7
16.6
23.6
6.2
6.4
5.4
--
__
5.2
Concentration
Range Tested
(mg/2. N02-N)
0.22-1.70
0.08-0.59
0.08-0.48
0.11-0.78
0.10-0.79
2.26-7.29
2.30-7.52
0.82-2.66
2.68-8.75
8.41-26.3
21.6-66.7
LC50 (95% C.I.)
(mg/d NO?-N)
72 hr
RAINBOil TROUT
0.29
(0.24-0.36)
0.32
(0.27-0.38)
0.35
(0.29-0.41)
0.30
(0.25-0.36)
0.22
(0.17-0.28)
FATHEAD MINNOW
5.54
(3.G6-7.95)
3.94
(2.37-6.55)
MOTTLED SCULPIN
No mortalities
No mortalities
No mortalities
No mortalities
96 hr
..
0.28
(0.24-0.33)
0.27
(0.22-0.33)
0.27
(0.23-0.32)
0.19
(0.15-0.25)
2.99
(2.35-3.81)
2.30
in 96 hr
in 96 hr
in 72 hr
in 154 hr
Temp (C)
Avg (Range)
10.1
(10.0-10.2)
10.2
(10.1-10.3)
9.8
(9.7-10.0)
9.8
(9.7-9.9)
10.4
(10.3-10.7)
13.0
(12.7-13.2)
12.7
(12.5-12.8)
12.8
(12.3-13.9)
13.1
(12.6-13.6)
13.2
(12.6-14.3)
13.6
(13.1-14.1)
pH
Avg (Range)
8.05
(7.94-8.12)
8.10
(7.99-8.33)
7.68
(7.58-7.79)
7.76
(7.71-7.83)
7.81
(7.76-7.86)
8.05
(8.03-8.09)
8.04
(7.96-8.10)
8.10
(7.93-8.19)
8.06
(8.03-8.13)
8.14
(8.09-8.19)
8.08
(8.00-8.20)
-------
TABLE 3. RESULTS OF ACUTE NITRITE BIOASSAYS ON RAINBOW TROUT (SALMO GAIRDSERl) WITH ADDITION OF CHLORIDE ION.
ro
Test
No.
357
362
363
366
369
377
Avg Fish Size
Wt (g) Length (cm)
69.5
69.1
79.0
86.4
99.3
113
16.7
17.0
17.6
18.4
19.4
20.0
Concentration
Range Tested
(mg/A N02-N)
0.
0.
0.
1.
2.
4.
16-1.08
88-5.99
97-6.74
56-10.76
57-18.33
92-33.90
Cl" Concn
(mg/A)
1.2
5.1
10.4
20.2
40.9
40.8
96-hr LC50 (95% C.I.)
(mg/j, N02-N)
0.46
2.36
3.54
6.69
12.2
12.5
(0.
(1-
(2.
(5.
(8.
(9.
36-0.58)
86-3.00)
30-4.32)
54-7.93)
06-18.4)
96-16.0)
Temp (C)
Avg (Range)
10.4
(10.3-10.
, 10.4
(10.4-10.
10.4
(10.2-10.
10.5
(10.4-10.
10.3
(10.0-10.
10.4
5)
5)
6)
5)
5)
PH
Avg (Range)
7 92
(7.87-7.99)
8.01
(7.96-8.05)
7.90
(7.83-7.96)
7.84
(7.80-7.88)
7.74
(7.71-7.78)
7.69
(10.2-10.5) (7.57-7.72)
-------
1.0
ro
en
0.8
en
e
. 0.4
O
«0
O
0.2
Test
(323).
(244)^
— I88g
TIME, DAYS
Figure 1. Acute toxicity of nitrite to rainbow trout (Salmo gairdnem-} (pH 7.7-8.1, temp.
9.8-10.4°C, Cl" <_0.4 mg/A).
-------
somewhat less susceptible to nitrite. Also, 3-g cutthroat trout (Salmo
elavki) tested under similar conditions were found to have 96-hr LC50 values
averaging 0.52 mg/1 NO?-N, and values for 1-g cutthroats were even higher
(Table 1). c
On fishes other than trout, we conducted two bioassays on fathead minnows
and four on mottled sculpins. The results of these bioassays are summarized
in Table 2. These fishes are much less susceptible to nitrite toxicity than
the rainbow trout. The 96-hr LC50 values for fathead minnows were an order
of magnitude higher, averaging 2.6 mg/1 N02-N. Mottled sculpins were tested
successively using 3, 9, 26, and 67 mg/1 NC^-N as the highest test concentra-
tions. In all four of these bioassays no mortalities were observed throughout
the test periods (except for one anomalous death in 4 hours at 6 mg/1).
Having determined the toxicity of nitrite to rainbow trout under a given
set of water quality conditions, we were interested in investigating the
effects of variations in those conditions. Nitrite ion establishes the
following aqueous equilibrium:
N02" + H+ t HN02
This equation suggests that pH might have an effect on nitrite toxicity if
either of these two chemical species (HN02 or N02~) were more or less toxic
than the other, or if they acted synergistically or antagonistically. A pH
decrease would cause an increase in HN02 concentration.
For a total N02-N concentration of 1 mg/1 at pH 8.5, and using a pKa of
3.29, the N0"2~ concentration is 7.14 x 10"5 M and the HN02 concentration is
4.37 x 10"10 M. For the same total N02-N concentration at pH 7.5, the NO?"
concentration is 7.14 x 10"5 M and the HN02 concentration is 4.37 x 10~9 M.
For both of these cases, and for the pH range in between, the N02~ concentra-
tion remains essentially constant, whereas the HN02 concentration varies by
an order of magnitude. Even so, the N02~ concentration remains 4-5 orders of
magnitude higher than the HN02 concentration, and because total nitrite is
toxic at such relatively low levels, it is difficult to conceive that the
lesser of the two nitrite chemical species (i.e., HN02) is the principal toxic
form. However, inasmuch as the concentration of HN02 does vary by an order of
magnitude within the pH range in question, it is reasonable to assume that its
toxic effect might be measurable.
To test this hypothesis, we conducted a series of nitrite bioassays in
which we varied the pH. Details of this study will be reported elsewhere, but
the major conclusions reached were that nitrite toxicity to rainbow trout is
independent of pH within the range tested (pH 7.5 - 8.5), and that nitrite
toxicity is correlated with N02~ concentration (essentially the same as total
nitrite concentration), and is not correlated with HN02 concentration. Thus,
is the principal toxic species of nitrite.
During our pH variation experiments we observed that when hydrochloric
acid (HC1) was used to lower the solution pH, the toxicity of nitrite was
greatly decreased. This indicated that chloride ion (Cl~) might be exerting
126
-------
an antagonistic effect on nitrite toxicity. We therefore conducted a series
of nitrite bioassays where we added C1~ (as NaCl) so that the Cl" concentra-
tion in the test tanks was 1, 5, 10, 20 and 4] mg/1. The rainbow trout used
in these bioassays were all from the same lot. The bioassay using 41 mg/1
Cl" was run in duplicate, and the duration of acclimation of the fish to the
NaCl solutions, before addition of NaN02> was 5 days in one case and 10 in
the other. Results of the two tests were in extremely close agreement, indi-
cating that the difference in acclimation times between 5 and 10 days had
little or no measurable effect. The results of these experiments are sum-
marized in Table 3; the toxicity curves are presented in Figure 2 and include
a zero-chloride bioassay for purposes of comparison.
It can be seen from the table and figure that Cl" exerts a marked effect
on nitrite toxicity; an increase in Cl" concentration causes a decrease in
nitrite toxicity. This effect is linearly correlated. Using a weighted
regression analysis (where the observation is weighted as I/variance of the
LC50), we obtain the following correlation coefficients: for 48 hr, .9964
(p=.00000); for 72 hr, .9957 (p=.00000); for 96 hr, .9873 (p=.00003).
Comparison of LC50 values for two bioassays in which chloride ion con-
centration of 10 mg/1 was achieved, in one case by addition of NaCl (Table
3, Test 363), and in the other case by addition of HC1 (data not given here),
indicates that Cl~ is exhibiting this inhibitory effect; there is no
evidence that it is attributable to Na+. Thus, we have established that Cl"
ion exhibits an antagonistic effect on nitrite toxicity. From preliminary
results of other research in our laboratory, we have found that bromide ion
exhibits a similar inhibitory effect. Other ions may also exhibit this kind
of antagonism to nitrite toxicity.
CONCLUSIONS
The results of our nitrite bioassays on fishes (including pH variation
data not yet published), and the reported results of others, lead us to
conclude that: (a) Exposure to nitrite causes an increase in methemoglobin
concentration in fish blood, although this may not be the only toxic action
of nitrite on fishes, (b) There are differences in susceptibility to nitrite
among fish species, with rainbow and cutthroat trouts being much more suscep-
tible to nitrite than fathead minnows or sculpins. (c) Rainbow trout fry are
less susceptible to nitrite than are larger rainbow trout; there is no
readily apparent size-related difference in toxicity among rainbow trout
between 12 and 235 g. (d) the toxicity of nitrite is related to nitrite ion
concentration, not nitrous acid concentration; changes in pH in the range
7.5 - 8.5 do not affect nitrite toxicity. (e) An increase in chloride con-
centration causes a decrease in nitrite toxicity; this relationship is linear
at least up to 40 mg/1 CT.
ACKNOWLEDGMENT
This work was funded by the U. S. Environmental Protection Agency,
Duluth, Minnesota, Research Grants No. R800861 and R803950. Robert J.
Luedtke and Charles Chakoumakos provided valuable assistance with the
127
-------
28
£4
I
CM
o
O 12
in
O
(377),
(S66)
(326)
(357)-
T1ME, DAYS
cr, irfl/i
Figure 2. Effect of chloride on nitrite toxicity to rainbow trout
(Salmo gaipdneri).
128
-------
biological procedures and chemical analyses. We thank Kenneth Emerson and
Martin A. Hamilton for assistance with some of the chemical and statistical
calculations.
129
-------
LITERATURE CITED
Anthonisen, A. C., R. C. Loehr, T. B. S. Prakasam, and E. G. Srinath. 1976.
Inhibition of nitrification by ammonia and nitrous acid. J. Water
Poll. Cont. Fed. 48(5): 835-852.
American Public Health Association et al. 1976. Standard methods for the
examination of water and wastewater. 14th ed. Am. Public Health Assoc.,
Washington, D.C. xxxix + 1193 p.
Bodansky, 0. 1951. Methemoglobinemia and methemoglobin-producing compounds.
Pharmacol. Rev. 3(1): 144-196.
Brown, D. A., and D. J. McLeay. 1975. Effect of nitrite on methemoglobin
and total hemoglobin of juvenile rainbow trout. Prog. Fish-Cult.
37(1): 36-38.
Cameron, J. N. 1971. Methemoglobin in erythrocytes of rainbow trout. Comp.
Biochem. Physio!. 40(3A): 743-749.
Colt, J. E. 1974. Evaluation of the short-term toxicity of nitrogenous
compounds to channel catfish. Unpublished Ph.D. thesis. Univ.
California, Davis. 94 p.
Colt, J. [E.], and G. Tchobanoglous. 1976. Evaluation of the short-term
toxicity of nitrogenous compounds to channel catfish, Ictali&us
punatatus. Aquaculture 8(3): 209-224.
Gillette, L. A., D. L. Miller, and H. E. Redman. 1952. Appraisal of a
chemical waste problem by fish toxicity tests. Sewage Ind. Wastes 24
(11): 1397-1401.
Hamilton, M. A., R. C. Russo, and R. V. Thurston. 1977. The trimmed
Spearman-Karber method for estimating median lethal concentrations in
toxicity bioassays. Environ. Sci. Techno!. 11. (In press.)
Jaffe, E. R. 1964. Metabolic processes involved in the formation and
reduction of methemoglobin in human erythrocytes. Pp. 397-422 in C.
Bishop and D. M. Surgenor (eds.), The red blood cell. Academic Press,
New York. 566 p.
Kiese, M. 1974. Methemoglobinemia: a comprehensive treatise. CRC Press,
Cleveland. 259 p.
Klingler, K. 1957. Natriumnitrit, ein langsamwirkendes Fischgift. Schweiz.
Z. Hydrol. 19(2): 565-578. [In English translation.]
Konikoff, M. A. 1973. Comparison of clinoptilolite and biofilters for
nitrogen removal in recirculating fish culture systems. Ph.D. thesis,
Southern Illinois Univ., Carbondale. 98 p. (Diss. Abst. 1974. 34:
4755B.)
130.
-------
Konikoff, M. 1975. Toxicity of nitrite to channel catfish. Prog. Fish-
Cult. 37(2): 96-98.
McCoy, E. F. 1972. Role of bacteria in the nitrogen cycle in lakes. Water
Poll. Cont. Res. Ser. 16010 EHR 03/72. Office of Research and Monitor-
ing, U. S. Environmental Protection Agency, Washington, D. C. vii +
23 p.
Mount, D. I., and W. A. Brungs. 1967. A simplified dosing apparatus for
fish toxicology studies. Water Res. 1(1): 21-29.
Russo, R. C., C. E. Smith, and R. V. Thurston. 1974. Acute toxicity of
nitrite to rainbow trout (Salmo gairdneri], J. Fish. Res. Bd. Canada
31(10): 1653-1655.
Russo, R. C., and R. V. Thurston. 1974. Water analysis of the East Gallatin
River (Gallatin County) Montana 1973. Tech. Rept. 74-2. Fisheries
Bioassay Laboratory, Montana State University, Bozeman. 27 p.
Shterman, L. Ya. 1970. Methemoglobin in fish blood. J. Ichthyol. 10(5):
709-712.
Smith, C. E., and R. C. Russo. 1975. Nitrite-induced methemoglobinemia in
rainbow trout. Prog. Fish-Cult. 37(3): 150-152.
Smith, C. E., and W. G. Williams. 1974. Experimental nitrite toxicity in
rainbow trout and chinook salmon. Trans. Am. Fish. Soc. 103(2): 389-
390.
U. S. Environmental Protection Agency. 1974. Methods for chemical analysis
of water and wastes. EPA-625-/6-74-003. Methods Development and
Quality Assurance Research Laboratory, National Environmental Research
Center, Cincinnati, Ohio. pp. 215-216.
Wallen, I. E., W. C. Greer, and R. Lasater. 1957. Toxicity to Garnbusia
affinis of certain pure chemicals in turbid waters. Sewage Ind.
Wastes 29(6): 695-711.
Westin, D. T. 1974. Nitrate and nitrite toxicity to salmonoid fishes.
Prog. Fish-Cult. 36(2): 86-89.
131
-------
COPPER TOXICITY: A QUESTION OF FORM
G. A. Chapman and J. K. McCrady
Western Fish Toxicology Station
U. S. Environmental Protection Agency
1350 S.E. Goodnight
Corvallis, Oregon 97330
An abundance of literature indicates that copper toxicity is one of the
more intensively investigated areas of fish toxicology. Much of the data on
copper toxicity comes from acute toxicity studies on a wide variety of fish
species, in waters of differing quality, using diverse methods. Compilations
of these data can provide valuable information, but are little help in under-
standing and predicting toxic levels of copper. A second large area of
research into copper toxicity involves studies whose goal is to increase the
understanding of the role of receiving water quality on copper toxicity and
from this understanding to generate a better predictive capability for esti-
mating potentially adverse levels of copper in various natural waters.
Although there are many factors which complicate this predictive capability
(e.g. variable copper exposure levels, biological acclimatization, complex
wastes, sublethal effects), it has generally been sought through relatively
simple experiments relying on continuous short-term exposure at constant
copper concentrations and utilizing death as the indicator.
Our interest in copper form was stimulated because of the toxicity of
relatively low levels of copper in a series of flow through toxicity tests
conducted at the Western Fish Toxicology Station (WFTS). The 96-hr LC50
values obtained in these tests with juvenile salmonids ranged from 15 to 38
yg/ji; these values were significantly lower than most related data in the
literature. In my presentation today I wish briefly to trace the contempo-
rary history of research into the effects of receiving water quality on
copper toxicity and to present some recent data dealing with this subject.
For additional information on the toxicity of copper to fish I recommend the
critical review of the literature by Doudoroff and Katz (1953), the toxicity
compendium of McKee and Wolf (1963), and the discussion and recommendations
in Water Quality Criteria 1972 (Nat. Acad. Sci. and Nat. Acad. Engr. 1973).
In the latter report the primary effect noted of receiving water quality
on copper toxicity was the effect of hardness. This effect is generally
recognized, the best known reference being that of Lloyd and Herbert (1962).
