Full text of DOI:10.1002/etc.5620191214

Environmental Toxicology and Chemistry, Vol. 19, No. 12, pp. 2937–2942, 2000
2000 SETAC
Printed in the USA
0730-7268/00 $9.00 .00
᭧
ϩ
EFFECTS OF SHORT-TERM OXYGEN DEPLETION ON FISH
J
OHN
SEAGER, IAN MILNE,* MIKE MALLETT, and IAN SIMS
WRc plc, Henley Road, Medmenham, Marlow, Buckinghamshire SL7 2HD, United Kingdom
(
Received 26 May 1999; Accepted 25 April 2000)
Abstract—Laboratory experiments were undertaken to investigate the influences of exposure duration and frequency on the toxicity
of short-term pulses of low dissolved oxygen (DO) to fish. For the investigation of exposure duration, rainbow trout (Oncorhynchus
mykiss [Walbaum]) and roach (Rutilus rutilus L.) were exposed to a range of DO concentrations in single pulses of 1, 6, or 24 h.
For the investigation of exposure frequency, brown trout (Salmo trutta L.) were exposed to 24-h pulses of DO concentrations of
4.0 and 5.5 mg/L at frequencies of once or twice weekly over a period of 75 d. The results suggest that, for a given duration, there
is a narrow threshold concentration range above which mortality does not occur and below which mortality rapidly becomes high.
This threshold concentration range increases as exposure duration increases. Roach were able to survive lower DO concentrations
than trout. Observations on experimental animals following exposure indicated no significant postexposure effects, even at very
low DO levels. For the exposure frequencies used here, DO concentration rather than frequency of exposure was the important
factor in terms of effect on fish. No significant effects on growth rate were observed but there were significant differences in
hemoglobin levels, hematocrit, and organ weights. These results have important implications for the derivation of environmental
quality standards aimed at the control of episodic pollution in rivers.
Keywords—Dissolved oxygen
Short-term exposure
Toxicity tests
Intermittent pollution
Environmental quality
standards
INTRODUCTION
lution in rivers. The results should be seen as preliminary,
pending further validation and testing of other species.
Dissolved oxygen depletion is a common problem in rivers
polluted by the discharge of organic wastes. As a consequence,
numerous studies have been carried out to establish how re-
duced dissolved oxygen (DO) concentrations affect aquatic
organisms and to derive appropriate water quality standards.
This work as been extensively reviewed [1–3]. Most reported
studies have concentrated on continuous exposure to low DO
over periods from days up to several weeks. While some pol-
luted rivers suffer prolonged oxygen depletion, particularly
during summer months, many others experience only transient
DO depletion caused by pollution episodes that may last only
a few hours. This has been shown to be the case, e.g., in rivers
affected by runoff of farm wastes [4] and those receiving rain-
fall-generated waste discharges from combined sewer over-
flows [5]. Although short-term, these transient events may ex-
ert a significant influence on fish and other aquatic life.
The study described in this paper was initiated to address
the question of how fish respond to short-term episodes of low
DO of differing duration and frequency. There has been some
previous work on short-term exposure, most notably during
the 1950s [6–8]. These earlier studies were primarily con-
cerned with lethal effects, whereas the focus of the present
study is defining conditions that will avoid lethality and long-
term detrimental effects. The series of experiments reported
here was conducted as part of a major research program to
investigate the effects of exposure to pollutant pulses of vary-
ing magnitude, duration, and frequency. The objective of this
program was to apply the ecotoxicological information derived
from these experiments to the derivation of environmental
quality standards relevant to the control of intermittent pol-
MATERIALS AND METHODS
Test organisms
Three species of fish were used, rainbow trout (Oncorhyn-
chus mykiss [Walbaum]), roach (Rutilus rutilus L.), and brown
trout (Salmo trutta L.). The rainbow trout were obtained from
Warnford trout farm (West Meon, Hampshire, UK) and when
tested were three months old. The roach were obtained from
Anglian Water Authority (Norwich Division) Fish Farm (Nor-
wich, Norfolk, UK) and when tested were approximately one
year old. The brown trout were obtained from Berkshire Trout
Farm (Hungerford, Berkshire, UK). Prior to testing, the fish
were held in large tanks with a groundwater flow giving a
replacement rate of between 10 and 24 volumes per day. The
fish were fed pelleted trout food at 1 to 1.5% wet body weight
per day for the trout and 2% for the roach.
