United States
?
Office of Water
Environmental Protection
?
Regulations and Standards
Agency?
Criterio and
aandarcts Division
Washington, DC 20460
EPA 44'
April 1986
Water
Dissolved Oxygen
Ambient Aquatic Life Water Quality
Criteria for Dissolved Oxygen.
,( Freshwater)
U.S. Environmental Protection Agency
Office of
Research and Development,
Environmental
Research Laboratories
Duluth, Minnesota.
Narragansett, Rhode Island
NOTICES
This document has been reviewed by the Criteria and Standards Division,
Office of Water Regulations and Standards, U.S. Environmental Protection
Agency, and approved for publication.
Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
This document is available to the public through the National Technical
Information Service (NTIS), 5285 Port Royal Road, Springfield, Virginia 22161.
ii
FOREWORD
Section 304(a)(1) of the Clean Water Act of 1977 (Pt. 95-217) requires the
Administrator of the Environmental Protection Agency to publish water quality
criteria that accurately reflect the latest scientific knowledge on the kind
and extent of all identifiable effects on health and welfare that might be
expected from the presence of pollutants in any body of water, including
groundwater._ This document is a revision of proposed criteria based upon a
consideration of comments received from other Federal agencies, State
agencies, special interest groups, and individual scientists. Criteria
contained in this document replace any previously published EPA aquatic life
criteria for the same pollutant(s).
The term "water quality criteria" is used in two sections of the Clean
Water Act, Section 304(a)(1) and Section 303(c)(2). This term has a different
program impact in each section. In Section 304, the term represents a non-
regulatory, scientific assessment of ecological effects. Criteria presented
in this document are such scientific assessments. If water quality criteria
associated with specific stream uses are adopted by a State as water quality
standards under Section 303, they become enforceable maximum acceptable
pollutant concentrations in ambient waters within that State. Water quality
criteria adopted in State'water quality standards could have the same numer-
ical values as criteria developed under Section 304.
?
However, in many
situations States might want to adjust water quality
.
criteria developed under
Section 304 to reflect local environmental conditions and human exposure
patterns before incorporation into water quality standards. It is not until
their adoption as part of State water quality standards that criteria become
regulatory.
Guidelines to. assist States in the modification of criteria presented in
this document, in the development of water quality standards, and in other
water-related programs of this agency, have been developed by EPA.
William A. Whittington
Director
Office of Water Regulations and Standards •
iii'
ACKNOWLEDGEMENTS
Gary Chapman
Author
Environmental Research Laboratory
Narragansett, Rhode Island
Clerical Support: Nancy Lanpheare
iv.
CONTENTS
Page
AcknowledgementsForeword
??
??
0
iv
Tables??
vi
Figures
?
vii
Introduction??
1
Salmonids
?
4
Physiology ?
4
Acute Lethal Concentrations
??
5
Growth
?
5
Reproduction ?
8
Early Life Stages ?
0
8
Behavior ?
10
Swimming ?
11
Field Studies
??
11
Non-Salmonids
?
12
Physiology ?
12
Acute Lethal Concentrations
??
12
Growth
?
13.
Reproduction
?
17
Early Life Stages??
17
Behavior ?
18
Swimming ?
19
Field Studies??
19
Invertebrates ?
20
Other Consideration ?
23
Effects of Fluctuations
23
Temperature and Chemical Stress
??
25
Disease Stress ?
26
Conclusions ?
0
27
National -Criterion??
33
References ? 0
39
TABLES
Page
1.
Percent reproduction in growth rate of salmonids at various dissolved
oxygen concentrations expressed as the median value from n tests with
2.
each
Influence
species
of
?
temperature on growth rate of chinook salmon held 'at
3.
"Influence
various dissolved
of temperature
oxygen concentrationson
growth rate
?
of coho salmon
4.
various
Percent
dissolved
reduction
oxygen
in growth
concentrationsrate
of some
?
nOnsalmonid fish held at
various dissolved oxygen concentrations
expressed as the median value
from n tests with each species. ?
15
5. Effects of temperature on the percent reduction in growth rate of
largemouth bass exposed to various dissolved oxygen concentrations in
ponds
?
?
16
6.
Acutely lethal concentrations.of dissolved oxygen to aquatic insects
?
22
7. Survival of rainbow trout embryos as- a function of intergravel
dissolved oxygen concentrations and water velocity as compared
to
dissolved oxygen concentrations established
as
criteria or estimated
as producting various levels of production impairment'
?
32
Water quality criteria for ambient dissolved oxygen concentrations..
?
34
9.?
Sample calculations for determining daily means and
.
7-day mean
dissolved oxygen concentrations (30-day averages are calculated in"a
similar fashion using 30-day data). ?
35,
vi
held
7
FIGURES
Page
1.
?
?Effect
of continuous exposure to various mean dissolved oxygen
concentrations on survival of embryos and larval stages of-eight
-species of nonsalmonid fish
?
14
vii
Ambient Water Quality Criteria for Dissolved Oxygen
FRESHWATER AQUATIC LIFE
I.?
Introduction
A sizable body of literature on the oxygen requirements of. freshwater
aquatic life has. been thoroughly_ summarized (Doudoroff and Shumway, 1967,
1970;'Warren et al., .1973; 'Davis, 1975a,b; and Alabaster and Lloyd, 1980).
These reviews and other documents describing .the dissolved oxygen requirements
of aquatic organisms (U.S. Environmental Protection Agency, 1976; -Inter-
national Joint Commission, 1976; Minnesota.Pollution Control
.
Agency, 1980) and
more recent data were considered in the preparation of this document. The
references cited below are limited to- those considered to be the most defin-
itive and most representative of the preponderance of scientific evidence
concerning the dissolved oxygen requirements of freshwater organisms. The
guidelines used in
deriving aquatic life criteria for toxicants (Federal
Register, 45 FR 79318, November 28, 1980) are not applicable because-of"the
different' nature of the data bases. Chemical toxicity data bases rely on
standard 96-h LC50 tests and standard chronic tests;-there are very few data
of either type onAissolved oxygen.
Over the last
.
10 years the dissolved oxygen criteria proposed by various
agencies and researchers have generally reflected two basic. schools of
thought. One maintained that a dynamic approach should be.used so that the
criteria would vary with natural ambient dissolved oxygen minima in the waters
of concern (Doudoroff and Shumway, 1970) or with dissolved oxygen'requirements
of fish expressed in terms of percent saturation (Davis, I975a,b). The other
maintained. that, while not ideal, a single minimum allowable concentration
should adequately protect-the diversity of aquatic life 'in fresh waters (U.S.
Environmental 'Protectfon Agency, 1976). Both approaches relied on a simple
minimum allowable dissolved oxygen concentration as the basis for their
criteria. A 'simple minimum dissolved oxygen concentration was also the most
practicable approach in waste load allocation models of the time.
Expressing the criteria in terms of the actual amount of dissolved oxygen
available to aquatic organisms in Milligrams per liter (mg/1) is considered
more direct and easier to administer compared to expressing the criteria in
terms of percent saturation. Dissolved oxygen criteria expressed as percent
saturation, such as discussed by Davis (1975a,b),'are more complex and could
often result in unnecessarily stringent criteria in the cold months and
potentially unprotective criteria during periods of high. ambient temperature
or at high elevations. Oxygen partial pressure is subject to the same
temperature problems as percent saturation.
The approach recommended by Doudoroff and Shumway (1970), in which the
criteria vary seasonally with the natural minimum dissolved oxygen concentra-
tions in the waters of concern, was adopted by the National Academy of
Sciences and National Academy of Engineering (NAS/NAE, 1973). This approach
has some merit, but the lack of data (natural minimum concentrations) makes
its application difficult, and it can also produce unnecessarily stringent or
unprotective criteria during periods of extreme temperature.
The more simplistic approach to dissolved oxygen criteria has been
supported by the findings of a select committee of scientists specifically
established by the Research Advisory Board of the International Joint
Commission to review the dissolved oxygen criterion for the Great Lakes
(Magnuson et al., 1979). The committee concluded that a simple criterion (an
average criterion of 6.5 mg/1 and a minimum criterion of 5.5 mg/1) was
preferable to one based on percent saturation (or oxygen partial pressure) and.
was scientifically sound because the rate of oxygen transfer across fish gills
is directly dependent on the mean difference in oxygen partial pressure across
the gill. Also, the total amount of oxygen delivered to the gills is
a more
specific limiting factor than is oxygen partial pressure
per
se. The format
of this otherwise simple criterion
,
was more sophisticated than earlier
criteria with the introduction of a two-concentration criterion comprised of
both a mean and
a
minimum. This 'two-concentration criteria structure is
similar to that currently used for toxicants (Federal Register, 45 FR 79318,
November 28, 1980). EPA agrees with the International Joint Commission's
conclusions'and will recommend a two-number criterion for dissolved oxygen.
The national criteria presented herein represent the best estimates,
based on the data available, of dissolved oxygen concentrations necessary to
protect aquatic life and its uses. Previous water quality criteria have
either emphasized (Federal Water Pollution Control Administration, 1968) or
rejected (National Academy of Sciences and National Academy of Engineering,
1972) separate dissolved
.
oxygen criteria for coldwater and warmwater biota. A
warmwater-coldwater dichotomy is made in this criterion. To simplify discus-
sion, however, the text, of the document 'is split into
. salmonid and non-
salmonid sections. The salmonid-nonsalmonid dichotomy is predicated on the
much greater knowledge regarding the dissolved oxygen
.
requirements of
salmonids and on the critical influence of intergravel dissolved oxygen
concentration on salmonid embryonic and larval development. Nonsalmonid fish
include many other coldwater and coolwater fish plus all warmwater fish. Some
of these species are known to be less sensitive_ than salmonids to low dis-
solved oxygen concentrations. Some other nonsalmonids
. may prove to be at
least as sensitive to low dissolved oxygen concentrations as the salmonids;
among the nonsalmonids of likely sensitivity are the herrings (Clupeidae), the
smelts (Osmeridae), the pikes (Esocidae), and the sculpins (Cottidae).
Although there is little published data regarding the dissolved oxygen
requirements of most nonsalmonid species, there is apparently enough anecdotal
information to suggest that many coolwater species are more sensitive to
dissolved oxygen depletion than are warmwater species. According to the
American Fisheries Society (1978), the term "coolwater fishes"• is not vigor-
ously defined, but it refers generally 'to those species which are distributed
by temperature preference between the "coldwater" salmonid communities to the
north and the more diverse, often centrarchid-dominated "warmwater" assem-
2
blages to the south. Many states have pore stringent dissolved oxygen
standards for colder waters, watef
.
s" that contain either salmonids, nonsalmonid
coolwater fish, or the sensitive centrarchid, the smallmouth bass.
The research and sociological emphasis for dissolved oxygen has been
biased towards fish, especially the more economically important species in the
family Salmonidae. Several authors (Doudoroff and Shumway, 1970; Davis,
1975a,b) have discussed
this bias in considerable detail and have drawn
similar conclusions regarding the effects of low
,
dissolved oxygen on fresh-
water invertebrates. Doudoroff and Shumway (1970) stated that although some
'invertebrate species are about as sensitive as the
.
moderately susceptible
fishes, all invertebrate species need not be protected in order to protect the
food source for fisheries because many invertebrate species, inherently more
tolerant than fish, would increase in abundance.
?
Davis (1975a,b) also
concluded that invertebrate species would probably be adequately protected if
the fish populations are protected.
?
He stated that the composition of
invertebrate communities may shift to more tolerant forms selected from the
resident community or recruited from outside the community. In general,
stream invertebrates that are requisite riffle-dwellers probably have a higher
dissolved oxygen requirement than other aquatic invertebrates. The riffle
habitat maximizes .the potential dissolved oxygen flux to organisms living in
the high. water velocity by rapidly replacing the water in the immediate
micinity.of the organisms. This may be especially important for organisms
that.exist clinging to submerged substrate in the riffles. In the absence of
data to the contrary, EPA will follow the assumption that a dissolved oxygen
criterion protective of fish will be adequate.
One of the most' difficult problems faced during this attempt to gather,
interpret, assimilate, and generalize the scientific data base for
dissolved
oxygen effects on fish has been the variability in test conditions used by.
investigators. Some toxicological methods for measuring the effects of
chemicals on aquatic life have been standardized for nearly 40 years; this has
not been true of dissolved oxygen research. Acute lethality tests with
dissolved oxygen vary in the extreme with respect to types of exposure
(constant vs. declining), duration of exposure (a few hours vs. a week or
more), type of endpoint (death vs. loss of equilibrium), type of oxygen.
control (nitrogen stripping vs. vacuum degassing), and type of exposure
chamber (open to the atmosphere vs. sealed). In addition there are the normal
sources of variability that influence standardized toxicity tests, including
.seasonal differences in the condition of test fish, acclimation or lack of
acclimation to test conditions, type and level of feeding, test temperature,
age of test fish, and stresses due to test conditions. Chronic toxicity tests
are typically of two types, full life cycle tests or early life stage tests.
These have come to be rather rigorously standardized and are essential to the
toxic chemical criteria established by EPA. These tests routinely are assumed
to include the most sensitive life stage, and the criteria then 'presume to
protect all life stages. With dissolved oxygen research, very few tests would
be considered legitimate chronic tests; either they fail to .include a full
life cycle, they fail to include both embryo and larval stages, or they fail
to include 'an adequate period of post-larval feeding and growth.
3'
Instead of establishing year-round Criteria to protect all life stages,
it may be possible to establish seasonal criteria based on the life stages
present. Thus, special early life stage'criteria are routinely accepted for
salmonid early life stages because of their usual intergravel environment.
The same concept. may be extended to any species that appear to have more
stringent dissolved oxygen requirements during one period of their life
history. The flexibility afforded by such a dichotomy in criteria carries
with it the responsibility•to accurately determine the presence or absence of
the more sensitive stages prior to invocation of the less stringent criteria.
Such presence/absence data must be more site-specific than national in scope,
so that temperature, habitat, or calendar specifications are not possible in
this document. In the absence of such site-specific determinations the
default criteria would be those that would protect all life stages year-round;
this is consistent with the present format:for toxic chemical criteria.
II. Salmonids
The effects of various dissolved oxygen concentrations on the well-being
of aquatid organisms have been studied more extensively for fish of the family
Salmonidae (which includes the genera Coregonus, Oncorhynchus, Prosopium,
Salmo, Salvelinus, Stenodus, and Thymallus) than for any other family of
organisms. Nearly all these studies have 'been conducted under laboratory
conditions, simplifying cause and effect analysis, but minimizing or eliminat-
ing potentially important environmental factors, such as physical and chemical
stresses associated with suboptimal water- quality, as well as competition,
behavior, and other related activities. *Most laboratory studies on the
effects of dissolved oxygen concentrations on salmonids have emphasized
growth, physiology,
-
or embryonic development. Other studies have described
acute lethality or the effects of dissolved oxygen concentration on swimming
performance.
?
'
A.?
Physiology
Many studies have reported a wide variety of physiological responses to
low dissolved oxygen concentrations. Usually, these investigations were of
short duration, measuring cardiovascular and metabolic alterations resulting
from hypoxic exposures of relatively rapid onset. While these data provide
only minimal guidance for establishing environmentally acceptable dissolved
oxygen concentrations, .they do provide considerable insight into the mechan-
isms responsible for the overall effects observed in the entire organism. For
example, a good correlation exists between oxygen dissociation curves for
rainbow trout blood (Cameron, 1971) and curves depicting the reduction in
growth .of salmonids (Brett and Blackburn, 1981; Warren et al., 1973) and the
reduction in swimming ability of salmonids (Davis et al., 1963). These
correlations indicate that the blood's reduced oxygen loading capacity at
lower dissolved oxygen concentrations limits the amount of oxygen delivered to
the tissues, restricting the ability of fish to maximize metabolic perform-.
ance.