Their data (Figure 1) indicated that higher copper concentrations were
required to produce lethality as the total hardness increased. When hardness
132
-------
1000
500
300
g 200
o
00
•53-
ct:
Q_
O
O
100
50
30
20
10
Rainbow Trout (Lloyd and Herbert 1962}
j i
20 30 50 100 200 300 500
TOTAL HARDNESS ( mg/liter CaCo3 )
Figure 1. The relationship between total hardness and the 48-hr LC50 of
copper to rainbow trout.
133
-------
increased. When hardness increased over a range from 15 to 320 mgA as CaCC>3
the 48-hr LC50 for rainbow trout (Salmo gairdneri) increased from about 45
wgA to about 450 yg/£.
Nine years after Lloyd and Herbert's 1962 paper, a related paper came
out of the same Water Pollution Research Laboratory at Stevenage, England,
and in this paper, Stiff (1971) developed an explanation of the results
reported in the 1962 paper. Realizing that levels of hardness and alkalinity
usually are related and approximately directly proportional in natural waters,
Stiff proposed that the phenomenon noted by Lloyd and Herbert could have been
largely due to the greater formation of copper carbonate complexes at the
higher alkalinities which accompanied the higher hardness values. Utilizing
published equilibrium values for chemical reactions involving Cu++HC03, and
H+ he computed the theoretical amount of free copper (cupric ion) in waters
having various alkalinities and pH values (Figure 2). Stiff's results indi-
cated that as alkalinity increased at a given pH, the amount of free copper
decreased sharply. Further, as pH increased, the amount of free copper also
decreased greatly at a given alkalinity.
Utilizing a computer program (REDEQL 2) developed at Cal Tech by Morgan,
Morel, and McDuff and modified by Ingel (1976), we computed free copper con-
centrations for a variety of alkalinities and pH values using equilibria data
for reactions among Cu++ (10~6 MA). Ca++, MG++, NA+, K+, C0|, SO/"1", Cl",
H+, and OH~ when present in ratios recommended for reconstituted freshwaters
for toxicity tests (Table 1). The results we obtained were qualitatively
identical to Stiff's and very close quantitatively (Figure 2). It should be
noted that Stiff ignored what he termed "the slowly formed complex" of copper
in his computations and we allowed no precipitation in our model; both con-
straints were based on observations in the laboratory using copper specific
ion electrodes. Apparently attaining final equilibrium concentrations of
some copper complexes may require longer than the residence time of aquaria,
mixing zones, and some rivers.
The matrix defined by the interactions among free copper, alkalinity,
and pH can be simplified to a relationship similar to that described by the
line for reconstituted freshwater shown in Figure 3. Since pH and alkalinity
are rather closely related in most natural waters, i.e. high alkalinity and
high pH occur together, it is possible to generalize a relationship between
alkalinity and free copper (or conversely between pH and free copper). The
alkalinity-pH relationships observed in 110 samples from 52 stations on 37
western Oregon streams are included in Figure 3 showing the pH-alkalinity
regression line as well as lines enclosing the extreme values (Samuelson
1976}. I draw these comparisons primarily to point out that studies which I
will discuss in which we used these four reconstituted freshwaters are
generally applicable to natural waters and do not refer solely to four
arbitrary points in the pH-alkalinity matrix.
The reconstituted freshwaters to which I refer were recommended in
"Methods for Acute Toxicity Tests with Fish, Macro-invertebrates, and
Amphibians" (U. S. Environmental Protection Agency 1975). We decided to use
these waters because we wanted maximum uniformity in water quality and we
134
-------
to
en
TABLE 1. QUANTITIES OF REAGENT-GRADE CHEMICALS USED TO PREPARE RECOMMENDED RECONSTITUTED FRESH WATERS
AND THE RESULTING WATER QUALITIES.3
Salts Required (mg/£)
Name
Very soft
Soft
Hard
Very Hard
NaHC03
12
48
192
384
CaS04'2H20
7.
30.
120.
240.
5
0
0
0
MgS04
7.5
30.0
120.0
240.0
KC1
0.5
2.0
8.0
16.0
Nominal Range and Observed Mean Value"
PH
6.4-6.
7.2-7.
7.6-8.
8.0-8.
8 (7.2)
6 (7.6)
0 (8.1)
4 (8.5)
Hardness
10-13
40-48
160-180
280-320
(13)
(46)
(182)
(359)
Alkali
10-13
30-35
110-120
225-245
nity
(12)
(35)
(125)
(243)
The Committee on Methods for Toxicity Tests with Aquatic Organisms, 1975.
5Mean value (in parenthesis) from bioassays.
-------
o:
LU
O
o
LU
LU
CC
LU
O
(r
LU
Q_
O
100
50
30
20
10
3
2
1.0
LU .5
o:
a 3
i -°
h .2
.1
10
x-Sfjff,M.J. 1971
• -WFTS Data
20 30 50 100 200 300 500
Figure 2.
TOTAL ALKALINITY (mg/liter CaC03)
Calculated percent free copper (cupric ion) at indicated
alkalinities and pH values.
136
-------
100
50
30
tr 20
Ld
CL
CL
8 I0
LU
LU
h-
z:
UJ
o
o:
UJ
o_
u
(-
UJ
a:
o
UJ
3
2
1.0
.2
.1
Western Oregon River Data
pH7.5
pHS.O
Reconstituted
Freshwater
pH8.5
Figure 3.
10 20 30 50 100 200 300 500
TOTAL ALKALINITY (mg/liter CaC03)
Relationship between pH, alkalinity, and theoretical percent
free copper for natural waters and reconstituted freshwaters.
137
-------
wanted waters whose copper complexors were essentially known both qualita-
tively and quantitatively. This decision simplified the use of the chemical
equilibrium computer model and simplified interpretation of the copper
specific ion data. In addition, we were interested in comparing copper
toxicity data from waters of known quality, particularly carbonate copper
complexing systems, with the hardness-copper mortality relationship published
by Lloyd and Herbert (1962). In so doing we could determine whether effects
of non-carbonate copper complexors (e.g. phosphates, organics) contributed
appreciably to the widely used hardness-copper toxicity relationships of
Lloyd and Herbert.
Although reconstituted freshwaters are primarily used in static toxicity
tests, we were able to utilize one of our existing flow-through diluter
diluter systems (Figure 4) with the reconstituted water. Well water was
passed through a reverse osmosis unit producing water with a conductivity of
about 1 umho/cm to which reagent grade chemicals were added in appropriate
quantities to make up the reconstituted freshwaters in Table 1. Pumps
continually agitated the water, providing aeration and mixing, and temperature
was maintained at 12 C. Toxicity tests were conducted in 19 liter aquaria
92x26x41 cm deep and containing 14 liters of water. Aquaria were dosed with
a diluter modified from that described by Mount and Brungs (1967). Twelve
aquaria were used (6 concentrations X 2 replicates per concentration). Time
for 50% aquarium volume replacement was 1 hour based on the flow-volume
relationship shown by Sprague (1969). Photoperiod was set to coincide with
sunrise-sunset tables for Corvallis, Oregon (dim illumination was used in
lieu of complete darkness).
Tests were conducted for 96-hr and test fish were acclimated to the
reconstituted freshwater for one week prior to the toxicity tests. Tests
were conducted with 3-month-old chinook salmon (OneovkynQhus tshowytssha]
having a mean weight of 1.35 g. Fish were fed Oregon Moist Pellet up to 48
hours prior to the start of the test, and were not fed thereafter.
Water analyses were conducted daily for dissolved oxygen, pH, total
hardness, total alkalinity, and total copper at each copper concentration
(one aquarium per duplicate pair). Copper analysis was by flameless atomic
absorption. Daily cupric ion activity measurements were made in situ using
an Orion cupric ion electrode*. In order to eliminate interference due to
light, the aquarium was provided with a black plastic cover during cupric ion
measurements. The electrode was calibrated using copper standards in acetate
buffer.
The results of our tests with reconstituted waters of various hardnesses
and alkalinities conformed to the familiar relationship of higher LC50 values
at higher hardnesses and alkalinities. The 96-hr LC50 values ranged from
about 10 pg/£ in very soft water to about 125 pg/£ in very hard water. We
found the resulting copper toxicity relationship to be nearly parallel to that
of Lloyd and Herbert (1962), utilizing either hardness or alkalinity as the
determinant (Figure 5). This result indicated that the hardness-copper
*Mention of product does not constitute endorsement by the Environmental
Protection Agency.
138
-------
10
WELL WATER
SAND
FILTER
FILTER
REJECT
CHEMICAL
REVERSE
OSMOSIS
UNIT
PRODUCT
STORAGE
AND
MIXING
HEAD BOX
OVER-
FLOW
CHILLER
DILUTER
STORAGE
HEATING
AERATION
SCHEMATIC OF RECONSTITUTED FRESH WATER FLOW-
THROUGH BIOASSAY SYSTEM
Figure 4. Flow diagram of the system used to supply reconstituted freshwater for copper toxicity
tests.
-------
A-48hr. LC50, Rai nbow Trout {Lloyd & Herbert, 1962) (T. Hard.)
B - 48 hr. LC50, Chinook Salmon r=. 998 (T. Hard.)
C - 96 hr. LC50, Chinook Salmon r=. 998 (T. Hard.)
D - 96 hr, LC50, Chinook Salmon r«. 998 (T. Alk.)
1000
500
300
i 200
"Si
3
o 100
ITS
O
LU -n
a. 50
Q_
O
30
20
10
20 30 50
100
200 300 500
( mg/!iter CaCo3) TOTAL HARDNESS OR ALKALINITY
Figure 5. Comparison of relationships between hardness or alkalinity and
acutely lethal levels of copper to rainbow trout and chinook
salmon.
140
-------
toxicity relationship shown in both studies could be explained by the alka-
linity dependent carbonate complexation of copper as suggested by Stiff
(1971).
However the 48-hr LC50 value for a given alkalinity from our study was
lower than that of Lloyd and Herbert by a factor of 3 to 4. The most reason-
able explanations for this divergence between the two studies lies in three
areas of difference: fish species and size, test methods, and water quality.
Studies at our lab have shown that steelhead trout (the anadromous form of
the rainbow trout studied by Lloyd and Herbert) are slightly more sensitive
to copper than are Chinook salmon, so fish species differences may not explain
the data; however, the Chinook slamon used in our studies may have been
smaller than the trout used by Lloyd and Herbert. The bioassays of these
authors were static, changed every 24 hours (R. Lloyd, personal communica-
tion), while ours were flow-through. In our experience, lower copper LC50
values are obtained in flow-through bioassays than in static bioassays. The
presence of strong copper complexing capacity of the type described by Chau,
Gachter, and Lum-Shue-Chan (1974) in the water used by Lloyd and Herbert can-
not be discounted; this phenomenon could produce the effect of raising the
copper LC50 values in just the manner noted. However, the difference in bio-
assay methods was probably the biggest contributor to the observed differ-
ences in LC50 values between the two studies. Regardless of the difference
between the two studies, it appeared that Stiff's (1971) explanation of the
effects of carbonate in modifying copper toxicity were tenable based on our
studies with reconstituted water.
One aspect of the alkalinity effect proposed by Stiff was that free
cupric ion was a primary toxic form. This conclusion was supported by
equilibrium calculations from published copper toxicity data compiled by
Pagenkopf, Russo, and Thurston (1974) although they determined that CuOH
might also be involved. Additional evidence was developed by Andrew (1976)
who showed that Daphnia survival time was proportional to cupric ion activity
(Figure 6). Based on these results it appeared that copper toxicity was
directly related to cupric ion activity and could be due primarily to that
form.
Andrew (1976) also showed that the 96-hr LC50 of cupric ion activity
was nearly equal for fathead minnows (Pimaphales pyomelas] in tests conducted
in two appreciably different waters (Figure 7). Thus, while 96-hr LC50
values for total copper were about 200 and 800 yg/£, the cupric ion activity
was only 0.70 and 0.55 M/& respectively. An exciting aspect of these results
was that a given free copper concentration might be determined to produce a
given effect regardless of water quality and total copper concentration.
However, when we looked at our bioassay data with respect to cupric ion
activity we found that this relationship did not occur. Indeed, we looked at
the 96-hr LC50 values in five different ways in regard to copper form and
found no simplifying result in any case (Figure 8).
Interestingly, while total copper 96-hr LC50 concentrations increased
with increasing alkalinity, cupric ion 96-hr LC50 values decreased with
141
-------
12 r
10
h-
CO
o
LU
o:
to
Q_
R=0.95
(N=19)
0.02 0.04 0.06
CU** ACTIVITY(uM/liter)
Figure 6. Rclationhip of reciprocal survival time of Rtphnia magna to
cupric ion activity (Andrew, 1976).
142
-------
0.30
0.20
5
*
0
0.10
Fathead Minnow
96 hr. LC
and Confidence Limits
200 400 600 800 1000
TOTAL Cu CONCENTRATION (ug/liter)
Figure 7. Relationship of cupric-ion activity, total copper concentration,
and 96-hr LCSO's for fathead minnows (Andrew, 1976).
143
-------
200
100
50
o>
~ 10
or
UJ
a.
Q- 5
o
o
o
IT)
O
cp 1.0
CD
0.5
O.I
CCutotal]
Meas.
[Cu++]Meas.+
r +t
CcuOH
r
Comp.
J I
lj
10
50 100
500 1000
Figure 8.
TOTAL ALKALINITY ( mg / liter CaC03)
Relationship between alkalinity and copper 96-hr LC50 values for
chinook salmon, expressed as measured total copper, measured
cupric ion, theoretical cupric ion, and theoretical cupric
hdroxide ion concentrations.
144
-------
increasing alkalinity. Considerably more cupric ion activity was measured
in situ (in the aquaria) than was predicted by the computer chemical equilib-
rium model. However, both the computer model and the cupric ion electrode
yielded similar results when the electrode analyses were made on a sample
gently stirred in a beaker. We presumed that the difference between aquaria
and beaker determinations reflected higher cupric ion levels in the aquaria
due to the relatively short reaction time of the cupric ions with the corn-
plexors in the dilution water.*
The data obtained in our chinook bioassays and those reported by Andrew
(1976) for fathead minnows appear to differ, in that Andrew's data could sug-
gest a constancy of cupric ion LC50 values while ours do not support such a
constancy. However, if pH is treated as a variable the data from the two
studies become more similar (Figure 9). Andres (personal communication) has
also found that apparent cupric ion toxicity increases with increasing pH.
This tentative analysis suggests that pH may be an important factor in
copper toxicity in addition to its usual association with alkalinity and the
effect of pH on copper complexation equilibria. In a search for a simplify-
ing premise we now wonder if the acutely lethal level of copper for a given
species of fish would be some constant cupric ion activity level for a given
pH value.
If this pH-constant relationship held true, then it should be possible to
determine how much total copper would be required to produce an acutely lethal
level of cupric ion activity in a specific water. Individual samples of water
could be titrated with copper and a titration curve established from which
cupric ion activity could be determined for any total copper concentration.
Sample titration curves for two reconstituted waters are shown in Figure 10.
The water without EDTA yields a straight line, with complexation due essen-
tially to reactions with carbonate and hydroxide. The nearly parallel line of
the water with EDTA added indicates a similar complexing capacity but only
after the EDTA has strongly complexed an equimolar (10~° M/J.) amount of
copper. The X intercept in this instance represents a strong copper complex-
ing capacity of the type described by Chau, Gachter, and Lum-Shue-Chan (1974).
As would be expected there is essentially no strong complexing capacity in the
reconstituted soft water.
We have run few copper titration curves for natural waters, but we are
currently conducting a study to determine the copper complexing capacity of
a variety of regional water. Initial results for two Oregon rivers are shown
in Figure 11. We found that these titration curves did not yield straight
lines, a result which we presume is due to the presence of multiple complex-
ors at differing concentrations. Both river waters had a strong copper
complexing capacity of <20 yg of copper/liter, and beyond the strong complexa-
tion, one river water sample (Alsea) had about twice the copper complexing
capacity as the other (Willamette).
*Recent experiments indicate no significant changes in cupric ion activity
when aquarium 50 percent volume replacement time was increased from 30 min
to 5 hrs.
145
-------
-------
0.20
0.\5
0.10
>
I-
u
o 0.05
0
Reconstituted Soft Water
Without EDTA
Plus 10 v M EDTA
20 40 60 80 100 120 140
TOTAL COPPER CONCENTRATION (yu.g/I iter)
160
Figure 10.
Relationships between total copper concentration and cupric ion activity obtained with
reconstituted freshwater with and without EDTA.
-------
CO
0.20r
.-? °'15
\
s
5
t 0.10
H
o
3
o
0.05
0
Willamette River pH 7.5
Alsea River pH 7.7
I
j
20 40 60 80 100 120
TOTAL COPPER CONCENTRATION
140 160
Figure 11. Relationships between total copper concentration and cupric ion activity obtained with two
natural waters.
-------
Returning to the copper activity 96-hr LC50 vs. pH relationship (Figure
9) we find for pH 7.5 and 7.7 corresponding LC50 values of 0.06 and 0.05 yM
copper activity/A. Utilizing these copper activity values and the titration
curves shown, in Figure 11, yields estimated 96-hr LC50 values for total copper
of about 70 yg/A for the Willamette River and 100 pg/Ji for the Alsea River.
If this procedure should prove tenable one could estimate lethal levels of
copper for various waters by chemical means rather than by biological means.
In some instances use of such chemical procedures could be highly advanta-
geous. Regardless of its direct applicability, the knowledge about the
relationships between receiving water quality and pollutant toxicity can aid
in understanding the variability observed in studies related to fish toxicol-
ogy.
I would like to conclude by placing the matter of copper form in a more
general perspective. First, even if acutely lethal copper levels can be
determined on the basis of cupric ion activity, a similar relationship with
chronic toxicity is not assured. Therefore field studies, chronic toxicity
studies, application factors, or short-cut indicator tests (e.g. the ventila-
tion cough response) would still be required to estimate safe levels of
copper. (It would be instructive to measure cough response and cupric ion
activity in several freshwater matrices to see what relationships occur.)
Second, based on the differences on copper activity observed in situ in the
aquaria and in samples equilibrated in beakers such factors as pH, the copper
form in the waste, and the reaction time in rivers or test aquaria are
important determinants of cupric ion activity and presumably of copper toxic-
ity. Finally, the data relating copper toxicity to cupric ion activity are
far from being definitive. Nevertheless, this area of research promises to
add appreciably to the understanding and predictive capabilities with regard
to copper toxicity.
149
-------
REFERENCES
Andrew, R. W. 1976. Toxicity relationships to copper forms in natural
waters. Pp. 127-143 In^ R. W. Andrew, P. V. Hodson, and D. E.
Konasewich (eds.), Toxicity to biota of metal forms in natural waters.
(Proceedings of a workshop held in Duluth, Minn. Oct. 7-8, 1975.)
Committee on the Scientific Basis for Water Quality Criteria, Great
Lakes Research Advisory Board, International Joint Commission. 329 p.
Chau, Y. K., R. Gachter, and K. Lum-Shue-Chan. 1974. Determination of the
apparent complexing capacity of lake waters. 0. Fish. Res, Bd. Canada
31(9): 1515-1519.
Doudoroff, P., and M. Katz. 1953. Critical review of literature on the
toxicity of industrial wastes and their components to fish. II. The
metals, as salts. Sewage Ind. Wastes 25(7): 802-839.
Ingle, S. E. 1976. Users' guide to REDEQL.EPA: A chemical equilibrium pro-
gram. Con/all is Environ. Res. Lab., Coastal Poll, Branch, U. S.
Environmental Protection Agency, Corvallis, Ore. 29 p. Mimeo.
Lloyd, R., and D, W. M. Herbert. 1962. The effect of the environment on the
toxicity of poisons to fish. J. Inst. Public Health Engin., pp. 132-
143.
Marking, L. L., and V. K. Dawson. 1973. Toxicity of quinaldine sulfate to
fish. Invest. Fish Control 48. U. S. Fish Wild!. Serv., Washington,
D. C. 8 p.
McKee, J, E., and H. W. Wolf (eds.). 1963. Water quality criteria. 2nd ed.
California State Water Quality Control Board Publ. 3-A. Sacramento,
Cal. xiv + 548 p. + map.
Mount, D. I., and W. A. Brungs. 1967. A simplified dosing apparatus for
fish toxicology studies. Water Res. 1(1): 21-29.
National Academy of Sciences and National Academy of Engineering. 1973.
Water quality criteria 1972. A report of the Committee on Water Quality
Criteria, Environmental Studies Board. Ecol. Res. Ser. EPA-R3-73-033.
U. S. Environmental Protection Agency, Washington, D. C. xix + 594 p.
Pagenkopf, G. K., R. C. Russo, and R. V. Thurston. 1974. Effect of complex-
ation on toxicity of copper to fishes. J. Fish. Res. Bd. Canada 31(4):
462-465.
Samuelson, D. F. 1976. Water quality: Western Fish Toxicology Station and
western Oregon rivers. Ecol. Res. Ser. EPA-600/3-76-077. Environ. Res.
Lab., Office of Res. & Devel., U. S. Environmental Protection Agency,
Duluth, Minn, viii + 56 p.
150
-------
Sprague, J, B. 1969. Measurement of pollutant toxicity to fish I. Bioassay
methods for acute toxicity. Water Res. 3(11): 793-821.
Stiff, M. J. 1971. Copper/bicarbonate equilibria in solutions of bicarbon-
ate ion at concentrations similar to those found in natural water.
Water Res. 5(5): 171-176.
U. S. Environmental Protection Agency, Committee on Methods for Toxicity Tests
with Aquatic Organisms. 1975. Methods for acute toxicity tests with
fish, macroinvertebrates, and amphibians. Ecol. Res. Ser. EPA-660/3-
75-009. Natl. Environ. Res. Center, Office of Res. & Devel., U. S.
Environmental Protection Agency, Corvallis, Ore. 61 p.
151
-------
THE ROLE OF CYANIDE AS AN
ECOLOGICAL STRESSING FACTOR TO FISH
Gerard Leduc
Department of Biological Sciences
Concordia University
1455 de Maisonneuve
Montreal H3G IMS, Quebec
Canada
ABSTRACT
_, Cyanide, at concentrations as low as 0.01 mg
l" HCN, produces individual stresses on fish which,
when integrated into a single total response, so ser-
iously affect the energy supply processes, that both
the range and scope for activity are reduced. The pro-
posed toxicological model suggests that, at least under
laboratory conditions, the fish could not continue to
exist as populations.
This conclusion was drawn after evaluating the
discrete effects of chronic cyanide poisoning on var-
ious fish tested under laboratory conditions in flow-
through aquaria. The physiological responses tested
were: embryological development; growth (wet, dry and
fat weights) where a fixed or variable food ration,
swimming velocity and initial size of the fish were
tested; respiration and swimming after the poisoning
period; histopathology of the liver; iono-and osmo-
regulation in varying salinities. Cyanide was also
tested jointly with arsenic showing an additive
deleterious effect on growth.
Not all physiological parameters were equally
affected by cyanide but it appears that the greater
energy-demanding processes, such as fat biosynthesis,
osmoregulation and swimming were more seriously af-
fected by this respiratory poison.
These results were integrated into a single eco-
physiological response curve - a Relative Performance
Index - which was used to develop a cyanide-stressed
152
-------
Scope for Activity model. This model suggests a 50%
reductionin the overall performance of the fish at
0.01 mg l" HCN and supports the previously established
water quality criteria for cyanide, i.e. a maximum
permissible level of 0.005 mg 1" HCN.
INTRODUCTION
The continuous introduction of new chemicals into the aquatic environ-
ment poses a double challenge to the water pollution biologist:
1. He must measure indivudual responses of organisms to these new
external stimuli at the physiological and/or biochemical levels keeping in
mind the relationships to food, climate, niche and the animal community.
Not only should these entities be recognized before testing, but their ap-
plication should be within realistic limits such as food quality and quantity;
temperature, flow and current of water; photoperiod; and, most important,
the form under which toxicants are administered to the test animals.
2. He must evaluate the impact of the responses at the ecological
1evel.
It is only in these terms that aquatic toxicology will produce new
knowledge applicable to the definition of sound water quality criteria. It
is essential to standardize the experimental conditions, thus requiring a lab-
oratory approach to minimize the complex interaction of the multiple and un-
controllable environmental factors that prevail in nature. It is therefore
the aim of the aquatic toxicologist to reach an ecological understanding of
toxicants so that test organisms are exposed to realistic amounts and chemical
species, for meaningful periods and through media (water, food, sediments,
etc.) under which they occur in nature.
To relate aquatic toxicology to natural conditions, despite the un-
realistic environment dictated by laboratory experimentation, one must first
evaluate the relative importance of single physiological responses to the
total performance of the animal in nature. This knowledge will come through
simple concepts of animal activity such as growth, movement, reproduction,
and behavior. Huntsman, (1948) defined the total response of animals to
their environment as Biapocrisis. This conceptual approach had been elab-
orated by Fry (1947) who quantified ecophysiology with the concept of Scope
for Activity, a measure under particular environmental conditions of the
animal's metabolic energy available for activity above and beyond the minumum
needs for maintenance (Figure 1). In a way, Scope for Activity is the total
metabolic capacity an animal has available to meet the ecological realities
of life in nature (Warren 1971, p. 148). Iverson and Guthrie (1969) have ex-
tended Fry's concept to natural populations integrating the total response of
animals to environmental factors taken one at a time or interacting together.
The "goodness of the habitat" which varies between upper and lower limits of
environmental identities reflects the distribution and abundance of animals
in nature from the center of distribution to the limit of their range (Figure
2). Iverson and Guthrie's most interesting contribution is the application
of the notion of environmental stress to populations responding to natural
and/or pollutional factors. The notion of stress must be taken positively
as a response - sometimes useful, sometimes harmful - to the population. If
153
-------
CO
J
o
u
s
N.
N
ZONE OF TOLERANCE'
'POTENTIAL RAN9E OF ACTIVITY-
OF UNSTRESSED ANIMAL
-H
FROM FRY (1947)
Figure 1. Diagram illustrating the standard and active metabolic rates of
an organism subjected to an environmental factor, under normal
conditions and under the influence of a stressing factor which
reduces both the scope and range of the Scope for Activity.
(Modified from Fry 1947.)
154
-------
<
K
m
x
u_
o
if>
UJ
z
o
o
LOWER LIMIT UPPER LIMIT
ENVIRONMENTAL IDENTITY X,
v ENVIRONMENTAL
V IDENTITY XL
MODIFIED FROM
IVERSON AND GUTHRIE 1969
Figure 2. Diagrams illustrating the response of a population: upper graph,
to a single environmental entity under normal conditions (dotted
line) and under the influence of a stressing factor (shaded area)
In lower graph, same as above but for a population responding to
two environmental entities without and under stress. (Modified
from Iverson and Guthrie 1969.)
155
-------
a population Is reduced in size and distribution when responding to a natural
factor(s), it is said to be under stress. This response might have a high
selective value and be good. However, there are reduction levels from which a
population could not recover and if, generation after generation, the pop-
ulation keeps shrinking, that stress is undesirable. "An environmental entity
which is not lethal in the toxicological sense of that term, but affects the
range and/or scope of activity of an organism or population is, ecologically
speaking, a stress" (Iverson and Guthrie 1969).
We surmise that toxicants, in nature, do not always produce readily
visible toxic effects on fish because of dilution and other masking factors.
Ecological potential may be reduced through reaction at the individual physi-
ological levels or stresses on other important related organisms, or both. By
measuring in the laboratory the effects of a toxicant on various reactions of
ecological importance it may he possible to model overall effects and pos-
tulate a safe application factor.
This paper is an overview of the effect of cyanide from the work of
several co-workers who for many years have contributed to this subject.
Cyanide, as simple molecular HCN or in the form of metal complexes, has
been extensively studied in water pollution research and the literature of
that subject has been extensively reviewed by Doudoroff (1976). Attention
to cyanide as a research subject is valid because of double toxicological
interest, basic and applied. Cyanide has long been known as a violent poi-
son. Its properties as a selective inhibitor have been recognized as a use-
ful research tool in respiratory physiology at the cell (Commoner 1940;
Stannard and Horecker 1948; Keilin and King 1960) and organismal levels
(Sumner and Doudoroff 1938) thus providing a good basic knowledge of its mode
of action. As to its practical implications, cyanide is a common pollutant
associated with mining. It is used in large quantities as a flotation re-
agent for silver and gold ores. It is also widely used as a complexing agent
in the electroplating of zinc, copper and silver. In addition, the steel
and chemical industries make wide use of this common chemical.
LABORATORY RESEARCH
Our laboratory studies with cyanide encompassed several aspects of the
life cycle of fish namely, embryological development, growth, swimming, osmo-
regulation, respiration, histopathology of reproductive organs and liver.
The test conditions varied in different experiments, but the most common ex-
perimental characteristics were as follows: all studies were conducted in the
laboratory, with flow-through test tank systems supplied with dechlorinated
water at pH of about 7.5, at temperatures of 10-25 C and for periods of 10-36
days. The test organisms were rainbow trout (Salmo gairdneri), a cichlid
(Cichlasoma bimaculatum), Atlantic salmon eggs^'(SaTmo salarJT and coho
salmon (Oncorhynchus kisutch).
156
-------
EMBRYOLOGICAL DEVELOPMENT
The purpose of this study was to evaluate the impact of chronic cyanide
poisoning on the early life stages of fish (Leduc 1977). Newly fertilized
Atlantic salmon eggs were obtained from a hatchery and within 24 hours dis-
posed into a series of test tanks, in renewed water and cyanide concentrations
of 0.01, 0.02, 0.04, 0.08 and 0.10 mg 1 as HCN. The observations extended
through hatching and up to the full resorption of the yolk sac of the control
fry when the experiments were terminated. The eggs/fry were continuously ex-
posed to cyanide during this whole period. The average temperature during
incubation was 4.4 C; after hatching the temperature ranged from 3.5 to
8.3 C.
Cyanide markedly affected developing Atlantic salmon embryos. Hatching
was delayed by three to six days and hatching success reduced by 20-40% in
the range of concentrations tested (0.01 to 0.10 mg 1" HCN). During incu-
bation, cyanide reduced the conversion efficiency of yolk into fish tissues
so that at hatching, the cyanide-exposed fry were smaller in length and
weight than the controls at all concentrations above 0.01 mg 1 . However,
this impairment was rapidly overcome after hatching when the cyanide-exposed
fry started to grow faster than the controls and, at the end of the experi-
ment, they were all bigger than, or equal to the controls. Accelerated
growth following a previous depression by cyanide has been observed by
Leduc (1966) in juvenile cichlids and coho salmon and in juvenile rainbow
trout (Dixon 1975; Speyer 1975). This phenomenon, which has not yet been
explained, may be of physiological interest but of little ecological signi-
ficance to the fry if other more serious effects of cyanide occurred during
exposure. Indeed, we noted that many cyanide-exposed fry, although alive,
were abnormal with gross deformities of the head, eyes, mouth and the verte-
bral column (Figure 3), anomalies that would be lethal in nature. The inci-
dence of these macroscopic congenital defects ranged from 6% at 0.01 mg 1"
to 19% at 0.10; the controls had less than 1% anomalies. To illustrate the
overall effects of cyanide on the early life stages of Atlantic salmon a
"realized viability" index was calculated by adding the values of percent
hatching, fry survival and of "normal" fry. It appears from Figure 4 that
cyanide reduced the "realized viability" at the lowest concentration, 0.01 mg
1 , by a significant amount and we believe that a much higher incidence of
abnormalities would have been observed had histological techniques been used.
GROWTH IN RESPONSE TO CYANIDE POISONING
Cyanide, as a respiratory poison, would be expected to reduce the
energy potential for growth in a way somewhat similar to the effects of low
dissolved oxygen in the water (Warren, Doudoroff and Shumway 1973) by re-
ducing food intake, conversion efficiency and/or biosynthesis. The study of
the effects of cyanide on growth was introduced by Leduc (1966) working with
a cichlid (Cichlasoma bimaculatum) fed unlimited rations of live tubificid
worms. The cichlids were held in rectangular troughs, supplied with spring
water heated to 25 C and subjected to various cyanide concentrations ranging
from 0.01 to 0.10 mg l" HCN for 36 days. Growth was measured as changes in
wet weight.
1 C-7
-------
Scale (- 1 I en:
Figure 3. Photograph showing typical body anomalies caused by continuous
exposure of Atlantic salmon eggs and fry to sublethal concen-
trations of cyanide.
158
-------
100,-
P90
§ 60
DC
u 7n
0_ 70
60
>-
I-
5 so
00
040
0 30
Ul
3"
O
1 I
j
L
i
i
i
i
0 .01 .02 .03 .04 .05 .06 .07 .08 .09 .10
HCN CONCENTRATION (mg/l)
Figure 4. Realized viability of Atlantic salmon fry at different cyanide
concentrations to which eggs and fry were exposed throughout
development. (See text for details.)
ifi
-------
Cyanide had different effects on growing cichlids, effects varying
with time of exposure and with the concentrations tested. Cyanide promoted a
higher food consumption; food conversion efficiency was initially higher at
low cyanide levels around 0.02 mg 1~ but lower at higher concentrations.