In general, the significance of metabolic and physiological studies on
the establishment of dissolved oxygen criteria must be indirect, because their
applicability to environmentally acceptable dissolved oxygen concentrations
requires greater extrapolation and more assumptions than those required for
data on growth, swimming, and survival.
4
Z.?
Acute Lethal ConcentratiOns•
Doudoroff and Shumway (1970) summarized studies on lethal concentrations
of dissolved oxygen for salmonids; analysis of these data indicates that the
test procedures were highly variable, differing
in
duration, exposure regime,
and reported endpoints. Only in a few cases could a 96-hr LC50 be calculated.
Mortality or loss of equilibrium usually occurred at concentrations between 1
and 3 mg/•.
Mortality of brook trout has occurred
in less than one hour at 10°C at
dissolved oxygen concentrations below 1.2 mg/1, and no fish survived exposure
at or below 1.5 mg/1 for 10 hours (Shepard, 1955). Lethal dissolved oxygen
concentrations increase at higher water temperatures and longer exposures. A
3.5 hr exposure killed all trout at 1.1 and 1.6 mg/1 at 10 and 20°C, respec-
tively (Downing and Merkens, 1957).
A 3.5-day exposure killed all trout at
1.3 and 2.4 mg/1 at 10 and 20°C, respectively. The corresponding no-mortality
levels were 1.9 and 2.7 mg/l. The difference between dissolved oxygen
concentrations causing total mortality and those allowing complete survival
was about 0.5 mg/1 when exposure duration was less than one week. If the
period of exposure to low dissolved oxygen concentrations is limited to less
than 3.5 days, concentrations of dissolved oxygen of 3 mg/1 or higher should
produce no direct mortality of salmonids.
More recent studies confirm these lethal levels in chronic tests with
early life stages of salmonids (Siefert et al., 1974; Siefert and Spoor, 1973;
Brooke and Colby, 1980); although studies with lake trout (Carlson and
Siefert, 1974) indicate that 4.5 mg/1 is lethal at 10°C (perhaps a marginally
acceptable temperature for embryonic lake trout).
C.?
Growth
Growth of salmonids is most susceptible to the effects of low dissolved
oxygen concentrations when-the metabolic demands or opportunities are great-.
est. This is demonstrated by the greater sensitivity of growth to low
dissolved oxygen concentrations- when temperatures are high and food most
plentiful (Warren et al., 1973).
A total of more than 30 growth tests have
been reported by Herrmann et al. (1962), Fisher (1963), Warren et al. (1973),
Brett and Blackburn (1981), and Spoor (1981). Results of these tests-are not
easily compared because the tests encompass a wide range of species, tempera-
tures, food types, and fish sizes. .These factors produced a variety of
control growth rates which, when combined with a wide range of test durations.
and fish numbers, resulted in an array of statistically diverse test results.
The results from most of these 30-plus tests were converted to growth
rate data for fish exposed to' low dissolved oxygen concentrations and were
compared to control growth rates by curve-fitting procedures (3R8 Associates;
1984). Estimates of growth rate reductions were similar regardless of the
type of curve employed, but the quadratic model was judged to be superior and
was used in the growth rate analyses contained in this document. . The apparent
relative sensitivity of each species to. dissolved oxygen depletion may be
influenced by fish size, test duration, temperature, and diet. Growth rate
data (Table 1) from these tests with salmon and trout fed unrestricted rations
indicated median growth rate reductions of 7, 14, and 25 percent for fish held
5
at 6, 5; and 4 mg/1, respectively (JRB Associates, 1984). However,
median
growth rate reductions for the various species ranged from 4 to 9 percent at 6
mg/1, 11 to 17 percent at 5 mg/1, and 21 to 2
.9 percent at .4 mg/l.
Table 1. Percent reduction in growth rate of salmonids at various dissolved
oxygen concentrations expressed as the median value from n tests
with each species (calculated from JRB Associates, 1984).
Species (number of tests)
Dissolved
Oxygen
?
?
Chinook
?
Coho?Sockeye
?
Rainbow
?
Brown?
Lake
(mg/1)?
Salmon (6) Salmon (12) Salmon (1) Trout (2) Trout (1) Trout (2)
9
?
0?
0?
0?
0?
0.
? 0
8? 0?
0
? 0?
1
?
0
?
0
7
?
1
?
1?
2? 5? 1? • 2
6?
7
?
4
?
6
?
9 ?
6? 7
5? 16
?
11?
12? 17? 13? 16
4?
29
?
21?
22? 25
? 23
?
29
3? 47? 37?
33
?
37? 36? 47
Median
Temp. (°C)
?
15
18
?
15?
12
?
12-
?
12
Considering the variability inherent in growth studies, the apparent
reductions in growth rate sometimes seen above 6 mg/1 are not usually statist-
ically significant. The reductions in growth rate occurring at:dissolved
oxygen concentrations below about 4 mg/1 should be considered severe; between
4 mg/1 and the threshold of effect, which variably appears to be.between•6 and
10 mg/1 in individual tests, the effect on growth rate is moderate to slight
if the exposures are sufficiently long.
Within the growth data presented by Warren et al. (1973), the greatest
effects and highest thresholds of effect occurred at high temperatures (17.8
to 21.7°C). In two tests. conducted at about 8.5°C, the growth rate reduction
at 4 mg/1 of dissolved oxygen averaged 12 percent. Thus, even at the maximum
feeding levels in these tests, dissolved oxygen levels down to 5 mg/1 probably
have little effect on growth rate at temperatures below 10°C.
Growth data from Warren et al. (1973) included chinook salmon tests
conducted at various temperatures. These data (Table 2) indicated that growth
tests conducted at 10-15°C would underestimate the effects of low dissolved
oxygen concentrations at higher temperatures by a significant margin. For
example, at 5 mg/1 growth was not affected at 13°C but was reduced by 34
percent if temperatures were as high as 20°C. Examination of the test
temperatures associated with the growth rate reductions listed in Table 1
shows that most data represent temperatures between 12 and 15°C. At the
higher temperatures often associated with low dissolved oxygen concentrations,
the growth rate reductions would have been greater if the generalizations of
6
the chinook salmon' data are applicable to salmonids in general. Coho salmon
growth studies (Warren et al., 1973) showed a similar result over a range of
temperatures from 9 to 18°C, but the trend was reversed in two tests near 22°C
(Table 3). Except for the 22°C coho tests, the coho and chinook.salmon
.results support the idea that effects of low dissolved oxygen become more
severe 'at higher temperatures. This' conclusion is supported by data on
largemouth bass (to be discussed later) and by the increase in metabolic rate
produced by high temperatures.
Table 2. Influence of temperature on' growth rate of chinook salmon held at
various dissolved oxygen concentrations (calculated from Warren et
al., 1973; JRB Associates, 1984).
Dissolved
?
Percent Reduction in Growth Rate at
Oxygen
.(mg/1)
?
8.4°C
?
13.0°C,
?
13.2°C?
17:8°C?
18'.6°C?
21.7°C
9
00000
0
8
000020
7
0
0
4
0.8
2
6
0
_O
8
5
19
14
5
0
0
16
16
34
34-
4
7
'4
25
33
53
65
3
26.
2g
36
57
77
'100
Table 3. Influence of temperature•on growth rate of coho salmon held at
various dissolved oxygen concentrations (calculated from Warren et
al., 1973; JRB Associates, 1984).
Dissolved
?
Percent Reduction in Growth Rate at
Oxygen?
.
(mg/1) .?
8:6°C?
12.9°C .?
13.0°C
?
18.0°C'?
21.6°C?
21.8°C .
10
0?
-
0
0
0
.0
0
9
0
?
.
0 0
5
0
0
8
.0
1
2
10
0
0
7
1
4
6
17.
.0
6
6
4
10
13
27
0.
1
5
9
18
' 23
38
0
7
4
17
29
36
51
4
19
3
.
?
28
42
51
67
6
37
•
Effects of dissolved oxygen concentration on the growth rate of salmonids
fed restricted rations have been less intensively investigated. Thatcher
(1974) conducted a series of tests with coho salmon at 15°C over a wide range
of food consumption rates at 3, 5, and 8 mg/1 of
.
dissolved oxygen. The only
significant reduction in
growth
.
rate was
observed at 3 mg/1 and food consump-
7
8
tion rates greater than about 70 percent of maximum... In these studies`,
Thatcher noted that fish at 5 mg/1 appeared to expend less energy in swimming
activity than those at 8 mg/l. In natural conditions, where fish may be
rewarded for energy expended defending preferred territory or searching for
food, a dissolved oxygen concentration of 5 .mg/1 may restrict these. activ-
ities.
The effect of forced activity and dissolved oxygen concentration on
the
growth of coho salmon was studied by Hutchins (1974). The growth rates
of
salmon fed to repletion at a dissolved oxygen concentration of 3 mg/1 and held
at current velocities of 8.5 and 20 cm/sec were reduced by 20 and 65 percent,
respectively. At 5 mg/1; no reduction of growth rate was seen at the slower
velocity, but a 15 percent decrease occurred at the higher velocity,
The effects of various dissolved oxygen concentrations on the growth rate'
of coho salmon
(ti 5 cm long) in laboratory streams with an average current
velocity of 12
cm/sec have been reported by Warren et al. (1973). In this
series of nine tests,• salmon consumed aquatic invertebrates living in the
streams. Results at temperatures from 9.5° to 15.5°C supported the results
of
earlier laboratory studies; at higher growth rates (40 to 50 mg/g/day),
dissolved oxygen levels below 5 mg/1 reduced growth rate, but at lower growth
rates (0 to 20 mg/g/day), no effects were seen at concentrations down to 3
mg/1
• The applicability of these growth data from laboratory tests depends on
the available food and required activity in natural situations. Obviously,
these factors will be highly variable depending on duration of exposure,
growth rate, species, habitat, season, and size of fish. However, unless
effects of these variables are examined for the site in question, the labora-
tory results should be used. The attainment of critical size is vital to the
smolting of anadromous salmonids and may be important for all salmonids
if
size-related transition to feeding on larger or more diverse food organisms is
an advantage. In the absence of more .definitive site-specific, species-
specific growth data, the data summary in Tables 1, 2, and 3 represent the
best estimates of the effects of dissolved oxygen concentration on the
potential growth of salmonid fish.
D. •
Reproduction
No studies were found that described the effects of low dissolved oxygen
on the reproduction, fertility, or fecundity of salmonid fish.
E.
Early Life Stages
Determining the dissolved oxygen requirements for salmonids, many of
which have embryonic and larval stages that develop while buried in the gravel
of streams and lakes, is complicated by complex relationships between the
dissolved oxygen supplies in the gravel and the overlying water. The dis-
solved oxygen supply of embryos and larvae can be depleted even when the
dissolved, oxygen concentration in the overlying body of water is otherwise
acceptable. Intergravel dissolved oxygen is dependent upon the balance
between the combined respiration of gravel-dwelling organisms, from bacteria
to fish
embryos, and the rate of dissolved oxygen supply, which is dependent
• upon rates
of
water percolation' and
,
convection, and dissolved oxygen dif-
fusion.
Water flow past salmonid eggs influences the dissolved oxygen supply to
the microenvironment surrounding each egg. Regardless of dissolved oxygen
concentration
.
in the gravel, flow rates below 100 cm/hr directly influence the'
oxygen supply' in
the microenvironment and hence the
size
at hatch of salmonid
fish. At dissolved oxygen levels below 6 mg/1 the time from fertilization to
hatch is longer as water flow decreases (Silver et al., 1963; Shumway et al.,
1964).
The
.
dissolved oxygen. requirements for growth of salmonid embryos and
larvae have not been shown to differ appreciably from those of older sal-
monids. Under conditions of adequate water flow
(.1100
cm/hr), the weight
attained by salmon and trout larvae prior to feeding (swimup) is decreased
less than 10 percent by continuous exposure to concentrations down
to
3 mg/1
(Brannon, 1965; Chapman and Shumway, 1978). The considerable developmental
delay which occurs at low dissolved
oxygen
conditions could have
survival and
growth implications if the time of emergence from gravel, or first feeding, is
critically related to the presence of specific food organisms, stream flow, or
other factors (Carlson and Siefert, 1974; Siefert and. Spoor, 1974). Effects
of low dissolved oxygen on early life stages are probably most significant
during later embryonic development when critical dissolved oxygen concentra-
tions are highest (Alderdice et al., 1958) and during the first few months
post-hatch when growth rates are usually highest. The latter authors studied
the effects
,
of 7-day exposure of
•
embryos to low dissolved oxygen at various
stages during incubation at otherwise high dissolved oxygen concentrations.
They found no effect of 7-day exposure at concentrations
above 2
mg/1 (at a
water flow of 85 cm/hr).
Embryos
of
mountain whitefish suffered severe mortality at a mean
dissolved oxygen concentration of 3.3 mg/1 (2.8 mg/1 minimum) and some
reduction in survival was noted at 4.6 mg/1 (3.8 mg/1 minimum); at 4.6 mg/1,
hatching was delayed by 1 to 2 weeks (Sieffert, et al., 1974). Delayed
hatching resulted in poorer growth at the end of the test, even at dissolved
oxygen concentrations of 6 mg/l.
Evaluating intergravel dissolved oxygen concentrations is
difficult
because of the great spatial and temporal variability produced by differences
in stream flow, bottom topography, and gravel composition. Even within the
same redd, dissolved oxygen concentrations can vary by 5 or 6 mg/1 at a given.
time (Koski, 1965). Over several months, Koski repeatedly measured the
dissolved oxygen concentrations in over 30 coho salmon redds and the overlying
stream water in three small, forested (unlogged) watersheds. The results of
these measurements indicated that the average intraredd dissolved oxygen
concentration was about 2
mg/1 below that
of the overlying water. The minimum
concentrations measured in the redds averaged about 3
mg/1 below
those of the
overlying water and probably occurred during the latter period of intergravel
development when water temperatures were warmer, larvae larger, and overlying
dissolved oxygen concentrations lower.
9
Coble (1961) buried steelhead trout eggs _in streambed gravel, monitored
nearby intergravel dissolved oxygen and 'water velocity, and noted embryo
survival. There was a positive correlation between dissolved oxygen concen-
tration, water velocity, and embryo survival. Survival ranged from 16 to 26
percent whenever mean intergravel dissolved oxygen concentrations were below 6
mg/1 or velocities were below 20 cm/hr; at dissolved oxygen concentrations
above 6 mg/i and velocities over 20 cm/hr, survival ranged from 36 to 6.2
percent. Mean reductions in dissolved oxygen concentration between stream and
intergravel waters averaged about 5 mg/1 as compared to the 2 mg/1 average
reduction observed
by
Koski (1965) in the same stream.
,
One explanation for
the different results is that the intergravel water flow may have been higher
in the natural redds studied by Koski (not determined) than in. the artificial
redds of Coble's investigation. Also, the density of eggs near the sampling
point may have been greater in Coble's simulated redds.
A study of dissolved oxygen concentrations in brook trout redds was
conducted in Pennsylvania
.