The resulting growth was then initially better at low cyanide levels but less
than the controls at higher cyanide levels (Figure 5). This pattern however
changed with tine. The initial growth advantage at low cyanide was lost
and the cichlids exposed to higher cyanide levels exhibited a marked increase
in growth rate by the end of the experiments. In other words, the cyanide-
exposed cichlids were making up for the initial growth reduction, the effects
increasing with the cyanide concentration. By the end of the 36-day periods
there were hardly any differences between the control and the cyanide-exposed
cichlids except a little depression at 0.10 mg/1 HCN (Figure 5). Similar re-
sults were obtained by Leduc (1966) with coho salmon fed unlimited ration of
earthworms in a flow-through system at 16 C.
Further studies of the effects of cyanide on the growth of fish were
pursued with another salmonid fish, rainbow trout, focusing attention on
other experimental and growth parameters. The temperature was lower
(11-12 C), an artificial diet was given at different limited rations, and the
effect of holding conditions was tested by comparing the growth of rainbow
trout with and without swimming requirements. One study also evaluated the
combined effects of cyanide with arsenic. As to the growth parameters, in
addition to the wet weight, special attention was given to dry and fat weight
changes.
Dixon (1975) exposed young rainbow trout to 0.01, 0.02 and 0.03 mg l"1
for two successive periods of 9 days at 12 C while feeding them at a ration
of 1.5 and 2.0% of body weight. Cyanide had an initial drastic effect, caus-
ing an almost complete arrest of growth at 0.03 mg l" (Figure 5), but again
the growth of cyanide-exposed trout showed a marked rebound in the second 10-
day period. However, this response was not sufficient to compensate for the
initial depressive effect of cyanide. Exposure to cyanide for 18..days re-
sulted in significant reductions of growth at 0.02 and 0.03 mg 1 HCN.
Speyer (1975) reached similar conclusions with rainbow trout tested at 0.02
mg 1-1 HCN and at 11 C.
McCracken (unpublished research, Department of Biological Sciences,
Concordia University) considered three important bioenergetic components of
fish growth affected by cyanide: activity, food ration, and size of the
fish. Using a series of annular growth chambers equipped with motor-driven
paddle wheels to maintain a constant.current (Kruzynski 1972), he measured
the effects of cyanide at 0.01 mg 1", using different food rations on young
rainbow trout swimming at 12 cm sec" at 10 C. The results shown in Figure 6
suggest a size-related response. Whereas the small fish (8g) showed no re-
sponse to cyanide at the different feeding levels it appears that for the
larger fish (18g) cyanide markedly impaired food utilization with increasing
ration. It should be noted that the food maintenance requirements (zero
growth) does not seem to have been affected by cyanide as was the case for
methoxychlor which markedly increased food requirements of brook trout
(Salvelinus fontinalis) tested under similar conditions (Oladimeji and
Leduc 1975). On the other hand when llg rainbow trout were simultaneously
160
-------
LU
o
(T
UJ
£L
140
130
120
LJ
c_> 100
5 90
CK
O
LJ. 80
CC
u
tt- 70
LJ
UJ
IT
60
50
40 _
GROWTH
(WET WEIGHT GAIN)
,CICHLIDS (24 DAYS,25°C)
I
COHO SALMON
(24 DAYS, I6'C)
RAINBOW TROUT
(20 DAYS, ire)
CICHLIDS
(36 DAYS,25°C)
I
I
.01 -02 .03 .04 .05 .06 -07 -08 -09 JO
HCN (mg/l)
Figure 5. Relative growth index of various species of fish exposed to
chronic cyanide poisoning throughout the experimental periods.
The studies with cichlids and coho salmon were performed by
Leduc (1966); rainbow trout were tested by Dixon (1975).
-------
+ 240,
-200
0 .5 1.0 1.5 0 .5 1.0 1.5 2.0
FOOD RATION (%/g FISH / DAY )
Figure 6 Comparison of the effects of various food rations during exposure
to 0.01 mg I'1 HCN between two size groups of rainbow trout at
10 C. (From unpublished McCracken data.)
162
-------
tested at 6, 12 and 20 cm sec-land fed one % of their body weight,,McCracken
(unpublished) found no significant effect of cyanide at 0.01 mg 1~ (Figure
7). These results suggest that the differential effect of cyanide noted
above is related to size more than activity.
The effect of size on the response of fish to toxicant has received
some attention in the past, mainly at acutely toxic concentrations. It is
noteworthy that Herbert and Merkens (1952) have clearly shown that the acute
toxicity of potassium cyanide was greater to large rainbow trout than to
smaller ones. On the other hand, Spear and Anderson (1975) found a reverse
relation with the pumpkinseed sunfish (Lepomis gibbosus) exposed to acute
levels of heavy metals.
In the natural environment fish are more likely to be exposed to a
mixture of toxicants rather than single ones. In the vicinity of certain
mines cyanide and arsenic occur together. Cyanide is used as a flotation re-
agent while arsenic leaches out of solid tailings from the oxidation of ar-
senopyrite. Speyer and Leduc (1975) found that exposure of rainbow trout to
mixtures of arsenic and cyanide at.the following concentrations: 0.02 mg I'1
HCN and 3.0 mg I'1 AS; 0.02 mg 1 HCN and 6.0 mg I"1 As, produced greater
growth impairment than either arsenic or cyanide tested separately (Figure 8)
the effects being additive following Finney's (1971) formula. These results
also showed differential effects of cyanide and arsenic on wet, dry and fat
weight gains (see Figure 8). This suggests on the one hand a greater water
retention in the poisoned fish than in the control due to some osmoregulatory
failure. Fat biosynthesis on the other hand, a high energy process, was ob-
viously hard hit by cyanide poisoning as can be expected from a chemical
acting directly on the respiratory-energy reaction chain. This phenomenon
has been further demonstrated by Dixon (1975) and McCracken (unpublished).
The ecological implications of the disturbance of fat synthesis are im-
portant indeed to the survival of fish populations in nature where adequate
fat reserves are essential for survival during adverse conditions and for yolk
deposition in the maturing ovaries. The critical needs of fat deposits in
yellow perch in nature were demonstrated by Newsome and Leduc (1975) and
shown in Figure 9 which illustrates the seasonal changes of fat in sexually
mature male and female yellow perch (Perca flavescens). During the fall, in
the mature females, there is an important translocation of fat from the body
to the ovaries which reduces the body fat content to a little over 2%, a level
barely sufficient to sustain survival during the winter prior to spawning.
This conversion is believed to be the cause of high winter female mortality
which accounts for the low proportion of females (20%) in the populations of
yellow perch inhabiting the cold mountain lakes of the Laurentians in which
they were introduced about 30 years ago. It seems that these low productive
lakes do not afford sufficient food for fat deposition to meet the maintenance
and egg production by the females.
Looking at Figure 5 it would appear that there are specific differences
of sensitivity to cyanide, cichlids being more resistant and rainbow trout
the least. There are undoubtedly some specific differences but temperatures
could have played a major role; cichlids were tested at 25 C., coho salmon at
16 C. and rainbow trout at 11 C. Recent findings by Kovacs (unpublished re-
-------
B 10
o:
LU
a.
8
X
o
LJ
I-
LU
O
CONTROL O
HCN •
06 12 18 24
SWIMMING SPEED (cm/sec)
Figure 7. Relationship between the growth of control and cyanide-exposed
rainbow trout (11.5g) held at different current velocities and
at 10 C for 20 days. (From unpublished McCracken data.)
164
-------
O O ARSENIC
• ARSENIC +0,02mg/l HCN
0.02 mg/I HCN
WET WEIGHT
DRY WEIGHT
1 i I I I
I I I I I I
FAT WEIGHT
01 23456
ARSENIC CONC. (mg/l As)
Figure 8. Effects of arsenic and cyanide, singly or in combination on the
wet, dry and fat gains of rainbow trout after 21 days and at
11 C. (From Speyer and Leduc 1975.)
-------
14
12
10-
8
6
4
2
o — OIMMATIWE FEMALES
n - D IMMATUM MALf.5
INTACT FEMALES
O—O FEMAUS LESS OVMICS
•—• INTACT MAUS
1 . . I . . I . . I
M JO
JAN
AM MMT JUN JUL
MEAN SAMPLE DATE
Figure 9. Seasonal fat content in immature and mature yellow perch in a
Laurential lake. (From Newsome and Leduc 1975.)
166
-------
search, Department of Biological Sciences, Concordia University) have now
confirmed that cyanide is more toxic at low temperatures than at higher
ones, both at the acute and sublethal levels. Tests were carried out
with juvenile rainbow trout at 6, 12, and 18 C.
SWIMMING ABILITY
One effective way to evaluate the effects of environmental factors on
fish activity is to measure their swimming ability. This response certainly
has important ecological implications considering migration, maintaining
position in a current or movements required in predator-prey interaction.
Compared to growth, swimming requires a rapid mobilization of energy re-
serves and therefore relies for its performance on the well functioning of
organs and intermediary metabolism. It is under stress, such as during swim-
ming, that the overall fitness of an animal can be better evaluated than un-
der the relatively passive conditions that exist when fish growth is meas-
ured in a tank, free of any rigorous swimming requirements.
Under low swimming velocities tested at 6, 12, and 20 cm sec ,
McCracken (unpublished) found no effect of cyanide (0.01 mg 1-1 on the
growth of rainbow trout fed at 1.0% of their body weight at 10.0 C. However,
cyanide had a profound effect on the swimming ability of fish tested at
higher velocities. Various studies summarized in Figure 10 illustrate the
great sensitivity of fish to chronic cyanide poisoning. Cichlids were tested
at 33.0 cm sec"1 and 25 C by Leduc (1966), rainbow trout at 47.0 cm sec"1 and
11 C by Speyer (1975), coho salmon at 48.8 cm sec-1 and 15 C by Broderius
(1970) and brook trout at 55.8 cm sec"1 and 8 C by Neil (1957). As noted
earlier for growth, the striking differences of swimming results cannot be
explained as reflecting simply the difference in cyanide tolerance of the
salmonids on the one hand and the cichlids on the other. The lower temper-
atures at which the salmonids were exposed during cyanide poisoning probably
account for much of the apparent greater sensitivity of the salmonids.
It is also important to note that resumption of normal swimming ability
after return to clean water is a very slow process, taking 15 to 20 days
(Neil 1957; Broderius 1970). These results are indicative of very serious
metabolic impairment by cyanide, and of inhibition of the oxidative pathways
responsible for the maintenance of swimming. However, since the inhibitory
action of cyanide on cytochrome oxidase is reversible (Stannard and Horecker
1948) and since cyanide is a non-cumulative cytoplasmic poison (Hewitt and
Nicholas 1963), one would expect a rapid recovery after removal from a toxic
environment unless there had been structural damage caused to the fish by
the toxicant.
RESPIRATION
Respiration rate is widely used in physiology as a biological parameter
integrating the overall metabolic activity of an animal in response to spe-
cific environmental entities. With fish, fundamental knowledge was acouired
and new physioloecological concepts arose from metabolic rate studies (pry
1947). Changes at the metabolic- or organ-functioning levels, or both, are
reflected by changes in oxygen consumption and can therefore be a useful
-------
100.
.01 .02 .03 .04 .05 .06 .07 .08 .09
HCN (mg/l)
.10
Figure 10. Effects of cyanide on the swimming endurance of various species
of fish. (See text for details.)
168
-------
diagnosis approach in fish toxicology. Indeed this proved to provide some
explanation to the effects of chronic cyanide poisoning of fish when Dixon
(1975) measured the resting metabolic rate of rainbow trout after exposure
to the toxicant. After two 19-day growth experiments where groups of rainbow
trout had been exposed to 0.0, 0.01, 0.02 and 0.03 mg I"1 HCN, individual
trout were placed in black tube flow-through respirometers and the oxygen con-
sumption measured for 6 consecutive days. The general pattern of the results
is shown in Figure 11.
The metabolic rate of the controls progressively dropped probably as a
result of quietening and starving. The cyanide-poisoned fish, however, first
boosted their respiration, then stabilized it at a higher level than the con-
trol. We may speculate that the initial rise reflects a surge of oxidation
of accumulated reduced metabolites upon return to clean water after 19 days
of cyanide exposure. This stabilized higher metabolic rate may also be in-
dicative of permanent damage by the previous exposure to cyanide that would
reduce the metabolic efficiency of the fish or impose an extra metabolic load,
both conditions resulting in an increased oxygen consumption. This increased
cost of operation cannot but reduce the Scope of Activity (Figure 1) and may
explain the very slow recovery of swimming performance of brook trout and coho
salmon after previous exposure to chronic cyanide poisoning (Neil 1957;
Broderius 1970). One has to look at some basic physiological functions before
attempting any explanation of this increased metabolic rate. Osmoregulation
appears to be promising, considering its bioenergetic and ecological impli-
cations.
OSMO- AND IONOREGULATION
Osmoregulation accounts for an important fraction of the basic metabolic
rate of a freshwater fish which has to continually excrete an excess of
water brought in by the osmotic gradient; this work also increases with ac-
tivity. This phenomena has been well demonstrated by Rao (1968) who has cal-
culated that Osmoregulation may account for up to 20% of the active metabolic
rate of rainbow trout. lonoregulation also is a critical function which
enables the fish to maintain the proper ionic strength in its tissues. The
studies of osmo- and ionoregulation related to the action of toxicants are of
double interest. They may provide highly informative clues to the overall
performance of a poison while in freshwater and/or when salmonid smolts mi-
grate to sea. Along these lines, Leduc and Chan (1975) have exposed rainbow
trout to cyanide (0.01, 0.015, 0.021, 0.028 and 0.037 mg I'1 HCN) in renewed
freshwater (10 C) for 28 days, then transferred them to artificial seawater
(10 C). at 19.1 ppt but containing no cyanide; later the fish were re-
turned to freshwater.
Cyanide affected both osmo- and ionoregulation in saltwater and in fresh
water. Figure 12 shows that after 260 hours in saltwater the plasma chloride
and plasma concentration were higher in cyanide-exposed fish, whereas upon re-
turn to freshwater (Figure 13) the reverse occurred, indicating loss of chlo-
ride and higher water content. In another test, we have shown that cyanide
had an immediate effect on osmo- and ionoregulation when saltwater-adapted
trout were transferred into freshwater-cyanide tanks (Figure 14). These
changes may look small and of questionable ecophysiological significance but
-------
(T
tn
.38
O
f •»«
.30
O
O
UJ
X
O
.26
.22
CONTROL
I
2345
DAYS AFTER EXPOSURE
Figure 11. Generalized pattern of the resting metabolic rate in clean water
of rainbow trout finger!ings comparing control fish to those
which had been exposed to low cyanide concentrations (0.01-03
mg 1 ) for 18 days at 11 C. (Modified from Dixon 1975.)
170
-------
tf
FRESHWATER
28 days
-HCNJ
7T
SALTWATER
260 hours
•5 15
.£
4>
a
6
a>
o
*
10
O Expt. I
• Expt. 2
I
o
£ l5
41
O»
J 10
o
S
^
41
a.
O Expt. I
• Expt. 2
0.0 0.01 0.02 0.03 0.04 0.0 0.01 OX>2 003 0.04
HCN CONCENTRATION (mg/l )
Figure 12. Relationships between the changes of plasma composition that
occurred during a 260-hour salinity (18.9 ppt) tolerance test,
and the cyanide concentrations to which juvenile rainbow trout
had been exposed for 28 days in flow-through aquaria at 10 C.
(From Leduc and Chan 1975.)
171
-------
0.0 0.01 0.02 0.03 0.04 0.0 0.01 0.02 0.03 0.04
HCN CONCENTRATION (mg/l)
Figure 13. Relationships between the changes of plasma composition that
occurred during fresh water exposure following transfer of the
fish from salt water, and the cyanide concentrations to which
juvenile rainbow trout had been exposed for 28 days in flow-
through aquaria at 10 C. (From Leduc and Chan 1975.)
172
-------
—
SALTWATER
W ^ 1
FRESHWATER-HCN 1
376 hours
9 6 hours
_o
o
(ft
0 -5
a
£
O
-10
£ -15
V
o
o
-5
-10
-15
-20
0.0
-20
•o
0.01 0.02 0.03 0.04 0.0 0.01 0.02 0.03 0.04
HCN CONCENTRATION (mg/l)
Figure 14. Relationships between the changes of plasma composition that
occurred during a 4-day fresh water-cyanide exposure following
transfer from salt water (18.9 ppt), and the cyanide concen-
trations to which juvenile rainbow trout were exposed in fresh
water flow-through aquaria at 10 C. (From Leduc and Chan 1975.)
173
-------
Wood and Randall (1973) have demonstrated that changes in water content as
small as 1% correspond to marked increase in urine flow which is maintained
fay glomeruli filtration concommittant with high energy expenditures. It
therefore appears that cyanide can have subtle effects measured in terms of
water content in the fish but which are indicative of serious physiological
impairment.
HISTOPATHOLOGICAL EFFECTS OF CYANIDE
Cichlids grown at 0.09 and 0.01 mg 1~* HCN not only had a markedly
reduced swimming performance, but some showed serious body injuries at the
end of the tests. Scales were falling off and short handling with a dip net
was sufficient to severely damage the fin rays. Some also showed swelling of
the body and extensive subcutaneous hemorrhaging. After the tests all fish
were returned to clean water (no current) for observation and some died with-
in 24 hours (Leduc 1966). No further examination was carried on these
cichlids, but Dixon (1975) and Ruby and Dixon (1974) have shown histopatho-
logical effects of cyanide on rainbow trout. Dixon (1975) found that a
9-day period of exposure to 0.01 mg I"1 HCN was sufficient to induce exten-
sive necrobiosis in the liver; gill tissue from the same fish showed no
apparent cyanide-induced histopathological damage. Ruby and Dixon (1974)
demonstrated blockage of mitosis in the testis of rainbow trout. Not only
was the number of dividing spermatogonia reduced by a previous exposure to
0.01 mg 1~1, but of those cells that were dividing, mitosis was blocked in
prophase with virtually no dividing spermatogonia reaching the later stages.
Under these circumstances spermatogenesis would be completely arrested,
preventing reproduction.
OVERALL SIGNIFICANCE AND CONCLUSION
We have shown that cyanide could markedly reduce the performance of
several physiological functions of fish tested in the laboratory, but it is
difficult to arrive at an overall significant judgement as to the toxicity of
cyanide to fish unless one can sythesize various responses into one. Not all
of the functions tested are of equal importance nor were they equally affected
by cyanide. In an attempt to arrive at an overall evaluation of the chronic
toxicity of cyanide, the approach taken by Warren, Doudoroff and Shumway (1973)
was followed. It consists in drawing on the same graph experimentally ob-
tained relative performance curves plotted against test concentrations while
giving a value of 100 to the controls. Performance curves have been plotted
in Figure 15 and a Relative Performance Index curve was then drawn, trying,
to the best of our judgement to integrate in one line (heavy trait) a gen-
eralized response curve to cyanide. The Relative Performance Index curve
suggests a 50% reduction of total performace at 0.01 mg/1, and further sug-
gests that even though fish could survive indefinitely at 0.03 mg r1 HCN in
the laboratory, the different physiological requirements necessary to survive
in nature could not be met.
The Relative Performance Index drawn out in Figure 15 was used in Figure
16 to model a Scope for Activity stressed by cyanide. The model shown in
Figure 16 points out that, due to a marked reduction in active metabolism
174
-------
LU
O
OC
Id
Q.
X
Ld
Q
UJ
O
a:
LJ
a.
LJ
UJ
or
*"~-Osmoregu lotion
.01
.02
.03
.04
.05
HCN CONCENTRATION (mg P)
Figure 15.
Relative performance of fish measured on various physiological
responses to long term exposure to sublethal concentrations of
cyanide. Details of the different experimental approaches are
given in the text. The heavy line is a generalized estimate of
the effects of cyanide on the overall performance of fish or a
Relative Performance Index.
175
-------
POTENTIAL RANGE OF ACTIVIT
Y WITHOUT STRESS-*
.02 ,03 .04
"1
0 .01
HCN (mg I") .
ZONE OF TOLERANCE— H
Figure 16. Comparison between a theoretical scope for activity of a fresh
water fish without stress (open dotted area)'and that under the
effect of chronic cyanide poisoning at about 11 C (shaded area),
176
-------
by cyanide, both the scope and range of activity were drastically reduced com-
pared to that of an hypothetical unstressed fish.
This model was then applied to actual activity values published by
Beamish (1964) who measured the routine scope for activity of brook trout at
5, 10, and 15 C. The model in Figure 17 shows that the routine scope for
activity would be reduced to zero by 0.03 mg 1~* HCN and by 50% at 0.01 rug
1~^. There are, however, two additional points to consider. The measure-
ments of the routine scope for activity of brook trout published by Beamish
(1964) showed a maximum at 10 C. The optimal temperature varies with dif-
ferent species of fish but if these data are directly transposable to the
natural environment, one would expect to find the brook trout most successful
at 10 C whereas 5 and 15 C would curtail the abundance and distribution of
this species. If, at these temperatures, (5 and 15 C) with safety factors
lower than at 10 C, the population was stressed by a toxicant it would exper-
ience such a reduction in its capacity to reproduce that it could lead to the
extinction of the population. With reference to cyanide, this effect would
be even greater at 5 C than suggested on our model since, as mentioned before,
cyanide, at low concentrations is more toxic at low temperature.
At present it is recommended that the level of cyanide in water never
exceeds 0.005 mg I'1 at any time or place (National Academy of Sciences and
National Academy of Engineering 1972J. According to our model, concentrations
between 0.01 and 0.005 mg 1~1 HCN would reduce the fish performance or scope
for activity by 30-50% (Figure 13). What reduction of scope can be accepted
as an application factor remains a difficult and somewhat arbitrary decision
to make. With regards to oxygen alone, Fry (1960) suggested that a 50% re-
duction could be a reasonable estimate of the oxygen requirements of fish in
nature. This view has however not been supported by Doudoroff and Shumway
(1970) and, to our knowledge, this question has been not considered with
toxicants. If Fry's suggestion of accepting a 50% reduction of scope is
taken, the recommended level of cyanide in water would fall within the values
of 0.01 mg 1~1 proposed by Jones in 1964 and 0.005 mg I"1 established by
the Environmental Protection Agency (.National Academy of Sciences and National
Academy of Engineering 1972). There is however some reservation to the
acceptance of these criteria.
Histopathological observations have revealed some very deleterious
effects of cyanide at 0.01 mg 1~1, a concentration that otherwise showed no
marked effect on growth. Also, the blockage of spermatogenesis and, possibly,
oogenesis could have dramatic effects leading to the disappearance of entire
populations resembling the effects of acid pollution on lakes in northern
Ontario (Beamish 1974).
Most water quality criteria developed for North America have undoubtedly
been developed to protect waters of the temperate regions. If, as our recent
studies suggest, cyanide chronic toxicity increases with decreasing temper-
ature then the recommended levels may not be safe under cold climates such
as in northern Canada and Alaska where cyanide is extensively used by the
mining industry.
177
-------
10 15 20
TEMPERATURE °C
Figure 17. Estimate of the effects of cyanide on the routine scope for
activity of a salmonid fish at different temperatures. This
model was drawn by applying the Relative Performance Index values
at different cyanide concentrations as determined from Figure 15
to an hypothetical salmonid fish using Beamish (1964) routine
scope for activity data obtained on brook trout, Salvelinus
fontinalis.
178
-------
Finally, further studies of cyanide should concentrate at levels lower
than 0.01 mg I"1 HCN. The lack of studies carried at lower concentrations is
mainly due to the lack of a suitable method of detection below 5 ppb which is
the lower limit of the techniques currently used.
ACKNOWLEDGEMENTS
The pursuit of scientific knowledge is a long, arduous, sometime obscure
path, but the presence of leadership, inspiration and close collaboration
along the way make achievement possible. This research on one common pollu-
tant, carried over many years with the hope that some benefit to our environ-
ment will be derived, would not have materialized without many inspirational
associations. I hereby wish to recognize the leadership of Dr. Peter
Doudoroff and Dr. Charles E. Warren, of Oregon State University, and the
generous assistance of George Chadwick at the beginning of these studies.
May I also acknowledge the excellent collaboration of Dr. Sylvia M. Ruby,
Associate Professor of Biology at Concordia University, and of graduate stu-
dents, Ken S. Chan, D. George Dixon, Ian R. McCracken, George E. Newsome,
Menno R. Speyer, Tibor G. Kovacs, Adebayo A. Oladimeji, Walter Banas, jr. and
George M. Kruzynski.
I also wish to acknowledge the financial support from the National Re-
search Council of Canada, the Department of Indian and Northern Affairs of
Canada (ALUR) and the Department of Education of the Province of Quebec
(FCAC).
179
-------
REFERENCES
Beamish, R. J. 1974. Loss of fish populations from unexploited remote lakes
in Ontario, Canada as a consequence of atmospheric fallout of acid.
Water Res. 8(l):85-95.
Beamish, F. W. H. 1964. Respiration of fishes with special emphasis on standard
oxygen consumption. II. Influence of weight and temperature on
respiration of several species. Canadian J. Zool. 42(2)-.177-188.
Broderius, S, J. 1970. Determination of molecular hydrocyanic acid in water
and studies of the chemistry and toxicity to fish of the nickelocyanide
complex. M. S. Thesis. Oregon State Univ., Corvallis. 93 p.
Commoner, B. 1940. Cyanide inhibition as a means of elucidating the mech-
anisms of cellular respiration. Cambridge Philosophical Society, Biol.
Rev. 15:168-201.
Dixon, D. S. 1975. Some effects of chronic cyanide poisoning on the growth,
respiration and liver tissue of rainbow trout. M.S. Thesis.
Concordia Univ., Montreal. 77 p.
Doudoroff, P., and D. l_. Shumway, 1970. Dissolved oxygen requirements of
freshwater fishes. FAO Fish. Tech. Paper 86. Food & Agric. Organ.
of the U.N., Rome, xi + 291 p.
Doudoroff, P. 1976. Toxicity to fish of cyanides and related compounds;
a review. Ecol, Res. Ser. EPA 600/3-76-038. Office of Res. & Devel.,
Environ. Res. Lab., U.S. Environmental Protection Agency, Duluth,
Minn, vi + 154 p.
Finney, D. J. 1971. Probit analysis. 3rd ed. Cambridge Univ. Press.
xv + 333 p.
Fry, F. E. J. 1947. Effects of the environment on animal activity. Univ.
Toronto Studies Biol. Ser. 55 Publ. Ontario Fish. Res. Lab. 68 The
University of Toronto Press. 62 p.
Fry, F. E. J. 1960. The oxygen requirements of fish. pp. 106-109 jm_
Biological problems in water pollution. (Trans, of the 1959 seminar).
Tech. Rept. W60-3. Robert A. Taft Sanitary Eng. Center, U.S. Publ.
Health Serv., Cincinnati, Ohio, xv + 285 p.
Herbert, D. W. M., and J. C. Merkens. 1952. The toxicity of potassium
cyanide to trout. J. Exp. Biol. 29(4):632-649.
Hewitt, E. J., and D. J. D. Nicholas. 1963. Cations and anions: inhibitions
and interactions in metabolism and in enzyme activity. Pp. 311-436 jn_
R. M. Hochster and J. H. Quastel (eds), Metabolic inhibitors; a
comprehensive treatise. Vol. II. Academic Press, London, xviii +
753 p.
180
-------
Hunsman, A. G. 1948. Method in ecology — biapocrisis. Ecology 29(0:30-42.
Iverson, S. L., and J. E. Guthrie. 1969. The ecological significance of
stress. The Manitoba Entomol. 3:23-33.
Jones, J. R. E. 1964. Fish and river pollution. Butterworths, London,
viii + 203 p.
Keilin, D., and T. E. King. 1960. Effect of inhibitors on the activity of
soluble succinic dehydrogenase and on the reconstitution of the succinic
dehydrogenase-cytochrome system from its components. Free. Royal Soc.
London, Ser. B 152 (947):163-187.
Kruzynski, G. M. 1972. Effects of dietary methoxychlor on brook trout
Sajye]jn us f ontina 1 is. M.Sc. Thesis. Sir George Williams University,
Montreal, 131 p.
Leduc, G. 1966. Some physiological and biochemical responses of fish to
chronic poisoning by cyanide. Ph.D. Thesis. Oregon State Univ.,
Corvallis, 146 p.
Leduc, G., and K. S. Chan. 1975. The effects of chronic cyanide poisoning
on the tolerance of rainbow trout to varying salinity, pp. 118-125
i_n_ T. C. Hutchinson (ed.), Water Pollution Research in Canada 1975,
vol. 10, incorporating the Proceedings of the Tenth Canadian Symposium
on Water Pollution Research, held at the University of Toronto,
February 1975, 236 p.
Leduc, G. 1976. The effects of cyanide on developing Atlantic salmon embryos.
Submitted for publication to J. Fish. Res. Bd. Canada.
National Academy of Sciences and National Academy of Engineering. 1973.
Water Quality criteria 1972. A report of The Committee on Water Quality
Criteria, Environmental Studies Board. Ecol. Res. Ser. EPA-R3-73-033.
U.S. Environmental Protection Agency, Washington, D. C. xix + 594 p.
Neil, J. H. 1957. Some effects of potassium cyanide on speckled trout
Salvelinus fontinalis. Pp. 74-96 _in. Papers presented at the 4th Ontario
Industrial Waste Conference, Honey Harbor, Ontario. Waste & Poll.
Advisory Comm., Ontario Water Resources Comm., Toronto. 156 p.
Newsome, G. E., and G. Leduc. 1975. Seasonal changes of fat content in the
yellow perch (Perca flavescens) of two Laurential lakes. J. Fish. Res.
Bd. Canada 32 (TTJ72214-2221.
Oladimeji, A. A., and G. Leduc. 1975. Effects of dietary methoxychlor on
the food maintenance requirements of brook trout. Prog. Water Techno!.
(Pergamon Press) 7(3/4):587-598.
Rao, G. M. M. 1968. Oxygen consumption of rainbow trout (Salmo gairdneri)
in relation to activity and salinity. Canadian J. Zool. 46(4):781-786.
181
-------
Ruby, S. M., and D. G. Dixon. 1974. Effects of sublethal concentrations
of cyanide on reproduction in immature rainbow trout. Paper presented
at the Aquatic Toxicity Coordination Workshop, Freshwater Institute.
(Held in Winnipeg, Aug. 1974.)
Spear, P., and P. D. Anderson. 1975. Fish size as a quantitative function
of tolerance to heavy metals. Pp. 170-178 jr^ T. C. Hutchinson (ed.),
Water Pollution Research in Canada, 1975, vol. 10, incorporating the
Proceedings of the Tenth Canadian Symposium on Water Pollution Research,
held at the University of Toronto, February 1975.
Speyer, M. R. 1975. Some effects of chronic combined arsenic and cyanide
poisoning on the physiology of rainbow trout. M.Sc. Thesis. Sir
George Williams Campus, Concordia Univ., Montreal, 76 p.
Speyer, M. R., and G. Leduc. 1975. Effects of arsenic trioxide on the growth
of rainbow trout. Pp. 17-19 jn_ Abstracts of International Conference on
heavy metals in the environment. Held October 27-31, 1975, in Toronto,
sec. C.
Stannard, J. N., and B. L. Horecker. 1948. The in vitro inhibition of cy-
tochrome oxidase by azide and cyanide. J. Biol, Chem. 172(2}:599-608.
Sumner, F. B., and P. Doudoroff. 1938. Some experiments upon temperature
acclimatization and respiratory metabolism in fishes. Biol. Bull.
74(3):403-429.
Warren, C. F. 1971. Biology and water pollution control. Saunders.
Philadelphia 434 p.
Warren, C. E., P. Doudoroff, and D. C. Shumway. 1973. Development of
dissolved oxygen criteria for freshwater fish. Ecol, Res. Ser.
EPA-R3-73-019. Office of Research & Monitoring, U.S. Environmental
Protection Agency, Washington, D.C. xviii + 121 p.
Wood, C. M., and D. J. Randall. 1973. The influence of swimming activity on
water balance in the rainbow trout (Salmo gairdneri). J. Comp. Physio!.
82(3):257-276.