(Hollander,' 1981). Brook trout generally prefer
areas of groundwater upwelling for spawning sites (Witzel and MacCrimmon,
1983). Dissolved oxygen and temperature data offer no indication .
of ground-
water flow in Hollander's study areas, however, so that differences between
water column and intergravel dissolved oxygen concentrations probably repre-
sent intergravel dissolved oxygen depletion. Mean dissolved oxygen concentra-
tions in redds averaged 2.1, 2.8, and 3.7 mg/liter less than the surface water
in the three portions of the study. Considerable variation of intergravel
dissolved oxygen concentration was observed between redds and within a single
redd. Variation from one year to another suggested that dissolved oxygen
concentrations will show
,
greater intergravel depletion during years of low
water flow.
Until more data are available, the dissolved oxygen concentration in the.
intergravel environment should be considered to be at least 3 mg/1 lower than
the oxygen concentration in the overlying water. The 3 mg/1 differential is
assumed in the criteria, since it reasonably represents the only two available.
studies based on observations in natural redds (Koski, 1965; Hollender, 1981).
When siltation loads are high, such as in logged or agricultural watersheds,
lower water velocity within the gravel could additionally reduce dissolved
oxygen concentrations around the eggs.. If either greater .or lesser differen-
tials are known or expected, the criteria should be altered accordingly.
F.?
Behavior
Ability of chinook and coho salmon to detect and avoid abrupt differences
in dissolved oxygen concentrations was demonstrated
by
Whitmore et al. (1960).
In laboratory troughs, both species showed strong preference for oxygen levels
.of 9 mg/1 or higher over those near 1.5 mg/1; moderate selection against 3.0
mg/1 was common and selection against 4.5 and 6.0 mg/1 was sometimes detected.
The response of young Atlantic salmon and brown trout to low dissolved
oxygen depended on their age; larvae were apparently unable to detect and
avoid water of low dissolved oxygen concentration, but fry 6-16 weeks of age
showed a marked avoidance of concentrations up to 4 mg/1 (Bishai, 1962).
Older fry (26 weeks of age) showed avoidance of concentrations up to 3 mg/l.
10
In a recent study of the rainbow trout.sport fishery of Lake Taneycomo,
Missouri, Weithman and Haas (1984) have reported that reductions in minimum
daily dissolved oxygen concentrations below 6 mg/1 are related to a decrease
in the harvest rate of rainbow trout
.
from the lake. Their data suggest that
lowering the daily minimum from 6 mg/1 to 5, 4, and 3 mg/1 reduces the harvest
rate by 20, 40, and 60 percent, respectively. The authors hypothesized that
the reduced catch was a result of reduction in feeding activity. This mechan-
ism of action is consistent with Thatcher's (1974) observation of lower
activity of coho salmon at 5 mg/1 in laboratory growth studies and the finding
of Warren et al. (1973) that growth impairment produced by low dissolved
oxygen appears to be primarily a function of lower food intake.
A three-year study of a fishery on planted rainbow-trout was published by'
Heimer (1984). , This study found that the catch of planted trout increased
during periods of low dissolved oxygen in American Falls reservoir. on the
Snake River in Idaho. The author concluded that the fish avoided areas of low
dissolved oxygen and high temperature and the increased catch rate was a
result of the fish concentrating in areas of more suitable oxygen
,
supply and
temperature.
G.
Swimming
Effects of dissolved oxygen concentrations on swimming have been demon-
strated by Davis et al. (1963). In their studies, the maximum sustained
swimming speeds (in the range of 30 to 45 cm/sec) of juvenile coho salmon were
reduced by 8.4, 12.7, and 19.9 percent at dissolved oxygen concentrations of
6, 5, and 4
mg/1,
respectively. Over a. temperature range from 10 to
20°C,
effects were slightly more severe at cooler temperatures. Jones (1971)
reported 30 and 43 percent reductions of maximal swimming speed of rainbow
trout at dissolved oxygen concentrations of 5.1 (14°C) and 3.8
(22°C)'mg/1,
respectively. At lower swimming speeds (2 to 4 cm/sec), cbho and chinook
salmon at 20°C were generally able to swim for 24 hours at dissolved oxygen
concentrations of 3 mg/i and above (Katz et al., 1958). Thus, the signif-
icance of lower dissolved oxygen concentrations on swimming depends on the.
level of swimming performance required for the survival, growth, and reproduc-
tion of salmonids. Failure to escape from predation or to negotiate a swift
portion of a spawning migration route.may be considered an indirect lethal
effect and, in this regard, reductions of maximum swimming performance can be
very important.. With these exceptions, moderate levels of swimming activity
required by salmonids are apparently little affected by concentrations of
dissolved 'oxygen that are otherwise acceptable for growth and reproduction.
H.
Field Studies
Field studies of salmonid populations 'are' almost non-existent with
respect to effects of dissolved oxygen concentrations. Some •
of the systems
studied by Ellis (1937) contained 'trout, but of those river systems in which
trout or other salmonids mere most likely (Columbia River and Upper Missouri
River) no stations were reported with dissolved oxygen concentrations below 5
mg/1, and 90 percent of the values exceeded 7 mg/l.
11
III. Non-Salmonids
The amount of data describing effects of low dissolved oxygen on non-
salmonid fish is
more limited than that for salmonids, yet must cover a group
of fish with much greater taxonomic and physiological variability. Salmonid
criteria must provide for the protection and propagation of 38 species in 7
closely related genera; the non-salmonid criteria must provide for the protec-
tion and propagation of some 600 freshwater species in over 40 diverse
taxonomic families. Consequently, the need for subjective technical judgment
is greater for the non-salmonids.
Many of the recent, most pertinent data have been obtained for several
species of Centrarchidae (sunfish), northern pike, channel catfish, and the
fathead minnow. These data demonstrate that the larval stage is generally the
most sensitive life stage. Lethal effects on larvae have been observed at
dissolved oxygen concentrations that may only slightly affect growth of
juveniles of the same species.
A.?
Physiology
Several studies of the relationship between low dissolved oxygen concen-
trations and resting oxygen consumption rate constitute the bulk of the
physiological data relating to the effect of hypoxia on nonsalmonid fish. A
reduction in the resting metabolic rate of .fish is generally believed to
represent a marked decrease
in the scope for growth and activity, a net
decrease in the supply of oxygen to the tissues, and perhaps a partial shift
to anaerobic energy sources. The dissolved oxygen concentration at which
reduction in resting metabolic rate first appears is termed the critical
oxygen concentration.
.?
•
Studies with brown bullhead (Grigg, 1969), largemouth bass (Cech et al.,
1979), and goldfish.and carp (Beamish, 1964), produced estimates of critical
dissolved oxygen concentrations for these species. For largemouth bass, the
critical dissolved oxygen concentrations were 2.8 mg/1 at 30°C, < 2.6 mg/1 at
25°C, and < 2.3 mg/1 at 20°C. For brown bullheads the critical concentration
was about
4
mg/l. Carp displayed critical oxygen concentrations near 3.4 and
2.9 mg/1 'at 10 and 20°C, respectively, and goldfish critical concentrations of
dissolved oxygen were about 1.8 and 3.5 mg/1 at 10 and 20°C, respectively.' A
general summary of these data suggest critical dissolved oxygen concentrations
between 2 and 4 mg/1, with higher temperatures usually causing higher critical
concentrations.
Critical evaluation of the data of Beamish (1964) suggest that the first
sign of hypoxic stress is not the decrease in oxygen consumption, but rather
an increase, perhaps as a result of metabolic cost of passing an increased
ventilation volume over the gills. These increases were seen in carp at 5.8
mg/1 at 20°C and at 4.2 mg/1 at 10°C.
B.?
Acute Lethal Concentrations
Based on the sparse data base describing acute effects of low dissolved
oxygen concentrations on nonsalmonids, many non-salmonids appear to be
considerably less sensitive than salmonids.
?
Except
for
larval forms, no
12
?
•
non-salmonids appear to be more sensitive than salmonids. Spoor (1977)
observed lethality of largemouth bass larvae at a dissolved oxygen concentra-
tion of 2.5 mg/1 after only a 3-hr exposure. Generally, adults and juveniles
of all species studied survive for at least a few hours at concentrations of
dissolved oxygen as low as 3 mg/l. In most cases, no mortality results from
acute exposures to 3 mg/1 for the 24- to 96-h duration of the acute tests..
Some non-salffonid fish appear to be able to survive a several-day exposure to
concentrations below 1 mg/1 (Moss and Scott, 1961; Downing and Merkens, 1957),
but so little is known about the latent effects of such exposure that short-
term .
survival cannot now be used as an indication of 'acceptable dissolved.
oxygen concentrations. In addition to the unknown latent effects of exposure
to very low dissolved oxygen concentrations, there are no data on the effects
of repeated short-term exposures. Most importantly, data on the tolerance to
low dissolved oxygen concentrations are available for only a few of the
numerous species of non-salmonid fish.
C.?
Growth
Stewart et al. (1967) conducted several growth studies with juvenile -
largemouth bass and observed reduced growth at 5.9 mg/1 and lower concentra-
tions.?
Five of ,
six experiments included dissolved oxygen concentrations
between. 5 and 6 mg/1; dissolved oxygen concentrations'of 5.1 and 5.4 mg/1
produced reductions in growth-rate of 20 and 14 percent, respectively, but
concentrations of 5.8 and 5.9 mg/1 had essentially no effect on growth. The.
efficiency Of food conversion was not reduced until dissolved oxygen concen-
trations were much lower, indicating that decreased food consumption was the
primary cause of reduced growth.
When. channel catfish fingerlings held at 8, .5, and 3 mg/1 were fed as
much as they could eat
in
three daily feeding's, there Were significant
reductions in feeding and weight gain (22 percent) after a 6 week exposure.to
5 mg/1 (Andrews et al., 1973). At a lower feeding rate, growth after 14 weeks
was reduced only at 3 mg/l. Fish exposed to
.
3 mg/1 swam lethargically, fed
poorly and had reduced response to loud noises. Raible (1975) exposed channel
catfish to several dissolved oxygen concentrations for up to 177 days and
observed a graded reduction in growth at each concentration below 6 mg/l.
However, 'the growth pattern for 6.8 mg/1 was comparable to that at 5.4 mg/1.
He concluded that each mg/1 increase in dissolved oxygen concentrations
between 3 and 6 mg/1 increased growth by 10 to 13 percent.
Carlson et al. (1980) studied the effect of dissolved oxygen concentra-
tion on the growth of juvenile channel catfish and yellow perch. Over periods
of about 10 weeks, weight gain of channel catfish was lower than that of
control fish by 14, 39, and 54
s
percent at dissolved oxygen concentrations of
5.0, 3.4, and 2.1.mg/1, respectively. These differences were' produced by
decreases .in growth rate of 5, 18, and 23 percent (JRB Associates, 1984),
pointing out the importance of differentiating between effects on weight gain
and effects on growth rate. When of sufficient duration, •small reductions in
growth rate can have large effects on relative weight gain. Conversely, large
effects on growth rate may have little effect on annual weight gain if they
occur only over a small proportion of the annual growth period. Yellow perch
appeared 'to be more tolerant to low dissolved oxygen concentrations, with
reductions in weight gain of 2, 4, and 30 percent at dissolved oxygen concen-
trations
of 4.9, 3.5, and 2.1 mg/1, respectively.
13
••120
•
100 --
?
tP
cig0
?
Alp
cg]
O
m
V
0?
0
0
cyr fria
-
80
. •l
0
o 60
V
•
a)
0 Largemouth Bass
?
?
•?
_ 0
Black Crappie
CL
op
?
40 —
o
A White
White
SuckerBass
?
•
II
20 —?
M?
M
Channel
Northern
CatfishPike
A
Walleye
•
Smallrriouth Bass
0
—3,646-8:12-11-4
2
?
?
3?
4 5
I
6 7
f
8'910
I l l I
Dissolved Oxygen (mg /L)
Figure 1. Effect of continuous exposure to various mean dissolved oxygen
concentrations on survival of embryonic and larval stages of eight
species of nonsalmonid fish.
?
Minima recorded in these tests
averaged about 0.3 mg/1 below the mean concentrations.
14
Species (number of'tests)
Northern
Pike (1)
Largemouth
Bass (6)
'Channel
Catfish (1)
Yellow
Perch (1)
00
0
0
1
0
0
0•
4
0
?
-
1
0
-
?
9
.
0
3
0
16
1
7
0
• 25
9.
13
0
35 .
17
20
7
51
29
22
19
26
25
.20
Dissolved
Oxygen
(mg/1)
9
8
7
6
5
4
3
2
Median
Temp (°C)
The data of Stewart et al. (1967), Carlson et al. (1980), and Adelman and
Smith (1972) were analyzed to determine the relationship between growth rate
and'dissolved oxygen concentration (JRB .
Associates, 1984). Yellow 'perch
appeared to be very resistant to influences of low dissolved oxygen concentra-
tions, northern pike may be about as sensitive as salmonids, while largemouth
bass and channel catfish are" intermediate in their response (Table
4).
The
growth rate relations modeled from Adelman-and Smith are based on only four
data points, with none in the critical dissolved oxygen region from 3 to 5
mg/i. Nevertheless, these growth
. data for northern pike are the best avail-
able for nonsalmonid coldwater fish. Adelman and Smith observed about a 65
percent reduction in growth of juvenile northern pike after 6-7 weeks at
dissolved oxygen concentrations of 1.7 and 2.6 mg/i. At the next higher
concentration (5.4
mg/1),
growth was reduced 5 percent.
Table 4. Percent reduction in growth rate of some nonsalmonid fish held at
various dissolved oxygen concentrations expressed as the median
value from n tests with each species (calculated' from JRB
Associates, 1984).
Brake (1972) conducted a series of studies
On juvenile largemouth bass in
two artificial ponds to determine the effect of reduced dissolved oxygen
concentration on consumption of mosquitofish and growth during 10 2-week
exposures. The dissolved oxygen in the control pond was maintained near
air-saturation (8.3 to 10.4 mg/1) and the other pond contained mean dissolved
oxygen concentrations from 4.0 to 6.0 mg/1 depending upon the individual test.
The temperature, held near the same level in both ponds for each test, ranged
from 13 to 27°C. Food .
consumption and growth rates of the juvenile bass,
maintained on moderate densities of forage fish, increased with temperature
and decreased at the reduced dissolved oxygen concentrations except at 13°C.
Exposure to that temperature probably slowed metabolic processes of the bass
so much that their total metabolic rates were not limited by dissolved oxygen
except at very low concentrations.?
These largemouth bass studies clearly
support the idea that higher temperatures exacerbate the adverse effects of
15
low dissolved oxygen on the growth rate of fish (Table 5). Comparisons of
Brake's pond studies with the laboratory growth studies of Stewart et al.
(1967) suggest that laboratory growth studies may significantly underestimate
the adverse effect of low dissolved
oxygen on
fish growth. . Stewart's six
studies with largemouth bass are summarized
in
Table 4 and Brake's data are
presented in Table 5. All of Stewart's tests were conducted at 26°C, about
the highest temperature in Brake's studies, but comparison of the data show
convincingly that at dissolved oxygen concentration's between 4 and 6 mg/1 the
growth rate of bass in ponds was reduced 17 to 34 percent rather than the 1 to
9 percent seen in the laboratory studies. These results suggest that the ease
of food capture in laboratory studies may result in underestimating effects of
low dissolved oxygen on growth rates in'nature.
in ponds (after Brake, 1972; JRB Associates, 1984).