182
-------
AN ASSESSMENT OF APPLICATION FACTORS
IN AQUATIC TOXICOLOGY
D. I. Mount, Ph.D., Director
Environmental Research Laboratory—Duluth
U. S. Environmental Protection Agency
6201 Congdon Boulevard
Duluth, Minnesota 55804
ABSTRACT
In the early 1950's, application factors to estimate
"safe" concentrations from LCSO's were proposed. Later,
an experimental method of estimating the numerical value
of the application factor was proposed to replace
arbitrary values such as 1/10. Both measured values and
arbitrary ones have been widely employed in water quality
criteria by regulatory agencies. An examination of the
data base for establishing application factors for
various pollutants in different water types and for
various species, reveals an unacceptable spread in their
numerical value. Several factors such as chemical effects
of the water on the pollutant, experimental error and
biological variability must be contributing to this spread
thereby making a determination of their real validity
difficult. A better method to predict concentrations that
will not affect survival, growth, and reproduction is
needed for present toxicological requirements.
Aquatic toxicologists of today are faced with an enormous pressure to
provide decisions regarding the potential effects of hundreds of chemicals
should they be released into the environment. While everyone agrees that
"more research is needed," we also must realize that urban and industrial
development will proceed regardless of the need for more research, and
therefore decisions must be made now. Some of us may find untenable the
passage of laws before sufficient data are available to confidently make the
needed decisions. As scientists we must agree that decisions or predictions
based on skimpy data can still be scientific. As long as the basis of the
decision and the proper confidence limits are provided, scientific integrity
is maintained. Indeed, is not science in essence a process of "concluding
from available data" that which can be concluded rather than requiring a
fixed or predetermined amount of information.
183
-------
It is in this framework that today's aquatic toxicologists must render
"scientific judgments" to meet today's problems. Answers derived from
available data, however scant, are better than pure guesses.
The Toxic Substances Control Act passed just prior to this writing,
brings to the field of aquatic toxicology demands for decisions that far
exceed any experienced heretofore. Under this new law, decisions on environ-
mental as well as other effects must be made long before a proposed product
is in use or introduced into the environment. Of necessity, then, regulatory
decisions will have to be based on laboratory studies of a relatively few
species and predictions made for many other species. Only after the early
decisions have been made regarding acceptable concentrations will field data
be obtained to verify predicted effects, because only then will the product
be in use and extant in the environment to produce these effects. Against
this background I want to discuss the present status of the application factor
hypothesis (Mount and Stephan 1967) for use in predicting acceptable concen-
trations for aquatic organisms.
Before proceeding, I want to acknowledge the difficult and laborious
effort of a special staff committee of the Environmental Research Laboratory
in Duluth. This committee, composed of Robert Andrew, Duane Benoit, John
Eaton, James McKim and Charles Stephan, now have in press their report
entitled "An Evaluation of an Application Factor Hypothesis" for which they
have sorted and assembled most of the toxicity data base pertinent to the
validity of experimentally derived application factors. Without their report
as source material, this paper could not have been presented at this time.
Let us first summarize the considerations that must be made when one
predicts the acceptable concentrations for aquatic organisms. I will leave
to others the difficult chore of extrapolating results from species to com-
munities to ecosystems; instead I will focus my comments on predicting effects
from test specimens to species in their normal niches.
Any prediction must consider whether the animals tested are typical of
the species as a whole. Certainly one would not choose inbred or geographi-
cally isolated populations if the data are to be broadly applicable and are
expected to account for extant species variability. Since one nearly always
is concerned about protecting more than a single species, the difference in
sensitivity between species also must be considered. The existing toxicolog-
ical data adequately demonstrate substantial differences between species for
many toxicants.
The physical and chemical changes brought about by the common components
in surface waters and the resulting changes in the toxicant, were recognized
early. Much early aquatic toxicological work involved metals and the effect
of pH and hardness on their toxicity. So great were these effects that inclu-
sion of "water hardness" effects on toxicity (even when there is no reason to
expect effects) have been routinely included in subsequent experimental work.
Even pollution control administrators who know little about the subject will
raise this issue in the standard setting process. Unfortunately, I think we
have failed to recognize the important role played by other components, for
184
-------
example, suspended solids. Clay or algal cells and chelating agents such as
hunric substances certainly must affect the toxicity of many contaminants,
particularly the less water soluble synthetic organics.
The length of exposure time to the toxicant and the organism's life
history stages likely to be exposed are other important considerations.
Often one does not have chronic exposure data for many species when acceptable
concentrations must be predicted. We recognize, however, that few or no
generalizations can be made about the shape of time-effect curves beyond
"acute" periods of exposure, thus leaving much uncertainty about the effect
of exposure length on acceptable concentrations and no way to extrapolate
effects for longer, untested periods of time.
Recently, the propensity of chemicals to form residues that produce
harmful effects has become an important concern in toxicity predictions.
These concerns are of two general types: 1) Accumulations that produce
objectional flavor, and 2} acculumations, usually of persistent chemicals,
that reach concentrations that are toxic to consumers of the organism bearing
the accumulations.
Finally, another concern when predicting acceptable toxicant concentra-
tions for aquatic organisms relates to the quality of the test animals. How-
ever, because that concern is relevant principally to the quality of data
obtained, I will not consider it further in this discussion. Many more
concerns could be identified, but these are sufficient to give a "feel" for
the complexity involved in predicting toxicity.
Let us now focus on one of the proposed concepts for predicting accept-
able concentrations—the application factor approach. Probably Hart,
Doudoroff and Greenbank (1945) were among the first to suggest the use of
application factors to (as they termed it) predict biologically safe concen-
trations. Even though aquatic toxicology was hardly born at the time they
published, their concepts and perceptions are still very much "on the mark"
and surprisingly current. Their approach included compensation for different
sensitivities of various species, variable toxicity due to different receiv-
ing waters, and the effect of length of exposure. Their use of what in
essence is the slope of the time-mortality curve to make inferences about
cumulative toxicity is truly remarkable considering the embryonic state of
aquatic toxicology at that time. They clearly cautioned workers about
toxicological consequences resulting from reactions of the toxicant with
various water constituents.
In subsequent papers, workers in the field, especially Doudoroff,
frequently discussed the need to lower LC50 values in order to arrive at a
safe exposure concentration. Times were such, and aquatic toxicology was so
embryonic and unrecognized, that no one ventured opinions about how much
reduction was needed at that time.
Henderson (1957) in the first of three seminars on "Biological Problems
in Water Pollution," ventured forth with a reduction of 1/10 of the 96-hour
185
-------
TLm as a "tentatively suggested" value. He considered this a "sizeable"
reduction and even dared then to illustrate with an example in which he
arrived at a factor of 1/12!
These and subsequent papers all stayed principally with arbitrary values
and generally did not suggest how values might be derived experimentally,
although Hart et al. (1945) did use experimental data as a part of their
proposed formula.
Warren and Doudoroff (1958) may have been the first to propose experi-
mentally derived application factors. They used 30-day toxicity tests in
artificial streams to determine application factors for pulp mill wastes.
Just after the powerful 1965 Water Quality Act was passed and just
before the "environmental awakening," Mount and Stephan (1967) proposed a
method of experimentally deriving an application factor (AF) for each
toxicant. They suggested that if one divides the highest concentration
tested, in which no adverse effects during a life cycle test were found by
the 96-hour TLm, the fractional value might be characteristic of the toxicant
and constant for most or all fish species and water types. While they did
not so state, there appears to be a sound toxicological basis for expecting
the ratio to be constant. Specifically, the mode of action of a given
toxicant is similar for various species of fish, but the threshold concentra-
tions producing the effects are different for various species, thus producing
various species sensitivities. Since the stage in an organism's life history
most sensitive to a toxicant will vary between species, then any consistency
in the AF value is probably caused by chance rather than a predictable toxi-
cological principle. It is probable that any consistency between AF values
for other animals (such as invertebrates) and fish is unlikely since the mode
of action is probably different.
Mammalian toxicologists have also used a similar predictive approach.
For example, Hayes (1967) described a chronicity factor to characterize the
1 a 4"? \/£\ +r\ v *i r* 4 i"\t r\f r*V\ j-.m*5 r* a 1 f
I VI ^-/Y %*• I I 1^ • W } • t+f-J \r*J \I.SV* / Vl\**/%r< 111
cumulative toxicity of chemicals,
The following evaluation of AF's is based largely on data from the com-
mittee report cited previously. The opinions expressed, however, are mine
and not those of the committee.
The objective in using an AF approach is to integrate effects of varia-
able species sensitivity, length of exposure and effect of water character-
istics on toxicity, and to enable one to estimate acceptable concentrations
without long expensive tests on a large number of species and waters. If
the 96-hour LC50 divided into the MATC* is a reasonably constant value for
most fish species, then the AF multiplied by the LC50 for any species of
fish in any water type would estimate the acceptable concentrations for that
species. The data in Table 1 are summarized from Andrew et al. (1977). The
*Maximum concentration that caused no significant effect on the reproduction,
growth, and survival of test animals during a full life cycle.
186
-------
quotient of extremes for.the MATC is found by dividing the highest concentra-
tion in a set of lowest concentrations just producing an effect by the lowest
concentration in a set of highest concentrations not producing an effect.
The quotient of extremes for the AF is calculated by dividing the largest
numerical value of the AF by the smallest value in a set of values. In both
cases, a set can include only one toxicant, since the AF value is expected
to be different for different toxicants.
TABLE 1. VARIABILITY OF MATC AND AF VALUES
Number Quotient of
Toxicant of Tests Extreme Limits
MATC AF
Atrazine
Cadmium
Chromium (IV)
Copper
Diazinon
Lead
Lindane
Malathion
Methyl mercury
Zinc
3
3
2
6
2
2
3
3
3
4
8.0
20
20
14
28
4.0
2.7
161
13
46
5.8
5.3
35
13
136
2.4
5.0
3.5
17
206
If one compares the quotient values in the columns for MATC's and AF's
for each toxicant, one finds that—among the ten toxicants for which data
are given involving 31 chronic tests—the AF has less variability in five
and the MATC has less variability in five. However, six of the ten values
for AF's and MATC's are not significantly different.
These comparisons suggest rather convincingly (given our present ability
to measure MATC and AF values) that one gains no more accuracy in estimating
acceptable safe exposure concentrations by using an LC50 and an application
factor than if one simply selects an MATC and uses that value for all fish
species.
Obviously, in nearly every instance the true MATC will be lower than the
LC50. If MATC's for only one or a few species are known, then using that
value as an acceptable concentration probably will result in the selection of
a lethal concentration for especially sensitive species for some portion of
toxicants. Prudence certainly dictates that the acceptable concentration
should be set below the MATC by some margin to protect us from our ignorance.
In the absence of any chronic data, but when a prediction of an accept-
able concentration must be made, arbitrary reductions below the LC50's should
be made. The amount of reduction can be derived by generalizing from the
ratios of MATC's to LC50's for many substances chemically similar, or can be
based on an average value representative of all toxicants for which such data
187
-------
are available. In other words, I'm suggesting that even though the present
data base does not show AF's to be constant values, for arbitrary estimates
they are as good as any other bases.
Many practical experimental problems can reduce reproducibility and make
the concept appear invalid when it is not. Certainly our inability to
measure biologically active forms of the toxicant can produce vast errors.
Few data exist for judging the reproducibility of the MATC. We know that
LC50 values vary substantially. These two sources of error can, and undoubt-
edly sometimes do, cancel or supplement each other to produce more experi-
mental error.
On the other hand, data are accumulating, as for example McKim and
Benoit (1971), to show that different species differ in their most sensitive
stages. As stated earlier, this would seem to undermine the toxicological
basis for the AF concept as proposed by Mount and Stephan (1967).
I began this paper by emphasizing the need to predict toxicity with
minimum effort and maximum speed, and now I have ventured an opinion that
the most commonly used predictive method is not supported by the data base.
What, then, is an alternative?
While far from desirable, we can see that the use of a single MATC as
the acceptable concentration is at least as good as the AF. Both the cough
test (Drummond 1977) and the embryo larval test (Macek 1977; McKim 1977)
offer promise as more accurate and non-arbitrary methods to estimate accept-
able concentrations for a variety of situations. Given demands of present
legislation, no research need is greater than the development of rapid and
accurate screening methods to estimate toxicity. I am firmly convinced,
however, that we must continue to use the life-cycle chronic test as our
laboratory guidepost to assess the suitability of rapid screening tests.
Without a solid chronic toxicity data base, we will be unable to judge the
value of any other method to predict chronic toxicity. In the last 5-10
years, the chronic toxicity data base has increased many fold and provides
an understanding that should be helpful in our search for better predictive
methods. Field monitoring should be used to assess our overall ability to
predict effects resulting from the use of our predictions but not for
initially measuring acceptable concentrations.
The present need for establishing biologically acceptable concentra-
tions of as many as 1500 new products each year, makes crystal clear that
our past pace of data generation will have to be increased two to three
orders of magnitude. Either more resources must be obtained or else a
faster means to produce data must be found. Probably no method will always
be correct, and we may have to be content with being right "most of the
time." Perhaps never before have we faced a challenge so important to our
national welfare as the one produced by the information needs of the Toxic
Substances Control Act. Since the consequences of being unnecessarily
restrictive are different, but perhaps as severe as being too liberal, our
best effort will be none too good.
188
-------
In conclusion, the biological validity of the AF concept certainly is
not yet disproven, but the present data base is such that even if the concept
is biologically valid, the practical problems involved in the determination
of AF's make the approach of questionable utility. Furthermore, the present
data base implies that the MATC of one species will provide with greater ease
an equally accurate estimate of an acceptable concentration for other species
if a safety factor is also applied. In view of current needs, we must
rapidly improve our ability to predict acceptable concentrations for aquatic
organisms.
189
-------
LITERATURE CITED
Andrew, R. W., D. A. Benoit, 0. 6. Eaton, 0. M. McK1m, and C. E. Stephan.
1977. Evaluation of an application factor hypothesis. (In press.)
Drummond, R. A., G. F. Olson, and A. R. Batterman. 1974. Cough response and
uptake of mercury by brook trout, Salvelinus fontinalis, exposed to
mercuric compounds at different hydrogen-ion concentrations. Trans.
Am. Fish. Soc. 103(2): 244-249.
Hart, W. B., P. Doudoroff, and J. Greenbank. 1945. The evaluation of the
toxicity of industrial wastes, chemicals and other substances to fresh
water fishes. Waste Control Laboratory, Atlantic Refining Co.,
Philadelphia. 317 p. + 14 + 43 fig.
Hayes, W. 0., Jr. 1967. The 90-dose LD50 and a chronicity factor as
measures of toxicity. Toxicol. Appl. Pharmacol. 11(2): 327-335.
Henderson, C. 1957. Application factors to be applied to bioassays for the
safe disposal of toxic wastes. Pp. 31-37 j_n C. M. Tarzwell (ed.),
Biological Problems in Water Pollution. (Trans, of the 1956 seminar.)
R. A. Taft Sanitary Engineering Center, U. S, Dept. of Health, Educa-
tion, and Welfare, Cincinnati, Ohio. 272 p.
Macek. K. J, 1977. Utility of toxicity tests with embryos and fry of fish
in evaluating hazards associated with the chronic toxicity of chemicals
to fishes, Jhi F. L. Mayer and J. M. Hamelink (eds.) Proceedings of a
symposium on pesticides sponsored by A.S.T.M. Committee 1-35, Memphis,
Tenn. (Oct. 25-26, 1976) (In press.)
McKim, J. M. 1977. Use of embryo-larval, early juvenile toxicity tests with
fish to estimating long-term toxicity. (In press.)
McKim, J. M., and D. A. Benoit. 1971. Effects of long-term exposures to
copper on survival, growth, and reproduction of brook trout (Salvelinus
fontinalis}. J. Fish. Res. Bd. Canada 28(5): 655-662.
Mount, D. I., and C. E. Stephan. 1967. A method for establishing acceptable
toxicant limits for fish--malathion and the butoxyethanol ester of
2,4-D. Trans. Am. Fish. Soc. 96(2): 185-193.
Warren, C. E., and P. Doudoroff. 1958. The development of methods for using
bioassays in the control of pulp mill waste disposal. Tappi 41(8):
211A-216A.
190
-------
CLOSING REMARKS—AN OLD FROG CROAKS AN APPEAL FOR LOGIC
P. Doudoroff
Department of Fisheries and Wildlife
Oregon State University
Corvallis, Oregon 97331
First, I want to express to the sponsors—the Department of Fisheries
and Wildlife of Oregon State University and the United States Environmental
Protection Agency—my deep appreciation of the honor that has been accorded
me by the dedication of this symposium. To the head of my department, Dr.