Temperature
(°C)
Percent Reduction in Growth Rate at
4.2 ± 0.2 mg/l.
4.9 ±
0.24g/1? 5.8
± 0.2 mg/1
13.3
0
13
.6
--
7
16.3
18
16.7
--
15
18.1
19
18.6
' 34
- -
18.7
.
18
--
23.3
26
- -
26
.7
--
17
27.4
31
Brett and Blackburn (1981) reanalyzed the growth data previously pub-
lished by other authors for largemouth bass, carp, and coho salmon in addition
to their own results for young coho and sockeye salmon. They concluded for
all species that above a critical level ranging from 4.0 to 4.5 mg/1,
decreases in growth rate and food conversion efficiency were not statistically
significant in these tests
of
relatively short duration (6 to 8 weeks) under
the pristine conditions. of laboratory testing. EPA believes that a more
accurate estimate of the dissolved oxygen concentrations that have no effect
on growth and a better estimate of concentration:effect relationships can be
obtained by curve-fitting procedures (JCB Associates, 1984) and by examining
these results from a large number of studies. Brett and Blackburn added an
additional qualifying statement that it 'was not the purpose of their study' to
seek evidence on .the acceptable level of dissolved oxygen in nature because of
the problems of environmental complexity involving all life stages and
functions, the necessary levels of activity to survive in a competitive world,
and the interaction of water quality
. (or lack of it) with varying dissolved
Table 5. Effect of temperature on the percent reduction in growth rate of
largemouth bass exposed to various dissolved oxygen concentrations
16-
oxygen concentrations. Their cautious concern
.
regarding the extrapolation to
the real world of results obtained under laboratory conditions is consistent
with that of numerous investigators.
D.
Reproduction
A life-cycle exposure of the fathead minnow beginning with 1- to 2-month
old juveniles was conducted and effects of continuous low dissolved oxygen
concentrations on various life stages indicated that the most sensitive stage
was the larval stage (Brungs, -1971). No spawning occurred at 1 mg/1, and the
number of eggs produced per female was reduced at 2
mg/1 but not at higher
concentrations. Where spawning occurred, the percentage hatch of embryos
(81-89 percent) was not affected when the embryos were exposed to the same
concentrations as their parents-
. Hatching time varied with temperature, which
was not'controlled, but with decreasing dissolved oxygen concentration: the
average incubation time increased gradually from the normal 5 to nearly 8
days. Mean larval survival was 6 percent at 3 mg/1 and 25 percent at 4 mg/l.
Mean survival of larvae at 5 mg/1 was 66 percent as compared to 50 percent at
control dissolved oxygen concentrations. However, mean growth of surviving
larvae at. 5 mg/1 was about 20 percent lower than control larval growth.
Siefert and Herman (1977) exposed mature black crappies to constant dissolved
oxygen concentrations from 2.5 mg/1 to saturation and temperatures of 13-20°C.
Number of spawnings; embryo viability, hatching success,-and survival through
swim-up were similar at all exposures.
E.
Early Life Stages
Larval ' and juvenile non-salmonids are frequently more sensitive to
exposures to low dissolved oxygen than are other life stages. Peterka and
Kent (1976) conducted semi-controlled experiments at,natural spawning sites
,
of
northern pike.,
bluegill, pumpkinseed, and smallmouth bass in Minnesota.
Dissolved •oxygen concentrations were measured 1 and 10 cm from the bottom,
with observations being made on hatching success and survival of embryos, 'sac
larvae, and, in some instances, larvae. Controlled exposure for up to 8 hours
was performed in situ in small chambers with the dissolved oxygen controlled
by nitrogen- stripping. For all species tested, tolerance to short-term
exposure
-'to
low concentrations decreased from embryonic to larval stages.
Eight-hour exposure of embryos and larvae of northern pike to dissolved oxygen
concentrations caused no mortality of embryos at 0.6 mg/1 but was 100 percent
lethal to sac-larvae and larvae. The most sensitive stage, the larval stage,
suffered complete mortality following 8 hours at 1.6 mg/1; 'the next higher
concentration, 4 mg/1, produced no mortality. Smallmouth bass were at least
as sensitive, with nearly complete mortality of sac-larvae resulting' from
6-hour exposure to 2.2 mg/1, but no mortality occurred after exposure to 4.2
mg/l. Early life stages, of bluegill were more hardy, with embryos . tolerating
4-hour exposure to 0.5 mg/1, a concentration lethal to sac-larvae; sac-larvae
survived similar exposure to 1.8 mg/1, however. Because the most sensitive
stage of northern pike. was the later larval stage,' and because the younger
sac-larval stages of smallmouth bass and bluegill were the oldest stages
tested, the tests with these latter species may not have included the most
sensitive stage. Based on these tests, 4 mg/1 is tolerated, at least briefly,
by northern pike and may be tolerated by smallmouth bass, but concentrations
as high as 2.2 mg/1 are lethal.
17
Several studies have provided 'evidence of mortality or other significant'
damage to young non-salmonids as a result of a few weeks exposure to dissolved
oxygen concentrations in the 3 to 6 mg/1 range. Siefert et al. (1973) exposed
larval northern pike to various dissolved oxygen concentrations at 15 and 19°C
and observed reduced survival at concentrations as high as 2.9 and 3.4.mg/1.
Most ,of the mortality at these concentrations occurred at the time the larvae
initiated feeding. Apparently the added stress of activity at that time or a
greater oxygen requirement for that life stage was the determining factor.
There was a marked decrease in growth at concentrations below 3 mg/l. In a
similar study lasting 20 days, survival of walleye embryos and larvae was
reduced at 3.4 mg/1 (Siefert and. Spoor, 1974), and none survived at lower
concentrations. A 20 percent reduction in the survival of smallmouth bass
embryos and larvae occurred at a concentration of 4.4 mg/1 (Siefert et al.,
1974) and at 2.5 mg/1 all larvae died
in
the first 5 days after hatching. At
4.4 mg/1 hatching occurred earlier than in the controls and growth among
survivors was reduced. Carlson and Siefert (1974) concluded that concentra-
tions from 1.7 to 6.3 mg/1 reduced the growth of early stages of the large-
mouth bass by 10 to 20 percent. At concentrations as high as 4.5 mg/1,
hatching was premature and feeding was delayed; both factors could indirectly
influence survival, especially if other stresses were to occur simultaneously.
Carlson et al. (1974) also observed that embryos and larvae of channel catfish
are sensitive to low dissolved oxygen during 2- or 3-week exposures. Survival
at 25°C was slightly reduced at 5 mg/1 and significantly reduced at 4.2 mg/i.
At 28
°C survival was slightly reduced at 3.8, 4.6, and 5.4 mg/1; total
mortality occurred at 2.3 mg/l. At all reduced dissolved oxygen concentra-
tions at both temperatures, embryo pigmentation was lighter, incubation period
was extended, feeding was delayed, and growth was reduced. No effect of
dissolved oxygen concentrations as low as 2.5 mg/1 was seen on survival of
embryonic and larval black
,
crappie (Sieffert and Herman, 1977). Other
tolerant species are the white bass and the white sucker, both of which
evidenced adverse effect to embryo larval exposure only at dissolved oxygen
concentrations of 1.8 and 1.2 mg/1, respectively (Sieffert et al., 1974;
Sieffert and Spoor, 1974).
Data (Figure 1) on the effects of dissolved oxygen on the survival
of
embryonic and larval nonsalmonid fish show some species to be tolerant
(largemouth bass, white sucker, black crappie, and white bass)
and others
nontolerant (channel' catfish, walleye, northern pike, smallmouth bass). The
latter three species are often included with salmonids in a grouping of
sensitive coldwater fish; these data tend to support that placement.
F.?
Behavior
Largemouth bass in laboratory studies (Whitmore et al., 1960) showed a
slight tendency to avoid concentrations of dissolved oxygen of 3.0 and 4.6
mg/1 and a definite avoidance of 1.5 mg/l. Bluegills avoided a concentration
of 1.5 mg/1 but not higher concentrations. The environmental significance of
such a response is unknown, but if large areas are deficient in dissolved
oxygen this avoidance would-probably not greatly enhance survival. Spoor
(1977) expo
s
ed largemouth bass embryos and larvae to low dissolved oxygen for
brief exposures of a few hours. At 23 to 24°C and 4 to 5 mg/1, the normally
quiescent, bottom-dwelling yolk-sac larvae 'became very active and swam
18
vertically
to a few inches above the substrate. Such behavior in natural
systems would probably cause significant losses due to predation and simple
displacement from the nesting area.
G.
Swimming?
.
Effects of low dissolved Oxygen on the swimming performance of largemouth.
bass were studied by Katz et al. (1959) and Dahlberg
et al.
(1968). The
results in the former study were highly dependent upon season and temperature,
with summer tests at
25°C
finding no effect on continuous swimming for 24 hrs
at 0.8 ft/sec unless dissolved oxygen concentrations fell below, .2 mg/l. In
the fall, at 20°C, no fish were able to swim for a day at 2.8 mg/1, and in the
winter and 16° no fish
.
swam for 24 hours at 5' mg/l. These results are
consistent with those seen in salmonids'in that swimming performance appears
to be more sensitive to low dissolved oxygen at lower temperatures.
Dahlberg et al• (1968) looked at the effect of dissolved oxygen on
maximum swimming speed at temperatures near 25°C. They reported slight
effects
(less than
10%
reduction in maximum swimming speed) at concentrations
between 3
.
and 4.5 mg/1, moderate reduction (16-20%) between 2 and 3 mg/1 and
severe reduction (30-50%) at 1 to 1.5 mg/l.
H. Field Studies
Ellis (1937) reported results of field studies conducted at. 982 stations
on freshwater streams and rivers during the months of June through September,
1930-1935. During this time, numerous determinations of dissolved oxygen
concentrations were made. He concluded that 5 mg/1 appeared to be the lowest
concentration which may reasonably be expected
to maintain varied warmwater
fish species in good condition in inland streams. Ellis (1944) restated his
earlier conclusion and also
added that his study had included the measurement
of dissolved oxygen concentrations at night and various seasons. He did not
specify the frequency or proportion of diurnal or seasonal sampling, but the
mean number 'of samples over the 5-year study was about seven samples per
station.
Brinley (1944) discussed a 2-year biological survey of the Ohio River
Basin. He concluded that in the zone where dissolved oxygen
is
between 3 and
5
mg/i the fish are more abundant than at lower concentrations,
but show a
tendency to sickness, deformity, and parasitization. The field results show
that the concentration of 5
mg/1 seems to represent a general dividing line
between good and bad conditions for fish.
A
three-year study of fish populations in the Wisconsin River indicated
that sport fish (percids and centrarchids) constituted
a significantly
greater
proportion of the fish population at sites having mean summer dissolved oxygen
concentrations greater than 5
mg/1 than at sites averaging below 5 mg/1
(Coble,
1982). The differences could not be related to any observed habitat
variables other than dissolved oxygen concentration.
These three field studies all indicate that increases in dissolved oxygen
concentrations above 5
mg/1 do not produce
noteworthy improvements in the
composition, abundance, or condition of non-salmonid fish populations, but
19
that sites with dissolved oxygen concentrations below 5 mg/I have fish
assemblages with increasingly poorer population characteristics as the
dissolved oxygen concentrations become lower. It cannot be stressed too
strongly that these field studies lack definition with respect to the actual
exposure 'conditions experienced by the resident populations and the lack of
good estimates for mean and minimum exposure concentrations over various
periods precludes the establishment of numerical criteria based on these
studies. The results of these semi-quantitative field studies are consistent
with the criteria derived later in this document.
IV. Invertebrates
As stated earlier, there is a general paucity of information on the
tolerance of the many forms of freshwater invertebrates to low dissolved
oxygen. Most available data describe the relationship between oxygen concen-
tration and oxygen consumption or short-term survival of aquatic larvae of
insects. These data are further restricted by their emphasis on species
representative of relatively fast-flowing mountain streams.
One rather startling feature of these data is the apparently high
dissolved oxygen requirement for the survival of some species. Before
extrapolating from these data one should be cautious in evaluating the
respiratory mode(s) of the species, its natural environment, and the test
environment. Thus, many nongilled species respire over their entire.body
surface while many other species are gilled. Either form is dependent upon
the gradient of oxygen across the respiratory surface, a gradient at least
partially dependent upon the rate of replacement of the water immediately
surrounding the organism. Some insects, such as some members of
.
the mayfly
'genus, Baetis, are found on rocks in extremely swift currents; testing their
tolerance to low dissolved oxygen in laboratory apparatus at slower flow rats
may contribute to their inability to survive at high dissolved oxygen concen-
trations, In addition, species of insects that utilize gaseous oxygen, either'
from bubbles or surface atmosphere, may not be reasonably tested for tolerance
of hypoxia if their source of gaseous oxygen is deprived in the laboratory
tests.
In spite of these potential problems, the dissolved oxygen requirements
for the survival of many species of aquatic insects are almost certainly
greater than those of most fish species. Early indication of the high
dissolved oxygen requirements of some aquatic insects appeared in the research
of Fox et al. (1937) who reported critical dissolved oxygen concentrations for
mayfly nymphs in a static test system. ?
Critical concentrations for six.
species ranged from 2.2 mg/1 to 17 mg/1; three of the species bad critical
concentrations in excess of air saturation. These data suggest possible
extreme sensitivity of some species and also the probability of unrealistic
conditions of water flow. More recent studies in water flowing at 10 cm/sec
indicate critical dissolved oxygen concentrations for four species of stonefly
are between 7.3 and 4.8 mg/1 (Benedetto, 1970).
In a recent study of 22 species of aquatic insects, Jacob et al. (1984)
reported 2-5 hour LC50 values at unspecified "low to moderate" flows in a
stirred exposure chamber, but apparently with no flow of replacement water.
Tests were- run at one or more of five temperatures
. from 12 to 30°C; some
20
, •
species were tested at only one.temperature, others at as Many as four. The
median of the 22 species mean LC5Os was about 3 mg/1, with eight species
having an average LC50 below 1 mg/1 and four excess of 7 mg/l. The four
most sensitive species were two mayfly species and two caddisfly species. The
studies of Fox et al. (1937), Benedetto (1970), and Jacob et al. (1984) were
all conducted with European species, but probably have general relevance to
North American habitats. A similar oxygen consumption study of a No'rth
American stonefly (Kapoor and Griffiths, 1975) indicated a possible critical
dissolved oxygen concentration of about 7 mg/1 at a flow rate of 0.32 cm/sec
and a temperature of 20°C.
One type of behavioral observation provides evidence of hypoxic stress in
aquatic insects. As dissolved oxygen concentrations decrease, many species
of
aquatic insects can be seen to increase their respiratory movements, movements
that provide for increased water flow over the respiratory surfaces. Fox and
Sidney (1953) reported. caddisfly respiratory movements over a range of
dissolved oxygen from 9 to mg/l. A dissolved Oxygen decrease to 5 mg/1
doubled the number of movements and at 1 to 2 mg/I the increase was 3- to
4-fold.
Similar
data were
published by Knight and Gaufin (1963) who studied a
stonefly common
•
in the western 'United States. Significant increases occurred
below 5 mg/1 at 16°C and below 2 mg/1 at 10°C. Increases in movements
occurred at higher dissolved oxygen concentrations
-
when water flow was 1.5
cm/sec than 7.5 cm/sec, again indicating the importance of water flow
-
rate on
the respiration of aquatic insects. A subsequent paper by Knight and Gaufin
(1965) indicated that species of stonefly lacking, gills are more sensitive to
low dissolved oxygen than are gilled forms.