Richard Tubb, who conceived this means of recognition of my services to the
department and my profession and who has worked diligently toward its success-
ful realization as a very special and memorable occasion at the time of my
retirement, and also to Mrs. Alma Rogers, who assisted him with the arrange-
ments, go my particular thanks. Also, to all the participants who have taken
the trouble to prepare papers for presentation here—contributions that have
been of great interest to me—and all those who have traveled long distances
to attend this symposium or have written to me to extend their greetings, I
am truly grateful. The cosponsorship of the symposium by EPA is signally
gratifying to me. Though never an employee of EPA—sometimes even its
opponent in adversary proceedings —I have felt since its inception as though
I were a kind of honorary member. My many years as a water pollution biolo-
gist with the U.S. Public Health Service and the encouraging support and many
courtesies extended to me by my former associates and other friends in EPA
laboratories, and by the Agency, have generated this special feeling of
affinity or fellowship, although I retired more than 11 years ago from the
federal government. To A. F. (Fritz) Bartsch, to Donald Mount, and to
Clarence Tarzwell (recently retired), I am particularly indebted in this
connection.
Because I have some highly critical remarks to make today about one
particular EPA publication, I want to make it very clear that I have great
respect for my many competent and dedicated colleagues in EPA and for their
notable research accomplishments. In no way can I hold them responsible
for the defects of the report in question, and I wish to fault nobody except
its anonymous authors in the Criteria and Standards Division, Office of Water
Planning and Standards. I well realize that in our overgrown federal
bureaucracy, monster agencies such as EPA can be many-headed like Hydra,
with one head often not knowing what another one knows, does, or thinks, and
not bothering to ask or to listen carefully. I am sure that some of my
friends in EPA are or will be as unhappy as I am with some of their
191
-------
organization's products, the quality of which they had no power to control.
They may welcome my saying more emphatically than they would want to say what
they too have been thinking.
What have been my thoughts concerning my career as the time of my
retirement approached? Naturally, I wish that my contributions to water
pollution biology and environmental protection were as important and influ-
ential as some of my friends have tried to assure me they have been. Long
ago, I believed that they would be. I started out as a smallish frog in a
little pond disdained and shunned by smarter frogs. It was the early 40's,
when water pollution control was primitive and my colleagues who were making
significant contributions to water pollution biology in the United States
could be counted on the fingers of my hands, or even on one hand. One did
not have to be great to be one of the top frogs in my unattractive puddle.
My early efforts to refine and standardize toxicity bioassays and to promote
their use in waste disposal control seemed well worthwhile and were soon
widely approved. Although I did little more than expedite inevitable
developments, the widespread adoption of the recommended bioassay methods in
this country and abroad was gratifying. My critical review of much of the
limited available literature in the field and my performance of a few simple,
carefully designed experiments soon made me an unchallenged expert. I moved
from Cincinnati to Gorvallis in 1953 at Professor R. E. Dimick's invitation.
I was to develop, with Charles Warren and others, an OSU-PHS cooperative
research program. As our joint research facilities and staff grew and
improved rapidly, the opportunities to make important contributions seemed
greatly enhanced. The need for a more aggressive attack on water pollution
was evidently being recognized. I thought that a rational plan of develop-
ment of our pertinent—although admittedly still very limited—ecological,
chemical, and toxicological knowledge, and an equally rational system of its
regulatory application would soon be designed and agreed upon by those in
charge of the effort. I was eager and ready to be one of those leading the
way, proud of our expanding laboratory complex here, which became a little
Mecca for the still small number of water pollution biologists. But then
came the flood, the unprecedented rapid expansion in the middle to late 60's,
of environmental protection activities in our country. I had become a bigger
frog in a pond somewhat enlarged by some busy beavers, but my pond now
suddenly became a large lake, whose often turbulent waters were soon invaded
by frogs coming from many other pools with all kinds of conflicting opinions.
My influence there consequently waned; it is now almost negligible, in spite
of my continued, sometimes frantic activity.
Impressed with signs of my apparent success and importance, such as the
extent of my travel and the size of my consulting fees, in recent years my
late brother Michael, the distinguished microbiologist, was no longer calling
me a "sewage worker." (This appellation he had gleefully assigned to me long
ago when he found me perusing the Sewage Works Journal, an early predecessor
of the Journal of the Water Pollution Control Federation.) Environmental
protection became a well-respected, well-funded, enthusiastically acclaimed
field of endeavor. However, I was not very pleased, for its too rapid, almost
chaotic development has not been conducive to careful discrimination between
fact and fancy, right and wrong, sense and nonsense. Now that my pretension
192
-------
of outstanding intellectual leadership can no longer be maintained, I am just
another frog contributing to a discordant chorus by croaking my discontent.
Now is a good time for me to retire completely from the fray. But, speaking
out on controversial issues in defense of rational positions, no matter how
futile it may be, is a habit difficult for me to break.
What future do I now see for aquatic toxicologists and aquatic biologists
in general in the field of water pollution control? I must say frankly that
I am not very optimistic. I see much bitter disappointment and frustration
for those competent, dedicated, and perceptive investigators who, like myself,
would like to see the results of their research promptly and intelligently
applied by the regulatory agencies. I see continued expenditure of much
talent, money, and effort on research of high quality that leads to no
visible, practical benefits, except perhaps, in the distant future. We can
hope, of course, that some day things will be different, the administration
of environmental protection laws will become entirely rational and truly
scientific, and incompetence, superficiality, and disregard for the elemen-
tary principles of logic in the application of our research results will no
longer be tolerated. Encouraged by this hope, or simply driven by intellec-
tual curiosity, many of my younger colleagues doubtless will continue to
exert their best efforts in seeking to advance knowledge in our field. But
the value of their most significant factual contributions and most pregnant
new ideas—even ideas that are not very profound or difficult to understand—
they should not expect to be soon recognized except by a small number of
colleagues also engaged in research. They should not assume that administra-
tive (regulatory) decisions on which these contributions and ideas obviously
have a direct bearing will be influenced and adjusted correctly to reflect
the new knowledge.
Why do I hold this pessimistic view? Well, let me give an example of
the kinds of frustration that I have recently experienced. My disappointment
was not unique, but it was somewhat more distressing and humiliating than most
of the others of its kind. And, it should be remembered that I am far from
being a beginner in my field; my views and contribution should not be quite
as easily ignored as those of numerous younger colleagues.
Last month, I examined a new publication just released by EPA {U.S.
Environmental Protection Agency 1976), a 510-page document entitled "Quality
Criteria for Water", a copy of which had kindly been supplied to me. Its
perusal in part left me quite shaken. The formulation of sound water quality
criteria pertaining to the protection of aquatic life and fisheries has been
my predominant interest or objective during most of the last 35 years. With
that end in view, I have done much thinking and have conducted intensive
experimental and literature research in the toxicology of the simple and
complex cyanides, the dissolved oxygen requirements of fishes, and other such
matters. Naturally, I want to know to what extent the water quality criteria
being proposed or used in water pollution control and the current regulatory
practices are being influenced by my efforts and recommendations. So, it was
with much interest that I began to examine the document presenting water
quality criteria now being recommended by EPA, that powerful government
agency charged with the administration of federal water pollution control
legislation.
193
-------
First, I looked at the section dealing with cyanides. As some of you
know, I have been able to demonstrate quite conclusively, with the invaluable
assistance of student and faculty colleagues, that the "total cyanide" con-
centration in water containing complex cyanides is toxicologically almost
meaningless (Doudoroff 1956; Doudoroff, Leduc, and Schneider 1966; Doudoroff
1976). The toxicity to bluegills, for example, of acutely toxic cyanide
solutions with total cyanide concentrations as low as 1 or 2 mg/1 or less is
determined entirely, or almost entirely, by the concentrations of free
cyanide or, more specifically, of molecular (un-ionized) hydrocyanic acid,
HCN. This relationship is usually true of the toxicity of much more concen-
trated solutions also, but there are known exceptions. At the pH of most
natural waters, most of the free cyanide (molecular HCN plus the CN" ion) is
present as HCN, the more toxic of the two forms of free cyanide (i.e., more
toxic than the CN" ion), therefore, the distinction between HCN and free
cyanide is of little practical importance. The level of one can be easily
calculated from that of the other when the pH is known. Undissociated
metallocyanide complex anions, which can be much more abundant than free
cyanide in cyanide-bearing wastes and polluted waters, are much less toxic
than HCN, or virtually nontoxic. For these reasons, it seemed obvious to
me that an entirely sound, basic, chemical water quality criterion pertaining
to the suitability of cyanide-polluted waters for aquatic life has to be
expressed as a concentration of free cyanide or of molecular HCN, and not of
total cyanide. A reliable and sensitive chemical analytical method that
distinguishes between the highly toxic and relatively harmless or toxicolog-
ically inactive forms of cyanide clearly was needed, I told my colleagues
long ago. Largely because of my early findings and urging, several quite
satisfactory methods for determination of molecular HCN have been developed
by my associates at Oregon State University (Schneider and Freund 1962;
Claeys and Freund 1968; Broderius 1973) and by other American and British
investigators (see Doudoroff 1976, pp. 9-10). Some of these methods were
used in confirming the toxicological conclusions stated above. Thus, through
intensive research, the technical problem to which I had addressed myself was
essentially solved, and I was very well pleased indeed with the accomplish-
ment, which seemed to call and point the way for much more research of the
same general kind.
But what did I find in the EPA report? I found that the great toxicity
of HCN is duly noted, as is the fact that the ratio of HCN to total cyanide
in waters polluted with cyanides is highly variable, depending not only on
the nature of the cyanide compounds introduced but also on the pH, illumina-
tion, and other conditions. In addition, I found this poorly worded but
nevertheless devastating statement (p. 132): "Since such chemical and physi-
cal conditions will dictate the form of cyanide, the cyanide criteria must be
based on the concentration of total cyanide present in the water" (emphasis
added). Accordingly, a cyanide concentration limit of 5 yg/1 (0.005 mg/1) is
recommended as a criterion for aquatic life without specifying that this
amount of cyanide must be free or present as molecular HCN.
Is the quoted conclusion a logical one? Apparently, the authors of the
report think that it obviously is; they make no effort to justify or defend
their assertion, although it flatly contradicts the published recommendation
of the National Academy of Sciences and National Academy of Engineering
194
-------
(1973), which I helped to prepare. Well, if that conclusion is accepted as
reasonable, then corresponding conclusions surely must be reached also with
respect to ammonia, sulfides, heavy metals, and other toxicants. It is well
known that the ratios of highly toxic molecular (un-ionized) ammonia and
hydrogen sulfide to total ammonia and total sulfide, respectively, in polluted
waters vary widely, depending on such factors as pH, temperature, and ionic
strength, and that their variation is toxicologically important. Thus, if
the EPA authors were at all consistent in applying the questionable reasoning
on which the statement quoted above is based, they should certainly have
concluded that, since chemical and physical conditions dictate the forms of
ammonia and sulfide, the ammonia and sulfide criteria must be based on the
concentrations of total ammonia and sulfide present in the water. But what
actually are the anmonia and sulfide criteria recommended by them? The
criterion for ammonia (p. 16) is 0.02 mg/1 of un-ionized ammonia only (not
total ammonia, for which no limit is proposed), and the sulfide criterion
(p. 410) is 2 ug/l of undissociated H^S only (not total sulfide or total dis-
solved sulfide). Evidently, the authors concluded that, since chemical and
physical conditions dictate the forms of ammonia and sulfide, the ammonia and
sulfide criteria must be based on the concentrations of molecular NH3 and H^S
only, disregarding the less harmful or relatively nontoxic NH4 and HS" ions.
What can be the reason for the obvious inconsistency? There can hardly
be any nice, logical justification. The only explanation that I can suggest,
other than sheer, negligent incompetence or dishonesty of the authors, is
that logic has gone out of style and consistency is no longer highly valued
in our field of environmental protection. Now, appeals to emotion and
prejudice prevail all too often over sound arguments, and a host of confused
"experts" have sprung up almost overnight like mushrooms. Immutable laws of
chemistry and physics dictate the transmutations of cyanide and ammonia, but
the choice of the water quality criteria evidently has been dictated only by
whim or caprice. Capriciously, the results of thorough, painstaking research
into the toxicology of the complex metallocyanides and careful development of
needed analytical methods that have made possible the establishment of sound
cyanide criteria like those previously developed for ammonia are totally
ignored—not even mentioned—in the EPA publication. They have been brushed
aside and made to seem irrelevant with a single, flat assertion that sounds
like a statement of an indisputable corollary of some natural law, but which
actually is groundless and contrary to reason. If this assertion were true,
there would be no good reason, of course, further to test or simplify the new
analytical methods for determination of HCN.
The possibility that a harmless form of cyanide present in water will be
soon converted, under certain conditions, into a highly toxic form should not
be overlooked in controlling water pollution. However, only after this trans-
mutation has actually occurred, a fact now readily demonstrable by chemical
analysis, is the suitability of the water as a medium for aquatic life af-
fected and it may or may not occur effectively. Photodecomposition of non-
toxic iron-cyanide complexes, for example, may be negligible in deep, turbid,
or shaded waters, and slowly liberated cyanide may decay or escape as rapidly
as it is released, free cyanide not being a presistent pollutant. A large
biochemical oxygen demand (BOD) of an effluent or receiving water is worthy
of attention, but the dissolved oxygen concentration (DO) is a much more
-------
meaningful Index of the suitability of polluted waters for aquatic life (ex-
cept for some decomposers} than is the BOD. When reaeration is rapid, an
Initially very large oxygen demand may be gradually satisfied without causing
any harmful depression of DO. It has long been generally recognized, there-
fore, that sound water quality criteria for the protection of aquatic life
against the oxygen-depleting effects of putrescible organic wastes must be
appropriate limits of DO and not of BOD. In what fundamental way is the
problem presented by the potential toxicity of nontoxic, complexed cyanide
different from that presented by the oxygen-depleting potential of organic
wastes? I can see no difference requiring diametrically different approaches
to the two problems.
Because of EPA's prestige and power, its ill-considered pronouncements
can block technical advances for years. Recently I have presented extensive
testimony in the State of Illinois in support of a proposal (by my clients,
the Illinois Petroleum Council, and others) that a free cyanide standard of
water quality be substituted for an outdated total cyanide standard that had
long been in force in that state. I hoped soon to see wide approval of such
improvement of standards by state regulatory agencies and I strove to bring
it about. But having seen the EPA report stating flatly that pertinent water
quality criteria "must be based on the concentration of total cyanide" and
implying that each recommendation contained in the report represents a con-
sensus or majority opinion of experts based on the latest available scientific
information, I now see almost no possibility of success. Although I do not
believe that such matters are best settled by the adversary method, I now
would like to see the issue litigated. Perhaps in a court of law, logic would
prevail. I hope that some of my influential, reasonable, and well-informed
friends in EPA will be willing and able to take some effective action leading
to early correction of the mistake.
In the section of the report on cyanide, I found other statements in
addition to the one quoted that are erroneous; some are incompatible (contra-
dictory). These errors are not of critical import, however, so they need not
be pointed out and discussed here. The treatment of the subject is generally
inadequate, and I think that attribution of the content of the entire volume
to "the efforts of many dedicated people" including "technical specialists
throughout the Agency's operational programs and in its research laboratories"
(p. ix) is not something that should greatly please competent members of the
EPA research staff.
After examining the section on cyanide, I turned to that dealing with
dissolved oxygen criteria—another subject of outstanding interest to me--
and found it no less depressing. There is no relation or resemblance at all
between the new EPA recommendations and the much more elaborate ones of
Doudoroff and Shumway (1970) or those of the National Academy of Sciences and
National Academy of Engineering (1973), which were based in large part on
those of Doudoroff and Shumway. Those recommendations have been ignored.
The DO criterion adopted by EPA is that proposed 40 years ago by Ellis (1937)
for warm-water fish habitats, simply a minimum of 5 mg/1. Its recommended
application has now been extended to all fresh waters, warm or cold, includ-
ing interstitial waters of the gravels of salmonid spawning beds. Applica-
bility of his criterion to cold-water fish habitats was not claimed by Ellis.
196
-------
The EPA criterion is said tfc be based primarily on observations made in the
field (mostly those of Ellis and his associates) on the relation between
observed DO levels in various sampled waters and the variety of fishes found
there; the presence of a "well-rounded fish population" was taken as an indi-
cation of satisfactory conditions.
The deficiencies of the evidence on which Ellis' conclusions were based,
that is, reasons for its unreliability, have been fully discussed by
Doudoroff and Shumway (1970, pp. 241-247). Their carefully developed
argument and the suppporting data, not mentioned by the EPA authors, can-
not be adequately summarized here. It was shown that good, mixed fish faunas,
as defined by Ellis, actually can occur 1n waters where DO levels do not ex-
ceed 4 mg/1 for very long periods, are often below 3 mg/1, and sometimes are
as low as 1.4 mg/1 or less. These results do not prove, of course, that fish
production is not seriously impaired at such low DO levels. Neither does
the observation, cited in the EPA report, that rainbow trout thrive in Lake
Titicaca, where, because of the altitude, DO does not exceed 5 mg/1, signify
that trout production is not reduced materially by reduction of DO to 5 rng/1
in other waters with much higher natural DO levels.