Two studies that provide
.
the preponderance
.
of the current data on
the
acute effects of low dissolved oxygen concentrations on aquatic insects are
those of Gaufin (1973) and Nebeker (1972) which together provide reasonable
96-hr LC50 dissolved oxygen concentrations for 26 species of aquatic insects
(Table 6). The two studies contain variables that make them difficult to
compare or evaluate fully. Test temperatures were 6.4°C in-Gaufin's study and
18.5°C in Nebeker's. Gaufin used a vacuum degasser while Nebeker used a
30-foot stripping column that probably produced an unknown degree of super-
saturation with nitrogen. The water velocity is not given in either paper,
although
flow
rates are given but test chamber dimensions are not deafly
specified. The overall similarity of the test results suggests that.potential
supersaturation and lower flow volume in Nebeker's tests did not have a
significant .
effect on the results.
Because half
of
the insect species tested had 96-h LC50 dissolved oxygen
concentrations between 3 and 4 mg/1 it appears that these species (collected
in Montana and Minnesota) would require at least 4 mg/1 dissolved oxygen to
ensure their survival. The two most sensitive species represent surprisingly
diverse habitats,
Ephemerella doddsi is found in swift rocky streams and has
an LC50 of 5.2 mg/1 while the pond mayfly,
Callibaetis montanus,
has an LC50
of 4.4 mg/l. It is possible that the test conditions represented too slow a •
flow for E.
doddsi and too stressful flow conditions for C.
montanus.
21?
.
22
Table 6.
?
Acutely
?
lethal
?
concentrations
insects.
af?
dissolved oxygen to
aquatic
96-h LC50
Species
(mg/1)
Source*
Stonefly
Acroneuria pacifica
?
.?
• 1.6 (H)**
G
Acroneuria lycorias
3.6
N
Acrynopteryx aurea
3.3 (H)
G
Arcynopteryx parallela
< 2?
(H)
G
Diura
knowltoni
3.6 (L)
G
Nemoura cinctipes
3.3 (H)
G
Pteronarcys californica
3.9 (L)
G
Pteronarcys californica
3.2 (H)
G
Pteronarcys dorsata
2.2
N
Pteronarcella badia
2.4 (H)
G
Mayfly
Baetisca laurentin'a
3.5
N
tT-TTEg
itis
montanus
4.4 (L)
G
Ephemerella doddsi
5.2 (L)
G
Ephemerella grandis
3.0 (H)
G
Ephemerella subvaria
3.9
N
Hexagenia limbata
1.8 (H)
G
Hexagenia limbata
1.4
N
Leptophlebia nebulosa
2.2
N
Caddis.fiy
Brachycentrus occidentalis
< 2?
(L)
G
Drusinus sp.
1.8 (H)
G
Hydropsyche sp.
3.6 (L)
G
Hydropsyche betteri
2.9 (21°C)
N
Hydropsyche betteri
2.6 (18.5°C)
N
Hydropsyche betteri
2.3 (17°C)
N
Hydropsyche betteri
1.0 (10°C)
N
Lepidostoma sp.
< 3
?
(H)
G
Limnophilus ornatus
3.4 (L)
G
14 '?
Neophylax sp.
3.8 (L)
G
Neothremma alicia
1.7
?
(L)
G
Diptera'
Simulium vittatum
3.2 (L)
G
Tanytarsus dissimilis
< 0.6
N
* G = Gaufin (1973) -- all tests at 6.4°C.
N = Nebeker (1972) -- all tests at 18.5°C except as noted/flow 125 ml/min.
** H = high flow (1000 ml/min); L-= low flow (500 ml/min).
Other freshwater invertebrates have been subjected to acute hypoxic
'stress •and their LC50 values determined. Gaufin (1973) reported a 96-h LC50
for the amphipod Gammarus limnaeus of < 3 mg/l. Four other crustaceans were
studied by Sprague (1963) who reported the following 24-h LC50s: 0.03 mg/1,
Asellus intermedius; 0.7 mg/1, Hyalella azteca; 2.2 mg/1, Gammarus pseudo-
limnaeus; and 4.3 mg/I, Gammarus fasciatus. The range of acute sensitivities
of
7
—
Efi
species appears similar
tothat Teported for aquatic insects.'
There are
.
few long-term studies of freshwater invertebrate tolerance to
low dissolved oxygen concentrations. Both Gaufin (1973) and Nebeker (1972)
conducted long-term survival studies with insects, but both are questioned
because of starvation and potential nitrogen supersaturation, respectively.
Gaufin's data for eight Montana species and 17 Utah species suggest that 4.9
mg/1 and 3.3 mg/1, respectively, would provide for 50 percent survival for
from •10 to 92 days. Nebeker lists 30-d LC50 values for five species, four
between 4.4 and 5
•0
mg/1 and one < 0.5 mg/l. Overall, these data indicate
that prolonged exposure to dissolved oxygen concentrations below 5 mg/1-would
have deterimental effects on a large proportion of the aquatic insects common
in areas like Minnesota, Montana, and Utah. Information from other habitat
types and geographic locations would provide a broader picture of invertebrate
dissolved oxygen requirements.
A more classic toxicological protocol wasused by Homer and Waller (1983)
in a study of the effects of low dissolved oxygen on Daphne magna. In a 26-d
chronic exposure test, they reported that 1.8 mg/1 significantly reduced.
fecundity and 2.7 'mg/1 caused,a 17.percent reduction In final weight of
adults. No effect was seen at 3.7 mg/1..
In summarizing the state of knowledge regarding the relative sensitivity
of fish and invertebrates' to low dissolved oxygen, it seems that some species
of insects and other crustaceans are
•
killed at concentrations survived by all
species of fish tested. Thus, while most fish will survive exposure to 3
mg/1, many species of invertebrates are killed by concentrations as high as 4
mg/l. The extreme sensitivity of a few species of aquatic inects may be an
artifact of the testing environment. Those sensitive species common to swift
flowing, coldwater streams may require very high concentrations of dissolved
oxygen. -On' the other hand, those stream• habitats are probably among the least
likely to suffer• significant dissolved oxygen depletion.
Long-term impacts of hypoxia are less well known for invertebrates than
for fish. Concentrations adequate to avoid impairment of fish production
probably will 'provide reasonable protection for invertebrates as long as
lethal concentrations are avoided.
V.?
•
Other Considerations
A.?
Effects of Fluctuations.
. Natural dissolved oxygen concentrations fluctuate on a seasonal and daily
basis, while in most laboratory studies the oxygen levels are held essentially
constant. In two studies on the effects of daily oxygen cycles the authors
concluded that growth of fish fed unrestricted rations was markedly less than
would be estimated from the daily mean dissolved oxygen concentrations
23
(Fisher, 1963; Whitworth, 1968). The growth of these fish was only slightly
above that attainable during constant exposure to the minimum concentrations
of the daily cycles. A diurnal dissolved oxygen pulse to 3 mg/1 for 8 hours
per day for 9 days, with a concentration of 8.3 mg/1 for.the remainder of the
time, produced a significant stress pattern in the serum protein fractions of.
bludgill and largemouth bass but not yellow bullhead (Bouck and Ball, 1965).
During periods of low dissolved oxygen the fish lost their natural color,
increased their ventilation rate, and remained very quiet. At these times
food was ignored. Several times, during the low dissolved oxygen concentra-
tion part of the cycle, the fish vomited food which they had eaten as much as
12 hours earlier. After comparable exposure of the rock bass, Bouck (1972)
observed similar results on electrophoretic patterns and feeding behavior.
Stewart et al. (1967) exposed juvenile largemouth bass to patterns of
diurnally-variable dissolved oxygen concentrations with daily minima near 2
mg/1 and daily maxima from 4 to 17 mg/l. Growth under any fluctuation pattern
was almost always less than the growth that presumably would have occurred had
the fish been held at a constant concentration equal to the mean concentra-
tion.
Carlson et al. (1980) conducted constant and diurnally fluctuating
exposures with juvenile channel catfish and yellow perch. At mean constant
concentrations of 3.5 mg/1 or less, channel catfish consumed less food and
growth was significantly reduced. Growth of this species was not reduced at'
fluctuations from about 6.2 to 3.6 and 4:9 to 2 mg/1, but was significantly
impaired at a fluctuation from about 3.1 to 1 mg/l. Similarly, at mean
constant concentrations near 3.5 mg/1,-yellow perch consumed less food but
growth was not impaired'until concentrations were near 2 mg/l. Growth was not
affected by fluctuations- from about 3.8 to 1.4 mg/l. No dissolved oxygen-
related mortalities were observed. In both the channel catfish and the yellow
perch experiments, growth rates during the tests with fluctuating dissolved
oxygen were considerably below the rate, attained in the constant exposure
tests. As a result, the fluctuating and constant exposures could not be
compared. Growth would presumably have been more sensitive in the fluctuating
tests if there had been higher rates of control growth.
Mature black crappies were, exposed to constant and fluctuating dissolved
oxygen concentrations (Carlson and Herman,
•
1978). Constant concentrations
were near 2.5, 4, 5.5, and 7 mg/1 and fluctuating concentrations ranged from
0.8 to 1.9 mg/1 above and below these original concentrations. Successful
spawning occurred at
.
allexposures except the fluctuation between 1.8 and 4.1
mg/i.
In considering daily or longer-term cyclic exposures to low dissolved
oxygen concentrations, the minimum values may be more important than the mean
levels. The importance of the daily minimum as a determinant of growth rate
is common to the results of Fisher (1963), Stewart (1967), and Whitworth
(1968). Since annual low dissolved oxygen concentrations normally occur,
during warmer months, the significance of reduced growth rates during the
period in question must be considered. If growth rates are normally low, then
the effects of low dissolved oxygen concentration. on growth could be minimal;
if normal growth rates are high, the effects could be significant, especially
if the majority of .the annual growth occurs during the period in question.
24
B.?
Temperature and Chemical Stress
When fish were exposed to lethal temperatures, their survival times were
reduced when the- dissolved
.
oxygen concentration was lowered from 7.4 to 3.8
mg/1 (Alabaster and Welcomme, 1962). Since .
high temperature and low dissolved
oxygen commonly occur together in natural environments, this likelihood of
additive or synergistic effects of these two potential stresses is a most
important consideration.
High temperatures almost certainly increase the adverse effects of low
dissolved oxygen concentrations. However, the spotty, irregular acute
lethality data base provides little basis for quantitative, predictive
analysis. Probably the most complete study is that on rainbow trout, perch,
and roach conducted by Downing and Merkens (1957). Because their study was
spread over an 18-month period, seasonal effects could have influenced the
effects at the various testtemperatures.. Over a range from approximately' 10
to 20°C, the lethal dissolved oxygen concentrations increased by an average
factor of about 2.6, ranging from 1.4 to 4.1 depending on fish species tested
and test duration. The influence of temperature on chronic effects of low
dissolved oxygen concentrations are not well known, but requirements for
dissolved oxygen probably increase to some degree with increasing temperature.
This generalization is 'supported by analysis of salmon studies reported by
Warren et al. (1973) and the largemouth bass studies of Brake (1972). .
Because most laboratory tests are conducted at temperatures near the
mid-range of a species temperature tolerance, criteria.based on these test.
data will tend to be under-protective at higher temperatures 'and over-
protective at lower temperatures. Concern for this temperature effect was a
consideration in establishing
these
criteria,, especially in the establishing
of those criteria intended to prevent short-term lethal effects.
A detailed discussion and model for evaluating interactions among
temperature, dissolved oxygen, ammonia, fish size,.and ration on the resulting
growth of individual fish (Cuenco et al., 1985a,b,c) provides an excellent,
in-depth evaluation of potential effects of dissolved oxygen on fish growth.
Several laboratory studies evaluated the' effect of reduced dissolved
oxygen concentrations on the toxicity of various chemicals, some of which
occur commonly in oxygen-demanding wastes. Lloyd (1961) observed that the
toxicity of zinc, lead, copper, and monohydric
phenols was increased at
dissolved oxygen concentrations as high as approximately 6.2 mg/1 as compared
to 9.1 mg/i. At 3.8 mg/1, the toxic effect of these chemicals was even
greater. The toxicity of
ammonia was enhanced
by low dissolved oxygen more
than that of other toxicants. Lloyd theorized that the increases in toxicity
of the chemicals were due to increased
,
ventilation at low dissolved oxygen
concentrations; as a consequence of increased ventilation, more water, and
therefore more toxicant, passes the fish's gills. Downing and Merkens (1955)
reported that survival times of rainbow trout at lethal ammonia concentrations
increased markedly over a range of dissolved'oxygen concentrations from
.
1.5
to
8.5 mg/l. Ninety-six-hr LC50 values for rainbow trout indicate that ammonia
became more toxic with decreasing dissolved oxygen concentrations from 8.6 to
2.6 mg/1 (Thurston et al., 1981). The maximum increase in toxicity was by
about a ;Factor of 2.
?
They also compared ammonia LC50 values at reduced
25.
dissolved oxygen concentrations after 12, 24, 48, and 72 hrs. The shorter the
time period, the more pronounced the positive relationship between the LC50
and dissolved oxygen concentration. The authors recommended that dissolved
oxygen standards for the protection of salmonids should reflect background
concentrations of ammonia which may be present and the likelihood of temporary
increases in those concentrations. Adelman and Smith (1972) observed that
decreasing dissolved oxygen concentrations increased the toxicity of hydrogen
sulfide to goldfish. When the goldfish were acclimated to the reduced
dissolved oxygen concentration before the exposure to hydrogen sulfide began,
mean 96-hr LC50 values were 0.062 and 0.048 mg/1 at dissolved oxygen concen-
trations of 6 and 1.5 mg/1, respectively. When there was no prior acclima-
tion, the LC50 values were 0.071 and 0.053 mg/1 at the same dissolved oxygen
concentrations. These results demonstrated a less than doubling
.
in toxicity
of hydrogen sulfide and little difference with regard to prior acclimation to
reduced dissolved oxygen concentrations. Cairns and Scheier (1957) observed
that bluegills were less tolerant to zinc, naphthenic acid, and potassium
cyanide at periodic low dissolved oxygen concentrations. Pickering (1968)
reported that an increased mortality of bluegills exposed to zinc resulted
from the added stress of low dissolved oxygen concentrations. The difference
in mean LC50 values between low (1.8' mg/1) and high (5.6 mg/1) dissolved
oxygen concentrations was a factor of 1.5.
Interactions between other•stresses and low dissolved oxygen concentra-
tions can greatly increase mortality of trout larvae. For example, sublethal
concentrations of pentachlorophenol and oxygen combined to produce 100 percent
mortality of trout larvae held at an oxygen concentration of 3 mg/1 (Chapman
and Shumway, 1978). The survival of chinook salmon embryos and larvae reared
at marginally high temperatures was reduced by any. reduction in dissolved
oxygen, especially at concentrations below 7 mg/1 (Eddy, 1972).
In general, the occurrence of toxicants in the water mass, iii
combination
with low dissolved oxygen concentration, may lead to a potentiation
of stress
responses on the part of aquatic organisms (Davis, 1975a,b). Doudoroff and
Shumway (1970) recommended that the disposal of toxic pollutants must, be
controlled so that their concentrations would not be unduly harmful at
prescribed, acceptable concentrations of dissolved oxygen, and these accept-
able dissolved oxygen concentrations should be independent of existing or
highest permitted concentrations of toxic wastes.
C.?
Disease Stress
In a study of 5 years of case records at
.
fish farms, Meyer (1970)
observed that incidence of infection with Aeromonas liquefasciens (a common
bacterial pathogen of fish) was most prevale7a7R776 June, July, and August.
He considered low oxygen stress to be a major factor in outbreaks of Aeromonas
disease during summer months. Haley et al. (1967) concluded that a kill of
American and threadfin shad in the San Joaquin River occurred as a result of
Aeromonas infection the day after the dissolved oxygen was between 1.2 and 2.6
mg/l. In this kill the'lethal agent was Aeromonas but the'additional
stress
of the low dissolved oxygen may have been a significant
factor.
.26
Wedemeyer (1974) reviewed the role
of
stress as a predisposing factor
in
fish diseases and concluded that facultative fish pathogens are continuously
present in most waters. Disease problems seldom occur, however, unless
environmental quality and 'the host defense systems
of
the fish also deter-
iorate.
?
He listed furunculosis, Aeromonad and Pseildomonad hemorrhagic
septicemia,- and vibriosis as diseases for which low dissolved oxygen is one
environmental factor predisposing fish to epizootics. He stated that to
optimize fish health, dissolved oxygen concentrations should be 6.9 mg/1 or
higher. Snieszko
(1974) also stated that outbreaks of diseases are probably
more likely if
the occurrence of stress coincides with the presence of
pathogenic microorganisms.
VI.
Conclusions
The primary determinant for the criteria is laboratory data describing
effect on growth, with developmental rate and survival included in embryo and
larval production levels. For the purpose of deriving criteria, growth in the
laboratory and production in nature are considered equally sensitive to low
dissolved oxygen. Fish production in natural communities actually may be
significantly more, or less, sensitive than growth in the laboratory, which
represents only one simplified facet of production.
The dissolved oxygen criteria are based primarily on data developed in
.the Taboratory under conditions which are usually artificial in several
important respects. First, they routinely preclude or minimize most environ-
mental stresses and biological interactions that under natural conditions are
likely to increase, to a variable and unknown extent, the effect of low
dissolved oxygen concentrations. Second, organisms are usually given no
opportunity to acclimate to low dissolved oxygen concentrations
prior to tests
nor
can they avoid the test exposure. Third, food availability is unnatural
because the fish have easy, often unlimited, access to food without signif-
icant energy. expenditure for search
. and capture. Fourth, dissolved oxygen
concentrations are kept nearly constant
so
that each exposure represents both
a minimum and an .average concentration. This circumstance complicates
application of the data to natural systems with fluctuating dissolved oxygen
concentrations.
Considering the latter problem only, if
the laboratory data are applied
directly as minimum allowable criteria,
the criteria will presumably be higher
than necessary because the mean dissolved oxygen concentration will often be
significantly higher than the criteria. If applied as a mean, the criteria
could allow complete anoxia and total mortality during brief periods of very
low dissolved oxygen or could allow too many consecutive daily minima near the
lethal threshold.
If
only a minimum or a mean can be given as a general
criterion, the minimum
must be chosen because averages are too independent of
the extremes.
Obviously, biological effects of low
dissolved oxygen concentrations
depend upon means, minima, the duration and frequency'of the minima, and the
period of averaging. In many respects, the effects appear to be independent
of the maxima; for example, including supersaturated dissolved
oxygen values
in the
average may produce mean dissolved oxygen concentrations that are
misleadingly high and unrepresentative of the true biological stress
of
the
dissolved oxygen minima,
27
28
Because most experimental exposures have been constant, data on the
effect of exposure to fluctuating dissolved oxygen concentrations is sketchy.
The few fluctuating exposure studies have used regular, repeating daily cycles
of an on-off nature with 8 to 16 hours at low dissolved oxygen and the
remainder of the 24 hr period at intermediate or high dissolved oxygen. This
is an uncharacteristic exposure pattern, since most daily dissolved oxygen
cycles are of a sinusoidal curve shape and not a square-wave variety.
The existing data allow a tentative theoretical dosing model for fluctu-
ating dissolved oxygen only as applied to fish growth, The EPA believes that
the data of Stewart et al. (1967) suggest that effects on growth are reason-
ably represented by calculating the mean of the daily cycle using as a maximum
value the dissolved oxygen concentration which represents. the threshold effect
concentration during continuous exposure tests. For example, with an effect
threshold of 6 Ag/1, all values in excess of 6 mg/1
should be averaged as
though they were.6 mg/l. Using this procedure, the growth effects appear to
be a reasonable function of the mean, as long as the minimum
is
not lethal.
Lethal thresholds are highly dependent upon exposure duration, species, age,
life stage, temperature, and a wide variety of other factors. Generally the
threshold is between 1 and 3 mg/l.
A most critical and poorly documented aspect of a dissolved oxygen cri-
terion is the question of acceptable and unacceptable minima during dissolved
oxygen cycles of varying periodicity. Current ability to predict effects of
exposure to a constant dissolved oxygen level is only fair; the effects of
regular, daily •dissolved oxygen cycles can only be poorly estimated; and
predicting the effects of more stochastic patterns of dissolved oxygen
fluctuations requires an ability to integrate constant and cycling effects.
Several general conclusions result from the synthesis of available field
and laboratory data. Some of these conclusions differ from earlier ones in
the literature, but the recent data discussed in this document have provided
' additional detail and perspective.
0?
Naturally-occurring dissolved oxygen concentrations may occasionally fall
below target criteria levels due to a combination of low flow, high
temperature, and natural oxygen demand. These naturally-occurrina
conditions represent a normal situation in which the productivity of fish
or other aquatic organisms
may
not be the maximum possible under ideal
circumstances, but which represent the maximum productivity under the
particular set of natural conditions. Under these circumstances the
numerical criteria should be considered' unattainable, but naturally-
occurring conditions which fail to meet criteria should not be inter-
preted as violations of criteria. Although further reductions in dis-
solved oxygen may be inadvisable, effects of any reductions should be
compared
,
to natural ambient conditions' and not to ideal conditions.
0
Situations' during which attainment of appropriate criteria is most
critical include periods when attainment of high fish growth rates is a
priority, when temperatures approach upper-lethal levels, when pollutants
are present in near-toxic quantities, or when other significant stresses
are suspected.
0
Reductions
in
growth rate produced by a given low dissolved oxygen
concentration are probably more severe as temperature increases. Even
during periods when- growth rates are normally low, high temperature
stress increases the sensitivity of aquatic-organisms to disease and
toxic pollutants, making the attainment of proper dissolved oxygen
criteria particularly important. For these reasons, periods of highest
temperature represent a critical portion of the year with respect to
dissolved oxygen requirements.
0
In salmonid spawning habitats, intergravel 'dissolVed.oxygen concentra7
tions.are significantly reduced by respiration Of fish
,
embryos and other
organisms: Higher water column concentrations of dissolved oxygen are
required to provide protection of fish embryos and:larvae which develop
in
.
the intergravel environment. A 3 mg/1 difference is used ln the
criteria to account for this factor.
0
The early life stages, especially the larval stage, of non-salmonid fish
are usually most sensitive to reduced dissolved oxygen stress. Delayed
development, reduced larval survival, and reduced larval and post-larval
growth are the observed effects. A separate early life stage criterion
for non-salmonids is established to protect these more sensitive stages
and is to apply from spawning through 30 days after hatching.
? •
0
Other life stages of salmonids appear to be somewhat more sensitive than
other life stages of the non-salmonids, but this difference, resulting in
a 1.0'
mg/1 difference in the criteria for other life stages, may be due
to a more complete and precise data,base for' salmonids. Also, this
difference is at least-partially due to the colder water temperatures at
which salmonid testS are conducted and the resultant higher dissolved
oxygen concentration in oxygen-saturated control water.
0
Few appropriate data are available on the effects of reduced dissolved
oxygen' on freshwater invertebrates. However, historical concensus states
that, if all life stages of fish are protected, the invertebrate commu-
nities, although not necessarily unchanged, should be adequately pro-
tected. This is a generalization to which there may
.
be exceptions of
environmental significance. Acutely lethal concentrations of dissolved
oxygen appear to be higher for many aquatic insects than for fish.'-.
0
Any dissolved oxygen criteria should include absolute minima to prevent
mortality due to the direct effects of hypoxia, but such minima alone may
not be
sufficient protection for the long-term-persistence of sensitive
populations under natural conditions. Therefore, the criteria minimum
must also provide reasonable assurance that regularly repeated or
prolonged exposure for days or weeks at the allowable minimum will avoid
significant physiological stress of sensitive organisms.
Several earlier dissolved oxygen criteria were presented in the form of a
family of curves (Doudoroff and Shumway, 1970) or equations (NAS/NAE, 1973)
which yielded various dissolved .oxygen requirements depending on the quali-
tative degree of fishery protection or risk deemed suitable at a given site.
Although dissolved oxygen concentrations that risk significant loss of
fishery
production are not consistent with the intent of water quality criteria, a
29
qualitative protection/risk assessment for a range of dissolved oxygen
concentrations has considerable value to resource managers. Using qualitative
descriptions similar to those presented in earlier criteria of Doudoroff. and
Shumway (1970) and Water Quality Criteria 1972 (NAS/NAE, 1973), four levels of
risk are listed below:
No Production Impairment. Representing nearly maximal protection of fishery
resources.
Slight Production Impairment. Representing a high level of protection of
important fishery resources; risking only slight impairment of production
in most cases.
Moderate Production Impairment. Protecting the persistence of existing fish
populations but causing considerable loss of production.
Severe Production Impairment. For low level protection of fisheries of some
value but whose protection in comparison with other water uses cannot be
a major objective of pollution control.
Selection of dissolved oxygen concentrations equivalent to each of these
levels of effect requires some degree of judgment based largely upon examina-
tion of growth and survival data, generalization of response curve shape, and
assumed applicability
.
of laboratory responses to natural populations. Because.
nearly all data on the effects of low dissolved oxygen on aquatic organisms.
relate to continuous exposure for relatively short duration (hours to weeks),
the resultant dissolved oxygen concentration-biological effect estimates are
most apOlicable to essentially constant exposure levels, although they may
adequately represent mean concentrations as well:
The production impairment values are necessarily subjective, and the
definitions taken from Doudoroff and Shumway (1970) are more descriptive than
the accompanying terms "slight," "moderate," and "severe." The impairment
values for other life stages are derived predominantly from the growth data
summarized in the text and tables in Sections II and III. In general, slight,
moderate, and severe impairment are equivalent to 10, 20, and 40 percent
growth impairment, respectively. Growth impairment•of 50 percent or greater
is often accompanied by mortality, and conditions allowing a combination of
severe growth impairment and mortality are considered as no .protection.
Production impairment levels for early life stages are quite subjective
and should be viewed as convenient divisions of the range of dissolved oxygen
concentrations between the acute mortality limit and the no production
impairment concentrations.
Production impairment values for invertebrates are based on survival in
both long-term and short-term studies. There are no studies of warmwater
species and few of lacustrine species.
The following is a summary of the dissolved oxygen concentrations (mg/1)
judged to be equivalent to the various qualitative levels of effect described
earlier; the value cited as the acute mortality limit is the minimum dissolved
oxygen concentration deemed not to risk direct mortality of sensitive
organisms:?
.
1.?
Salmonid Waters
a.
?
Embryo and Larval Stages
0
0
0
0
0
No Production. Impairment
?
= 11* (8).
Slight Production Impairment?
= 9* (6)
Moderate Production Impairment =
8* (5)
Severe Production Impairment?
= 7* (4)
Limit to Avoid Acute Mortality =
6* (3)
(* Note: These are water column concentrations recommended to achieve the
required intergravel dissolved oxygen concentrations shown in
parentheses. The 3 mg/1 difference is discussed in the criteria.
document.)
b.?
Other Life Stages
O
No. Production Impairment?= 8
O
Slight Production Impairment = 6
O
Moderate Production Impairment = 5
O
Severe Production Impairment = 4
O
Limit to Avoid Acute Mortality = 3
2.?
Nonsalmonid Waters
a.?
Early Life Stages
O
No Production Impairment
?
= 6.5
O
Slight Production Impairment = 5.5
O
Moderate Production Impairment = 5 .
O
Severe. Production Impairment = 4.5
O
Limit to Avoid Acute Mortality = 4
b.
?
Other Life Stages
O
No Production Impairment?
= 6
O
Slight Production Impairment = 5
0
Moderate Production Impairment = 4
O
Severe Production Impairment = 3.5
O
Limit to Avoid Acute Mortality = 3
3.?
Invertebrates
0
No Production Impairment
?
= 8
O
Some Production Impairment ?
= 5
O
Acute Mortality Limit?
= 4
Added Note
Just prior to final publication of this criteria document, a paper
appeared (Sowden and Power, 1985) that provided an interesting field valida-
tion of the salmonid early life stage criterion and production impairment
estimates. A total of 19 rainbow trout redds were observed for a. number of
31
parameters including percent survival of embryos, dissolved oxygen concentra-
tion, and calculated intergravel water velocity. The results cannot be
considered a rigorous evaluation of the criteria because of the paucity of
dissolved oxygen determinations per redd (2-5) and possible inaccuracies in
determining percent survival and velocity.
?
Nevertheless, the qualitative
validation is-striking.
The generalization drawn from Coble's (1961) study that good survival
occurred when mean intergravel dissolved oxygen concentrations exceeded 6.0
mg/1 and velocity exceeded 20 cm/hr was confirmed; 3 of the 19 redds met this
criterion and averaged 29 percent embryo survival. The survival in the other
16 redds averaged only 3.6 percent. The data from the study are summarized in
Table 7. The critical intergravel water velocity from this study appears to
be about 15 cm/hr. Below this velocity even apparently good dissolved oxygen
Table 7. Survival of rainbow trout embryos as a function of intergravel
dissolved oxygen concentration and water velocity (Sowden and Power,
1985) as compared to dissolved oxygen concentrations established as
criteria 'or estimated as producing various levels of production
impairment.
Criteria Estimates
Dissolved Oxygen
Concentration
mg/1
Percent
Survival
Water
Velocity,
cm/hr
Mean
Survival
(Flow
Mean
Minimum
> 15 cm/hr)
Exceeded Criteria
8.9
8.0
22.1
53.7
7.7
7.0
43.5
83.2
29.0
7.0
6.4'
1.1
9.8
6.9
5.4
21.3
20.6
Slight Production
7.4
4.1
0.5
7.2
Impairment
7.1
4.3
21.5
16.3
6.7
4.5
4.3
5.4
15.6
6.4
4.2
0.3
7.9
6.0
4.2
9.6-
17,4
Moderate Production
5.8
3.1
13.4
21.6
Impairment
5.3
3.6
5.6
16.8
6.5
5.2
3.9
0.4
71.0
Severe Production
4.6
4.1
0.9
18.3
0.9
Impairment
4.2
3.3
0.0
0.4
Acute Mortality
3.9
2.9
0.0
111.4
3.6
2.1
0.0-
2.6
2.7
1.2
0.0
4.2
-?