I was amused by the statement in the EPA report that, in seeking to
relate fish abundance and distribution to DO in the field, "enough observa-
tions have been made under a variety of conditions that the importance of
oxygen concentration seems clear." I cannot quarrel with that statement, but
is the mere demonstration of the importance of an environmental factor suf-
ficient for the establishment of a water quality criterion? The pertinent
experimental data, most of which have been thoroughly and critically reviewed
by Doudoroff and Shumway (1970), also show very clearly the importance of DO.
Why, then, has the vast amount of such information obtained during the past
40 years, in our laboratories and others, been mostly disregarded by the EPA
authors? Quite disturbing to me was this justification given by them of
their reliance predominantly or almost entirely on data from the field: "The
requirement that the data be applicable to naturally occurring populations
imposes limits on the types of research that can be used as a basis for the
criterion. Aside from a few papers on feeding, growth, and survival in
relation to oxygen concentration, very little of the laboratory based litera-
ture has a direct bearing; field data are in general more useful."
How many of the other water quality criteria, that have been recommended
in the same publication as defensible criteria pertaining to the requirements
of aquatic life (mostly criteria for toxic pollutants) are based predominantly
on field data? How many, I should ask, are based on any data other than data
from laboratory experiments? Not many, I am sure. What is the cyanide cri-
terion based on, for example? Only on laboratory data, and particularly on
observed effects of 10 yg/1 of free cyanide on the swimming performance of
salmonid fish. Actually, the vast amount of experimental (mostly laboratory)
data bearing on the DO requirements of fishes that is now available (data on
effects of DO reductions on survival, development, feeding, growth, fecundity,
swimming ability, behavior, respiration, and oxygen consumption) is a basis
for water quality criteria that is far more satisfactory than the bases for
most of the other recommended criteria. By contrast, the available data
from field studies on fish distribution and abundance (natural fish pop-
To?
-------
ulations) in relation to DO are still extremely limited, and their usefulness
in the verification or refinement of DO criteria is almost negligible. Again,
logic seems to have been abandoned. If the extensive data from laboratory
studies are indeed of almost no value or pertinence to the formulation of DO
criteria, does it not follow that there are no adequate bases at all for
most of the other water quality criteria pertaining to aquatic life that
have been advanced? Should not these other recommended criteria have been
withheld (not published) for lack of sufficient foundations?
I myself have been urging other investigators to pay more attention to
natural conditions and to their simulation (especially with regard to bio-
energetic considerations) in the design of experiments directed toward better
understanding of the effects of water pollution on aquatic life (Doudoroff
1977; Doudoroff and Shumway 1970). I know that fish, in their natural
habitats, are not usually exposed throughout their life cycles, or for very
long periods, to nearly constant concentrations of pollutants, or to unlim-
ited amounts of food obtainable almost without effort, or to an artificially
restricted food supply. I have repeatedly pointed out that interference with
reproduction in polluted waters of limited extent can be often fully compen-
sated for by increased growth rates (due to reduced competition for food) or
by the immigration of young from contiguous waters. I believe that some of
our water quality criteria based on results of unrealistic experiments may be
misleading, and some regulatory water quality standards directly derived from
them can be entirely too restrictive, particularly when the criteria derive
from life-cycle tests at constant concentrations of toxicants. But I certain-
ly would not go so far as to say that the experimental work of the past has
provided little useful information. I do not propose that we abandon our
laboratories and all take to the field to sample various polluted waters and
their fish populations in order to arrive at the best water quality criteria.
My impression is that, in the eyes of the authors of the EPA report, the
intensive experimental work on the DO requirements of fish and the chemistry
and toxicology of the complex cyanides that my co-workers and I have done
over the years has been almost completely wasted effort. Certainly, their
recommended water quality criteria would not have been any different had none
of this work ever been done. One may well be impelled to ask if it is not a
pity that so much time and money were spent so unproductively, because of my
poor judgment. And is not Gary Chapman of EPA, who spoke to us about the
different forms of copper and their relative toxicity, perhaps largely wast-
ing his time also when concerning himself with such matters? If water qual-
ity criteria for copper must, for some reason, be "based on" total copper, no
matter how successfully the toxic forms may be identified, their interactions
described, and analytical methods for their separate determination developed,
the subject of Chapman's report can be of academic interest only. Perhaps
he too should be out in the field collecting and identifying fish. Has
William Spoor also been wasting federal government money in Duluth by study-
ing effects of DO reduction on fish development?
I must say that I have not always felt that my efforts have been unap-
preciated or that my recommendations relative to water quality criteria have
been ignored. On the contrary, I have been often gratified by the attention
given to my findings and conclusions by my most respected professional
198
-------
colleagues, including leading EPA biologists. The honor accorded me at this
time clearly bespeaks abundant appreciation of my modest accomplishments. And
the authors of the important, recent publication "Water Quality Criteria
1972", the socalled "Blue Book", prepared for EPA by the National Academy of
Sciences and National Academy of Engineering (1973), having given me a cour-
teous and attentive hearing at no expense to me, accepted in large part
those of my views that were presented to them. As noted already, the cyanide
criteria recommended by that prestigious group are concentration limits of
free cyanide, not total cyanide. The DO criteria recommended, although not
entirely in agreement with the recommendations that I presented and defended,
did reflect my views in large degree, and I felt that their adoption was an
important step in the right direction. The adoption of graded criteria of
water quality appropriate to different "levels of protection" of aquatic
life (to be selected on the basis of socio-economic considerations), which
were recommended in dealing with pH and with suspended and settleable solids
as well as with DO, was most gratifying, because it had been first proposed
and strongly advocated by me.
Unfortunately, some important inconsistencies or illogical features
similar to those of the recommendations in the new EPA report mar also the
recommendations presented in the "Blue Book". At least one of the modifica-
tions made of the proposed DO criteria of Doudoroff and Shumway (1970) and
their related recommendations was not, in my opinion, justifiable; that
change, an incongruous kind of hybridization of old and new approaches,
clearly was adopted as a compromise because of reluctance of some of the
authors to depart entirely from precedents. Some serious errors and incon-
sistencies are to be expected in a work prepared in the manner and short
time in which the "Blue Book" was prepared. But it seems to me that in the
course of the preparation of the new EPA publication, on which work has been
going on for a long time, the inconsistencies and other mistakes to be found
in the somewhat too hastily prepared "Blue Book" should have been largely
corrected or avoided, not multiplied or aggravated.
The 1976 report is not the first such report prepared by EPA. This new
volume is a revision of proposed EPA Water Quality Criteria, presented in a
publication that was not widely distributed but whose limited availability
was announced by means of a notice published in October, 1973, in the Federal
Register (U. S. Environmental Protection Agency 1973a), It is noteworthy
that the cyanide criteria proposed by EPA in the earlier (1973a) report are
essentially identical with those recommended in the "Blue Book". The DO
criteria proposed at that time are somewhat different from the "Blue Book"
criteria, but were said in the Notice of Publication to be "generally consis-
tent" with them. I may have seen these proposed DO criteria but cannot now
recall examining them; a single DO level of 5 mg/1 was certainly not given as
a generally applicable water quality criterion. The disagreement between the
most recently published cyanide and DO criteria recommended by EPA and those
proposed in the "Blue Book" obviously are not attributable to inadvertence.
Why the criteria initially proposed by EPA have now been rejected and differ-
ent ones substituted, and who first proposed the drastic changes, I do not
know. In the 1976 report, it is stated that the revision of the previously
proposed criteria was "based on a consideration of comments received from
other federal agencies, state agencies, special interest groups, and individ-
-------
ual scientists." But it is not apparent that authors of the "Blue Book" and
other leading experts had an opportunity to review and comment on all the new
or revised criteria before publication, to object to proposed changes, and to
explain their objections. I understand that "pre-publication" copies of
"Quality Criteria for Water" were distributed in October or November of 1975
to a number of scientists or laboratories outside EPA for review. However,
I do not know how many of these copies were distributed or to whom they were
sent, and I have learned that the proposed DO criteria presented in those cop-
ies were still quite similar to the "Blue Book" criteria and those of
Doudoroff and Shumway. Thus, it seems reasonable to suppose that nobody of
the scientific community outside EPA was given the opportunity to examine
and object to the finally published DO criterion and supporting statement;
reviewers of the prepublication version had good reason to believe that the
"Blue Book" recommendations would not be entirely ignored or contradicted in
the published EPA report. I was never consulted nor asked my opinion of the
new cyanide and DO criteria by EPA, although my pertinent expertise could
hardly have been overlooked. Their publication was a complete surprise to
me, like a bolt from the blue.
It has been suggested to me that the real reasons for the drastic
revision of the original EPA criteria may perhaps have been political rather
than scientific, having something to do with possible difficulties of en-
forcement of regulatory standards based on them. The suggestion was that the
authors may have understood perfectly that the cyanide criteria can very well
be "based on" reliably determinate free cyanide or HCN levels and that
limits of free cyanide or HCN concentration are scientifically much sounder,
more reliable criteria than limits of total cyanide concentration, but
decided that acknowledgment of these scientific facts would be politically
inexpedient or embarrassing. However, deliberate obfuscation or concealment
of the truth obviously would have been intellectually dishonest, and I do not
want to accuse anyone of intellectual dishonesty. The administrator of EPA
had been directed by Congress to publish "criteria for water quality accur-
rately reflecting the latest scientific knowledge" and not reflecting his
staff's latest notions of how science or truth can best be twisted to achieve
some practical objective. In preparing my critical comments, I assumed that
the authors of the EPA report strove to fulfill this charge (as they implied
they did) and so were not intentionally inconsistent and purposely misleading.
It is noteworthy that the authors of "toxic pollutant effluent standards"
proposed by EPA about three years ago (U. S. Environmental Protection Agency
1973b) were aware of the importance of the distinction between free and
complexed cyanide. My clients, the American Iron and Steel Institute, and
many others objected to those proposed standards for various reasons, among
which were terminological and methodological vagueness and errors. At a
hearing in Washington, D.C., in 1974, I expounded extensively on the chemis-
try and toxicology of the cyanides, as did also my former student, Steven
Broderius, at a later hearing. I had hoped that our efforts to clarify the
complicated problems involved would lead to a better understanding by all
those in EPA concerned with effluent and water quality standards and criteria.
Because of the various objections raised, the proposed effluent standards,
which had some sensible features and could have been improved enough with a
few changes to make them fairly reasonable, were finally withdrawn by EPA.
200
-------
But the water quality criterion for cyanide now being recommended by EPA sug-
gests that understanding, if it has changed at all, has deteriorated. New
proposals concerning regulatory standards could well be totally wrong. I am
reminded again of the nature of Hydra, with which I have already drawn an
analogy. When you chopped off a head that threatened you, you were worse off
than before, because two more dangerous heads grew in its place.
I want to repeat, however, that my purpose here has not been to attack
EPA, an organization to which I still feel, justifiably or unjustifiably,
that I somehow belong. What I am really attacking is the shallow, careless,
and irresponsible thinking that pervades the environmental protection move-
ment. This irrationality is to be found outside EPA, in state regulatory
agencies for example, probably at least as often as in the powerful federal
agency; it is often to be found even in our universities, where we expect to
find models of detached rationality. I am objecting to all indifferent
tolerance in my profession of gross inconsistency, which betokens gross error,
for it can exist only when there is such error, I am croaking an appeal for
logic. If even old frogs like me refrain from raising their voices in pro-
test, for fear of offending some other frogs in our lake, who will? To whom
will the tadpoles in the lake be able to look for inspiring intellectual
guidance? At this stage of my career, I have nothing to lose by being out-
spoken, and I am sure that many of you, as well as others, no matter where
they work or seek support, will share my sentiments.
I thank you and wish you all a good year and successful researching
through 1977.
-------
LITERATURE CITED
Broderius, S. J. 1973. Determination of molecular hydrocyanic acid in water
and studies of the chemistry and toxicity to fish of metal-cyanide
complexes, Ph.D. thesis. Oregon State University, Corvallis. xvii +
287 pp.
Claeys, R. R., and H. Freund. 1968. Gas chromatographic separation of HCN
on Porapak Q--Analysis of trace aqueous solutions. Environ. Sci.
Techno!. 2(6): 458-460.
Doudoroff, P. 1956. Some experiments on the toxicity of complex cyanides to
fish. Sewage Ind. Wastes 28(8): 1020-1040.
Doudoroff, P. 1976. Toxicity to fish of cyanides and related compounds—A
review. Ecol. Res. Ser. EPA-600/3-76-038. U. S. Environmental Protec-
tion Agency, Duluth, Minn, vi + 155 pp.
Doudoroff, P. 1977. Keynote address—Reflections on pickle-jar ecology.
Pp. 3-19 jji J. Cairns, Jr., K. L. Dickson, and G. F. Westlake (eds.),
Biological monitoring of water and effluent quality. Pub. STP 607.
American Society for Testing and Materials, Philadelphia.
Doudoroff, P., G. Leduc, and C. R. Schneider. 1966. Acute toxicity to fish
of solutions containing complex metal cyanides, in relation to concentra-
tions of molecular hydrocyanic acid. Trans. Am. Fish. Soc. 95(1): 6-22.
Doudoroff, P., and D. L. Shumway. 1970. Dissolved oxygen requirements of
freshwater fishes. FAO Fish. Tech. Paper 86, FIRI/T86. Food and
Agriculture Organization of the United Nations, Rome, xi + 291 pp.
Ellis, M. M. 1937. Detection and measurement of stream pollution. (U. S.
Dept. of Comm., Bur. Fish. Bull. 22) Bull. Bur. Fish. 48: 365-437.
National Academy of Sciences and National Academy of Engineering. 1973.
Water quality criteria 1972. A report of the Committee on Water Quality
Criteria, Environmental Studies Board. Ecol, Res. Ser. EPA-R3-73-033.
U. S. Environmental Protection Agency, Washington, D. C. xix + 594 pp.
Schneider, C. R., and H. Freund. 1962. Determination of low level hydro-
cyanic acid in solution using gas-liquid chromatography. Anal. Chem.
34: 69-74.
U. S. Environmental Protection Agency. 1973a. Water quality criteria--
Notice of publication. Federal Register 38(206): 29646-29647.
U. S. Environmental Protection Agency. 1973b. Proposed toxic pollutant
effluent standards. Federal Register 38 (247): 35388-35395.
U. S. Environmental Protection Agency. 1976. Quality criteria for water.
EPA-440/9-76-023. U. S. Environmental Protection Agency, Washington,
D. C. ix + 501 pp.
202
-------
Dr. Charles E. Warren presented a paper on "The Interpretation of
Laboratory Results." The manuscript was not available at the time of
printing. Exclusion is not meant to imply any criticism of the paper
or the presentation.
203
-------
TECHNICAL REPORT DATA
(Phase read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/3-77-085
3. REPORT
4. TITLE ANDSUBTITLE
Recent Advances in Fish Toxicology—A Symposium
5. REPORTTTATE
July 1977
IS. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
j Richard A.
Tubb, editor
Oregon State University!
8. PERFORMING ORGANIZATION REPORT NO
Corvallis
9. PERFORMING ORGfiNIZATIQN NAME AN-P ADD
Department of Fisheries "
and Wildlife
Oregon State University
Corvallis, Oregon 97331
Lorvains Env. Research Lab.
Environmental Protection Agy
200 SW 35th St.
Corvallis, Oregon 97330
10. PROGRAM ELEMENT NO.
1BA608
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Research Laboratory—Corvallis
Office of Research and Development, EPA
200 SW 35th St.
Corvallis, Oregon 97330
13. TYPE OF REPORT AND PERIOD COVERED
proceedings — inhouse
14. SPONSORING AGENCY CODE
EPA-600-02
15. SUPPLEMENTARY NOTES
v ABSTRACT
The papers contained in this report were presented at the symposium—Recent Advance
in Fish Toxicology—held in Corvallis, Oregon on January 13-14, 1977. The Corvallis
Environmental Research Laboratory, U.S. Environmental Protection Agency and the
Oregon State University Deoartment of Fisheries and Wildlife cosponsored the sym-
posium to encourage the rapid communication of recent findings among fish toxicolo-
gists. The symposium was dedicated to Professor Peter Doudoroff on his retirement
from a long and active research and teaching career.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/GlOUp
Fish Toxicology
Water Quality
Aquatic Biology
3. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21.
------- |