0.0
2.4
0.8
0.0
1.1
2.0
0.8
0.0
192.0
Ow
32
characteristics do not produce reasonable survival. At water velocities fn.
excess of 15 .cm/hr the average percent survival in the redds that had
dissolved oxygen concentrations that met the criteria was 29.0 percent. There
was
.
no survival in redds that had dissolved oxygen minima below the acute
mortality limit. Percent survival in redds with greater than 15 cm/hr flow
averaged 15.6; 6.5, and 0.9 percent for redds meeting slight, moderate, .and
severe' production impairment levels, respectively.
Based on an average redd of 1000 eggs, these mean percent survivals
would
be equivalent to 290, 156, 65, 9, and 0 viable larvae entering the environment
to produce food
for
other fish, catch for fishermen, and eventually a new
generation of spawners to replace- the parents of the embryos in the redd.
Whether or not these survival numbers ultimately represent the impairment
definitions is moot in the light of further survival and growth uncertainties,
but the quantitative field results and the qualitative and quantitative
impairment.and criteria values are surprisingly similar.
VII.
National Criterion
The national criteria for ambient dissolved oxygen
concentrations
for the
protection of freshwater aquatic life are presented in Table 8. The
criteria
are
derived from the production impairment estimates on the preceding page
which are
in
turn based primarily upon growth data and information on tempera-.
ture,. disease, and pollutant stresses. The average dissolved oxygen concen-
trations selected are values 0.5 mg/1 above the slight production impairment
values and represent values between no production impairment and slight
production impairment. Each criterion may thus be viewed as
,
an estimate of
the threshold concentration below which detrimental effects are expected.-
Criteria for coldwater fish are intended to apply to waters containing a
population of one or more species in the family Salmonidae (Bailey et al.,
1970)
or to waters containing other coldwater or coolwater fish deemed by the
user to be closer to salmonids in sensitivity than to most warmwater species.
Although the acute lethal limit for salmonids' is at or below 3 mg/1, the
coldwater minimum has been established at 4 mg/1 because .a significant
proportion of the insect species common to salmonid habitats are less tolerant
of acute exposures to low dissolved oxygen than are salmonids. Some coolwater
species may require more protection than that afforded by the other life stage
criteria for warmwater fish and it may be desirable to protect sensitive
coolwater species with the coldwater* criteria. Many states have 'more
stringent dissolved oxygen standards for cooler waters, waters that contain
either saldonids, .nonsalmonid coolwater fish, or the sensitive centrarchid,
the smallmouth bass. The warmwater criteria are necessary to protect early
life stages of warmwater fish as sensitive as channel catfish and to protect
other life stages of fish as sensitive as largemouth bass. Criteria for early
life stages are intended to apply only where and when these stages occur.
These criteria represent dissolved oxygen concentrations which EPA believes
provide a reasonable and adequate degree of protection for freshwater aquatic
life.
The criteria do not represent assured no-effect levels. .However, because
the criteria represent worst case conditions (i.e., for wasteload allocation
and waste treatment plan design), conditions'wil• be better than the criteria
- 33
Table 8. Water quality criteria for ambient dissolved oxygen concentration.
Coldwater Criteria
Warmwater Criteria
Early Life
Stages 1 ' 2
Other Life
Stages
Early Life
Stages2
Other Life
Stages
30 Day Mean.
NA3
6.5
NA
5.5
7 Day Mean
9.5 (6.5)
NA
6:0
NA
7 Day Mean
NA
5.0
NA
4.0
Minimum
1 Day Minimum
4 '
5
8.0 (5.0)
4.0
5.0
3.0
1
These are water column concentrations recommended to achieve the required
intergravel
dissolved oxygen concentrations shown in parentheses. The 3
mg/1 differential is discussed in the criteria document. For species that.
have early life stages exposed directly to the water column, the figures in
parentheses'apply.
2
Includes all embryonic and larval stages and all juvenile forms to 30-days
following hatching.
3
NA (not applicable).
4
For highly manipulatable discharges, further restrictions apply (see•page
37)
5
All minima should be considered as instantaneous concentrations to be
achieved at all times.
nearly all the time at most sites. In situations where criteria conditions.
are just maintained for considerable periods, the criteria represent some risk
of production impairment. This impairment would probably be slight, but would
depend on innumerable other factors. If slight production impairment or a
small but undefinable risk of moderate production impairment is unacceptable,
then continuous exposure conditions should use the no production impairment
values as means and the slight production impairment values as minima.
The criteria represent annual worst case dissolved oxygen concentrations
believed to protect the more sensitive populations of organisms against
potentially damaging production impairment. The dissolved oxygen concentra-
tions in
the criteria are intended to be protective at typically high seasonal
environmental temperatures for the appropriate taxonomic and life stage
classifications, temperatures which are often higher than those used in the
research from which the criteria were generated, especially for other than
early life stages. .
. 34
Where natural conditions alone create dissolved oxygen concentrations
less than 110 percent of the
,
applicable criteria means or minima or both, the
minimum acceptable concentration is 90 percent of the natural concentration.
These values are similar to those presented graphically by Doudoroff and
Shumway (1970) and those calculated from Water Quality Criteria 1972 (NAS/NAE,
1973). Absolutely no anthropogenic dissolved oxygen depression in the
potentially lethal area below the 1-day minima should be allowed unless
special care is
taken to ascertain the tolerance of resident species to low
dissolved
oxygen'.
If daily cycles of dissolved oxygen are essentially sinusoidal, a
reasonable daily average is calculated from the day's high and low dissolved
oxygen values. A time-weighted
.
average may be required if the .dissolved
oxygen cycles' are decidedly non-sinusoidal. Determining the magnitude of
daily dissolved oxygen cycles requires at least two appropriately timed
measurements daily, and characterizing the shape of the cycle requires several
more appropriately spaced measurements.
Once a series of daily mean dissolved oxygen concentrations are calcu-
lated, an average of these daily means can be calculated (Table 9). For
embryonic, larval, and early life stages, the averaging period should not
exceed 7'days. . This short time is needed to adequately protect these often
Table 9. Sample calculations' for determining daily means and 7-day mean
dissolved oxygen concentrations (30-day averages are calculated in a
similar fashion using 30 days data).
Dissolved Oxygen (mg/1)
Day?
Daily. Max.
Daily Mln.
Daily Mean
1?
9.0
7.0
8.0
2?
10.0
7.0
8.5
3? 11.0,
4? 12."0
5•?10.0
8.0
8.0
8.0
9 . 5
9.5
9.0
6
?
11.0,
7?
12.0"
9.0
10.0
10.0
10.5'
57.0'
65.0
1-day Minimum
7.0
7-day Mean Minimum
8.1•
7-day Mean
.
9.3
a
Above
?
air?
saturation?
concentration (assumed?
to be11.0mg/1
for?
this
example).
?
(11.0 + 8.0)
?
2.
?
c
(11.0 +10.0)
?
2.
35
short duration, most sensitive life stages. Other life stages can probably be
adequately protected by 30-day averages. Regardless of the averaging period,
the average should be considered a moving average rather than a calendar-week
or calendar-month average..
. The criteria have been established on the basis that the maximum dis-
solved oxygen value actually used in calculating any daily mean should not
exceed the air saturation value. This consideration is based primarily on
analysis of studies of cycling dissolved oxygen and the growth of largemouth
bass (Stewart et al., 1967), which indicated that high dissolved oxygen levels
(> 6 mg/1) had no beneficial effect on growth.
During periodic cycles of dissolved oxygen concentrations, minima lower
than acceptable constant exposure levels are tolerable so long as:
1.
the average concentration attained meets or exceeds the criterion;
2.
the average dissolved oxygen concentration is calculated as recommended
in Table 9; and
3.
the minima are not unduly stressful and clearly are not lethal.
A daily minimum has been included to make
.
tertain that no acute mortality
of sensitive species occurs as a result of lack of oxygen. Because repeated
exposure to ,dissolved oxygen concentrations at or near the acute lethal
threshold will be stressful and because stress can'indirect
•
y produce mortal-
ity or other adverse effects (e.g., through disease), the criteria are
designed to prevent significant episodes of continuous or regularly recurring
exposures to dissolved oxygen concentrations at or near the lethal•threshold.
This protection has been achieved by setting the daily minimum for early life
stages at the subacute lethality threshold, by the use of a 7-day averaging
period for early life stages, by stipulating a
.
7-day mean minimum value for
other life stages, and by recommending additional limits for manipulatable
discharges.
The previous EPA criterion for dissolved
•
oxygen published in Quality
Criteria 'for Water
(USEPA,
1976) was a minimum of 5 mg/1-(usually applied as a
7Q10) which fi—Tfinilar
to the current criterion minimum except for other life
stages of warmwater fish which now allows a 7-day mean minimum of 4 mg/l. The
new criteria are similar to those contained in the 1968 "Green Book" of the
Federal Water Pollution Control Federation (FWPCA, 1968).
A.
The Criteria and Monitoring and Design Conditions
The acceptable mean concentrations should be attained most of the time,
but some deviation below these values would probably not cause significant
harm. Deviations below the mean will probably be serially correlated and
hence apt to occur on consecutive days. The significance of deviations below
the mean will depend on whether they occur continuously or in daily cycles,
the former being more adverse than the latter. Current knowledge regarding
such deviations is limited primarily to laboratory growth. experiments and by
extrapolation to other activity-related phenomena.
• 36
Under conditions where large daily cycles of dissolved oxygen occur, it
is possible to meet the criteria mean values and consistently violate the mean
minimum criteria. Under these conditions the mean minimum criteria will
clearly be the limiting regulation unless alternatives such as nutrient
control can dampen the daily cycles.
The significance of conditions which fail to meet the recommended
dissolved oxygen criteria depend largely upon five factors: (1) the duration
of the event; (2) the magnitude of the dissolved oxygen depression; (3) the
frequency of recurrence; (4) the proportional area of the•site failing to meet
the criteria; and (5) the
biological significance of the site where the event
occurs. Evaluation of an event's significance must be largely case- and
site-specific. Common sense would dictate that the magnitude of the depres-
sion would be the single most important factor in general, especially if the
acute value is violated. A logical extension of these considerations is that
the event must be considered in the context of the level of resolution of the
monitoring or modeling effort. Evaluating the extent, duration,.and magnitude
of an event must be a function of the spatial and temporal frequency of the
data. Thus, a single deviation below the criterion takes on considerably
,
less
significance where continuous monitoring occurs than where sampling is
comprised of once-a-week grab samples. This is so because based on continuous
monitoring the event is provably small, but with the much less frequent
sampling the event is not provably small and can be considerably worse than
indicated by the sample.
The frequency of recurrence is of considerable interest to those modeling
dissolved oxygen concentrations because the return period, or period between
recurrences, is a primary modeling consideration contingent upon probabilities
of receiving water volumes, waste loads, temperatures, etc. It should be
apparent that return period cannot be isolated from-the other four factors
discussed above. Ultimately, the question of return period may be decided on
a site-specific basis taking into account the other factors (duration,
magnitude, areal extent, and biological significance) mentioned above. Future.
studies of temporal patterns of dissolved oxygen concentrations, both within
and between years, must be conducted to provide a better basis for selection
of the appropriate return period:
In conducting waste load allocation and treatment plant design computa-
tions, the choice of temperature in the models
will
be important. Probably.
the best option would be to use temperatures consistent with those expected in
the receiving water over the critical dissolved oxygen period for the biota.
B.?
The Criteria and Manipulatable Discharges
If daily minimum dissolved oxygen concentrations are perfectly serially
correlated, i.e., if the annual lowest daily minimum dissolved oxygen concen-
tration is adjacent in time to the next lower daily minimum dissolved oxygen
concentration and one of these two minima is adjacent to the third lowest
daily minimum dissolved oxygen concentration, etc., then in order to meet the
7-day mean minimum criterion it is unlikely that there will be more than three
or four consecutive daily minimum values below the acceptable 7-day mean
minimum. Unless the dissolved oxygen pattern is extremely erratic, it is also
unlikely that the lowest dissolved oxygen concentration will be appreciably
37
below the acceptable 7-day mean minimum or that daily minimum values below the
7-day mean minimum will occur in more than one or two weeks each year., For
some discharges, the distribution of dissolved oxygen concentrations can be
manipulated to varying degrees. Applying the daily minimum to manipulatable
discharges would allow repeated weekly cycles of minimum acutely acceptable
dissolved oxygen values, a condition of probable stress and possible adverse
biological effect. If risk of protection impairment is to be minimized, the
application of the one day minimum criterion to manipulatable discharges
should either limit the frequency of occurrence of values below the acceptable
7-day mean minimum or impose further limits on the extent of excursions below
the 7-day mean minimum. For such
. controlled discharges; it is recommended
that the occurrence of daily minima below the acceptable 7-day mean minimum be
limited to 3 weeks per year or that the acceptable one-day minimum be
increased to 4.5 mg/1 for coidwater fish and 3.5 mg/1 for warmwater fish.
Such decisions could be site-specific based upon the extent of control', serial
correlation, and the resource at risk.
38.
VIII. REFERENCES
Adelman, I. R., and L. L. Smith. 1970. Effect of oxygen on growth and food
conversion efficiency of northern pike. ?
Prog. Fish-Cult. 32:93-96.
Adelman, I. R., and L. L. Smith. 1972. Toxicity of hydrogen sulfide to
goldfish (Carassius auratus) as influenced by temperature, oxygen, and
bioassay tOTicTiiT7
777-71711.
Res. Bd. Canada 29:1309-1317.
Alabaster, J. S., and R. L. Welcomme. 1962. Effect of concentration of
dissolved oxygen on survival of trout and roach in lethal temperatures.
Nature 194:107.
Alabaster, J. S., and R. Lloyd. 1980. Water Quality Criteria for Freshwater
•?
Fish. Butterworths, London. 297 p.
Alderdice, D. F., W. P. Wickett, and J, R. Brett. 1958. Some effects of
temporary exposure to low dissolved oxygen levels on Pacific salmon eggs.
J. Fish. Res. Bd. Canada 15:229-250.
American Fisheries Society.
?
1978.?
Selected Coolwater• Fishes of North
America.?L. Kendall, Ed.
?
Special Publication No. 11, American
Fisheries Society, Washington, D.C. 437 p..
Andrews, J.
W.,
T. Murai, and G. Gibbons. 1973. The influence of dissolved
oxygen on the growth of channel catfish: Trans. Amer. Fish. Soc.
102: 835.
Bailey, R. M., J. E. Fitch, E. S. Herald, E. A: Lachner, C. C. Lindsey, C. R .
Robins, and W. B. Scott. 1970. A list of common and scientific names of
fishes from the
.
United States and Canada (third edition). American
Fisheries Society Special Publication No. 6. Washington, D.C.. 150 p.
Beamish, F. W. H. 1964. Respiration of fishes with special emphasis on
standard oxygen consumption. - III. Influence of oxygen. Can. J. Zool.
42:355-366.
Benedetto,?
1970.
?
Observations on the oxygen needs of some species of
European plecoptera. Int. Rev. Ges. Hydrobiol., 55:505-510..
Bishai, H. M. 1962. Reactions of larval and young salmonids to water of low
oxygen concentration. .J. Cons. Perm. Int. •Explor. Mer., 27:167-180.
Bouck, G. R. 1972. Effects of diurnal hypoxia on electrophoretic protein
fractions and other health parameters of rock bass (Ambloplites
rupestris). Trans. Amer. Fish. Soc. 101:448-493..
Bouck, G. R., and R. C. Ball.- 1965. Influence of a diurnal oxygen pulse on
fish •
serum proteins. Trans. Amer. Fish, Soc. 94:363-370.
Brake, L. A. 1972. Influence of dissolved oxygen and temperature on the
growth of a juvenile largemouth bass held in artificial ponds. Masters
Thesis. Oregon State University, Corvallis. 45 p.
39.
Brannon, E. L. 1965. The influence of physical factors on the development
and weight of sockeye salmon embryos and alevins. International Pacific
Salmon Fisheries Commission, Progress Report No. 12. New Westminster,
B.C., Canada. 26 p.
Brett, J. R., and J. M. Blackburn. 1981. Oxygen requirements for growth of
young coho salmon (Orconhynchus kisutch) and sockeye (0. nerka) salmon at
15°C. Can. J. Fish. Aquat. Sci.
38:399-404.
Brinley, F. J. 1944. Biological studies. House Document 266, 78th Congress,
1st Session; Part II, Supplement F. p. 1275-1353.
Brooke,
L. T., and P. J. Colby. 1980. Development and survival of embryos of
lake herring at different constant oxygen concentrations and tempera-
tures. Prog. Fish-Cult. 42:3-9.
Brungs, W. A. 1971. Chronic effects of low dissolved oxygen concentrations
on fathead minnow (Pimephales promelas). J. Fish. Res. Bd. Canada,
28:1119-1123.
Cairns, J., and A. Scheier. 1957. The effects of periodic low oxygen upon
the toxicity of various chemicals to aquatic organisms. Proc, 12th
Industrial Waste Conf., Purdue Univ. Eng. Bull. No. 94.
?
p.
165-176.
Cameron, J. N. 1971. Oxygen dissociation characteristics of the blood of
rainbow trout, Salmo gairdneri.
?
Comp. Biochem. Physiol. 38:699-704.
Carlson, A. R., J. Blocker, and L. J. Herman. 1980. Growth and survival Of
channel catfish and yellow perch exposed to lowered constant and
diurnally fluctuating dissolved oxygen concentrations. Prog. Fish-Cult.
42:73-78.
Carlson, A. R., and L. J.
Herman. • 1978. Effect of long-term reduction and
diel fluctuation in dissolved oxygen on spawning of black crappie,
Pomoxis niqromaculatus. Trans. Amer. Fish. Soc. 107:742-746.
?
.
Carlson, -A. R., and R. E.
Siefert, 1974. Effects of reduced oxygen on the
embryos and larvae of lake trout (Salvelinus namaycush) and largemouth
bass (Micropterus salmoides).?
Fish. Res. , Bd. Canada, 31:1393-1396.
Carlson, A. R., R. E. Siefert, 'and L. J. Herman. 1974. Effects of lowered
dissolved oxygen concentrations on channel catfish (Ictalurus punctatus)
embryos and larvae. Trans. Amer. Fish. Soc. 103:623-626.
Cech, J. J., Jr., C. G. Campagna, and S. J. Mitchell. 1979.. Respiratory
responses of largemouth bass (Micropterus salmoides) to environmental
changes in temperature and dissolved oxygen. Trans. Amer. Fish. Soc.
108:166-171. •
Chapman, G. A., and D. L. Shumway, 1978. Effects of .
sodium pentachloro-
phenate on
the survival and energy metabolism of larval steelhead trout.
pp. 285-299. In: K. .Ranga Rao, ed. Pentachlorophenol: chemistry,
pharmacology, and environmental toxicology. Proceedings of a symposium
held in Pensacola, Florida, June 27-29, 1977. Plenum Press, New York.
40
Eddy, R. M. 1972. The influence of dissolved oxygen concentration and
temperature on the survival and growth of chinook salmon embryos and fry.
M.S. Thesis, Oregon State University, Corvallis. 45 p.
Ellis, M. M. 1937. Detection and measurement of stream pollution. Bull.
U.S. Bureau of Sport Fisheries and Wildlife 48(22):365-437.
Ellis, M. M. 1944. Water purity standards for freshwater fishes. Special
Scientific Report No. 2, U.S. Department of Interior, Fish and Wildlife
Service.
Federal Water Pollution Control Administration. 1968. Water Quality
Criteria. Report of the National Technical Advisory Committee of the
Secretary of Interior. U.S.. Dept..of Interior, Washington, D.C. 234 p.
Fisher, R. J. 1963. Influence of
.
oxygen concentration and its diurnal
fluctuation on the growth
,
of juvehil i
e,.coho
salmon. M.S. Thesis, Oregon
State University, Corvallis. 48 p.
Fox, H. M., and J. Sidney. 1953. The influence of dissolved 'oxygen on • the
respiratory
movements of caddis larvae. J. Exptl. Biol., 30:235-237.
Fox, H. M., C. A. Wingfield, and B. G. Simmonds. 1937. The oxygen consump-
tion of ephemerid nymphs from flowing and from still waters in relation
to the concentration of oxygen
,
in :the water. J. Exptl. Biol., 14:210-
• 218.
Gaufin, A. R: 1973. Water quality requirements of aquatic insects. EPA-660/
3-73-004, September 1973. Ecological Research Series. U.S. Environ-
mental Protection Agency, Washington, D.C. 79 p.
Grigg, G, C. 1969.
The
failure of oxygen-transport in a fish at low levels
of ambient oxygen. Comp. Biochem. Physiol. 29:1253-1257.
Haley,
R., S. P. Davis, and J. M.
'
Hyde. 1967. Environmental stress and
Aeromonas
liquefascians
in American and threadfin shad mortalities.
TRTT-riih-Cult.
29:193.
Heimer,
J.
T. 1984. American Falls-Snake River fisheries investigations.
Final Report to Idaho Power Company from Idaho
Dept. of ,Fish and Game.
35 p.
Herrmann, R. B., C. E. Warren, and P. Doudoroff. 1962. Influence of,oxygen
concentration on the growth of
. juvenile coho salmon. Trans. Amer. Fish.
Soc. 91:155-167.
Hollender, B. A. 1981. Embryo survival, substrate composition, and dissolved
oxygen in redds of wild brook trout, M.S. Thesis, University of
Wisconsin, Stevens Point. 87 p.
Homer, D. H., and W. E. Waller. 1983. Chronic effects of reduced dissolved
oxygen on
Daphnia magna.
Water, Air, and Soil Pollut., 20:23-28.
42
Hutchins, F. E. 1974. Influence of dissolved oxygen concentration and
swimming velocity on food consumption and growth of juvenile coho salmon.
M.S. Thesis, Oregon State University, Corvallis. 66 p.
International Joint Commission. 1976. Dissolved oxygen. In: Great Lakes
Water Quality, Annual Report of the Water Quality Objectives Subcommittee
and the Task Force on the Scientific Basis for Water Quality Criteria.'
83 p.
Jacob, U., H. Walther, and R. Klenke. . 1984.?
Aquatic insect larvae as.
indicators of limiting minimal contents of dissolved oxygen. Aquatic
• Insects, 6:185-190.
Jones, D. R. 1971. The effect of hypoxia and anemia on the swimming perform-
ance of rainbow trout (Salmo gairdneri). J. Exptl. Biol. 44:541-551.
JRB Associates. 1984. Analysis of data relating dissolved oxygen and fish
growth. Report submitted to EPA under contract 68-01-6388 by JRB
Associates, McLean, Virginia.
Kapoor,
N. N.,
and W. Griffiths. 1975. Oxygen consumption of nymphs of
Phasganophora capitata (Pictet) (Plecoptera) with respect to body weight'
and oxygen concentrations. Can. J. Zool., 53:1089-1092.
Katz, M., A. Pritchard, and C. E. Warren. 1959. Ability of some salmonids
and a centrarchid to swim in water of reduced oxygen content. Trans.
Amer. Fish. Soc. 88:88-95.
Knight, A. W., and A. R. Gaufin. 1963. The effect of water flow, tempera-
ture, and oxygen concentration on the Plecoptera nymph, Acroneuria
pacifica Banks. Proc. Utah Acad. Sci., Arts, and Letters, 40(II):175-
184
Knight, A. W., and A. R. Gaufin. 1965. Function of stonefly gills under
reduced dissolved oxygen concentration. Proc. Utah Acad. Sci., Arts, and
Letters, 42(II):186-190.
Koski, K. V. 1965. The survival of coho salmon (Oncorhynchus kisutch) from
.egg deposition to emergency in three Oregon coastal streams. M.S.
Thesis, Oregon State University, Corvallis. 81 p.
Lloyd, R. 1961. Effect of-dissolved oxygen concentration on thetoxicity of
several poisons to rainbow trout (Salmo gairdnerii Richardson). J.
Exptl. Biol. 38:447-455.
Magnuson, J. J., P. O. Fromm, J. R. Brett, and F. E.
J.
Fry. 1979. Report of
the review committee for the dissolved oxygen objective for the Great
Lakes. A report submitted_to the Great Lakes Science Advisory Board,
International Joint Commission, Windsor, Ontario, Canada.
Meyer, F.-P. 1970. Seasonal fluctuations
in
the incidence of disease on fish
farms.. In: •A Symposium on Diseases of Fish and Shellfishes. Spec.
Publ. No. 5. Amer. Fish. Soc. Washington, D.C. p. 21-29.
43
Minnesota Pollution Control Agency. 1980. Dissolved oxygen standard justifi-
cation.?
MPCA, Water Quality Division. Unpublished manuscript.
?
35•p.
Moss,
D.
D., and D. C. Scott. 1961. Dissolved oxygen requirements of three
species of fish. Trans. Amer. Fish. Soc. 90:377-393.
National Academy of Sciences/National Academy of Engineering- 1973. Water
Quality Criteria. 1972. p. 131-135. EPA-R/73-033. 594 p.
Nebeker, A. V.
?
1972. :Effect of low oxygen concentration on survival and
emergence of aquatic insects.
?
Trans. Amer. Fish. Soc., 101:675-679.
Peterka,
J. J.,
and J.
S. Kent.
?
1976.?
Dissolved oxygen., temperature,,
survival of. young at fish spawning sites.
?
Environmental Protection
Agency Report No. EPA-600/3-76-113, Ecological Research Series. 36 p.
Pickering, Q. H. 1968. Some effects of dissolved oxygen concentrations upon
the toxicity of zinc to the bluegill, Lepomis macrochirus Raf. Water
Res. 2:187-194.
Raible, R. W. 1975. Survival and growth rate of channel catfish as a
function of dissolved oxygen concentration. Water Resources Research
Center, Arkansas University, PB 244 708, NTIS, Springfield, Virginia.
Shepard, M. P. 1955. Resistance and tolerance of young speckled trout
(Salvelinus fontinalis) to oxygen lack, with special reference to low
oxygen acc imation. J. Fish. Res. Bd. Canada, 12:387-446.
Shumway, D. L., C. E. Warren, and P. Doudoroff. 1964. Influence of oxygen
. concentration and water movement on the ,growth of steelhead trout and
coho salmon embryos. :Trans. Amer. Fish. Soc. 93:342-356.
Siefert, R. E., A. R. Carlson, and
L:
J,
Herman. 1974. Effects of reduced
oxygen concentrations on the early life stages of mountain whitefish,
smallmouth bass, and white bass'. Prog. Fish-Cult. 36:186-190.
Siefert, -R. E., and L. J. Herman. 1977. Spawning success of the black
crappie, Pomoxis nitromaculatus, at reduced dissolved oxygen concentra-
tions: Trans. Amer. Fish. Soc. 106:376-379.
Siefert, R. E., and
W.
A. Spoor. 1974. Effects of reduced oxygen on embryos
and larvae of the white sucker, coho salmon, brook trout, and walleye.
pp. 487-495. In: J. H. S. Blaxter, ed. The early life history of fish.
The proceedings of an international symposium, Oban, Scotland, May 17-23,
1973. Springer-Verlag, Berlin.
Siefert, R. E., W. A. Spoor, and R. F. Syrett. 1973. Effects of reduced
oxygen concentrations on northern pike (Esox lucius) embryos and larvae.
J. Fish. Res. Bd. Canada, 30:849-852.
Silver, S. J., C. E. Warren, and P. Doudoroff. 1963. Dissolved oxygen
requirements of developing steelhead trout and chinook salmon embryos at
different water velocities. Trans. Amer. Fish. Soc. 92:327-343.
44
Snieszko,
S. F.• 1974. The effects of environmental stress on outbreaks of
infectious diseases of fish. Fish. Biol. 6:197-208.
§owden, T. K., and G. Power. 1985. Prediction of 'rainbow trout embryo
survival in relation to groundwater seepage and particle size of spawning
substrates. Trans. Amer. Fish. Soc., 114:804-812.
Spoor, W.
A. 1977. Oxygen requirements of embryo and larvae of the large-
mouth bass, Micropterus salmoides (Lacepede). J. Fish. Biol. 11:77-86.
Spoor, W. A. 1981. Growth of trout at different oxygen. concentrations.
Preliminary report from -USEPA, Environmental Research Laboratory --
Duluth, Minnesota. 9 p.
Sprague, J. B. 1963. Resistance of four freshwater crustaceans to lethal
high temperatures and low oxygen. J. Fish. Res. Bd. Canada; 20:387-415.
Stewart, N. E., D. L. Shumway, and P. Doudoroff. 1967. 'Influence of oxygen
concentration .on the growth of juvenile largemouth bass. J. Fish. Res.
Bd. Canada, 24:475-494.
Thatcher, 'T. O. 1974. Some effects of dissolved oxygen concentration on
feeding, growth, and bioenergetics of juvenile coho salmon.. Ph.D.
Thesis. Oregon State University, Corvallis. 70 p.
Thurston, R. V., G. R. Phillips, R. C. Russo, and S. M. Hinkins.
?
1981.
Increased- toxicity of ammonia to rainbow
trout (Salmo gairdneri) result-
ing from reduced concentrations of dissolved oxygen.
?
Can. J.
Fish.
Aquat. Sci. 38:983-988.
U.S. Environmental Protection Agency.. 1976.
?
Quality Criteria for Water.
Washington, D.C. 256 p.
U.S. Environmental Protection Agency. 1982. Water Quality Standards RegUla-
tion. Federal Register 47:49239. October 29.
Warren, C. E., P. Doudoroff, and D. L. Shumway.
?
1973.?
Development
of
dissolved oxygen criteria for freshwater fish.
U.S.
Environmental
Protection Agency,
Ecological Research Series Report. EPA-R3-73-019.
'Washington, D.C. 121 p.
Wedemeyer, F. A. 1974. Stress as a predisposing factor in fish diseases.
U.S. Department of the Interior, Fish and Wildlife Service Leaflet
FDL-38. 8 p.
Weithman, A. S., and M. A. Haas. 1984.
Effects
of dissolved oxygen depletion
on the rainbow trout fishery in Lake Taneycomo, Missouri. Trans. Amer.
Fish. Soc. 113:109-124.
-Whitmore, C. M., C. E. Warren, and P. Doudoroff. 1960. Avoidance reactions
of salmonid and centrarchid fishes to low oxygen concentrations. Trans.
Amer. Fish. Soc.' 89:17-26.•
45
Whitworth, W. R. 1968. Effects of diurnal fluctuations of dissolved oxygen
on the growth of brook trout. J. Fish. Res. Bd. Canada, 25:579-584.
Witzel, L. D., and H.:R. McCrimmon. 1983. Redd-site selection by brook trout
and brown trout in
southwestern Ontario streams. Trans. Amer. Fish.
Soc., 112:760-771.