Test procedure—Exposure duration
For each test, fish were placed in each tank (25 rainbow
trout or 20 roach per tank) and allowed to acclimate for at
least 2 h. Six tests were carried out separately, three for each
species, with exposure durations of 1, 6, and 24 h. The test
vessels were polypropylene tanks with clear plastic lids sealed
with a rubber strip and clamped down. Each tank was filled
with 150 L of groundwater, giving a 2-cm closed air space
between the water surface and lid. Each test had one control
tank, and the number of experimental tanks was four for tests
with rainbow trout and three for tests with roach, each tank
receiving a different DO concentration. The DO concentrations
were reduced and maintained at the desired levels by aeration
with a mixture of nitrogen, air, and carbon dioxide to achieve
as close to square-shape pulses as possible. The control tank
* To whom correspondence may be addressed
(ian.milne@sepa.org.uk).
2937

2938
Environ. Toxicol. Chem. 19, 2000
J. Seager et al.
Table 1. Experimental design for exposure frequency experiments
Target dissolved oxygen
were weighed every two weeks and the feeding rate adjusted
accordingly.
The low DO exposures lasted 24 h, and the procedure for
reducing DO concentration was the same as for the exposure
duration experiments. Each vessel received a different treat-
ment, as shown in Table 1. The test lasted 75 d, with vessels
2 and 3 receiving 10 low DO events and vessels 4 and 5
receiving 20 events.
Vessel No.
(mg/L)
Frequency/week
1 (control)
Air saturation
—
1
1
2
2
2
3
4
5
4.0
5.5
4.0
5.5
Water quality was measured throughout experiments. The
pH ranged from 7.28 to 8.53 and temperature from 12.1 to
14.3ЊC. Hardness was measured twice during the test and
received air and carbon dioxide only. At the end of the ex-
posure period, the nitrogen supply was stopped and the air
supply increased to bring the DO concentration back to air
saturation value. This took approximately 1 h. During the test,
the DO concentration in each tank was monitored continuously
by DO meters connected to a chart recorder. The tanks were
held in a constant ambient air temperature between 12 and
ranged from 272 to 290 mg CaCO3/L. Only slight differences
between test vessels in terms of water quality were observed.
For half the fish in each tank, the measurements made at
the end of the experiment were hematocrit (according to the
method of Blaxhall & Daisley [9]), hemoglobin (absorbance
at 540 nm [9]), and spleen, kidney, and liver weights. Two-
way analyses of variance were performed on the data to test
for the effects of DO concentration and frequency of exposure.
Gills were dissected out and sent to the National Rivers Au-
thority Fish Diseases Laboratory (Brampton, Cambridgeshire,
UK) for measurement of surface area and assessment of con-
dition.
The remaining fish were subjected to an acute low DO
exposure at 1.5 mg/L. The survival time for each fish was
recorded and median period of survival was calculated for each
group. Differences between groups were examined using the
method of Litchfield [10].
Before the acute exposure, the control fish and those from
vessel 4 (worst case treatment) were placed in a WRc Mark
III Fish Monitor (WRc, Medmenham, Buckinghamshire, UK)
[11] and their ventilatory responses to a low DO pulse mon-
itored. Two runs were carried out, each with four fish from
each group in the eight-channel monitor. After a 24-h accli-
mation period, the ventilation rates were monitored for a 30-
min period, DO was reduced to 4 mg/L, and ventilation mon-
itored for a further 30 min.
13
the roach tests, with constant low intensity illumination.
The physicochemical variables measured were (with total
range over all tests in parentheses) temperature (12.1–14.8 C),
ЊC for the rainbow trout tests and between 13 and 15Њ for
Њ
pH (7.37–8.44), and total hardness (150–288 mg/L CaCO3).
For each test, there was very little variation in these variables
between tanks.
For the 1- and 6-h exposure tests, the fish were observed
continuously for mortality and sublethal effects. For the 24-h
exposure tests, observation was continuous for the first 11 h,
no observations were made during the next 5.5 h, and obser-
vations were then made hourly to the end of the exposure
period. The fish were held in the tanks for a further 4 d for
1- and 6-h exposure tests, 6 d for 24-h exposure tests, and
observed at least once per day for any postexposure mortality.
Test procedure—Exposure frequency
The exposure frequency experiment used brown trout. (The
use of a different species arose from a request by the project
sponsors following the exposure duration test that UK native
species be used.) Fish were selected from stock, weighed
(mean wet wt 11.2 g), and randomly distributed between five
120-L polythene test vessels, 16 fish per vessel. During the
experiment, a groundwater flow of 1 L/min was maintained
to each tank, but this was stopped during the low DO exposure
periods. The fish were fed at 4% body weight per day and
RESULTS
Exposure duration
Oxygen concentrations during exposure. The mean, min-
imum, and maximum DO concentrations during the exposure
periods are given in Table 2. Dissolved oxygen concentrations
Table 2. Dissolved oxygen concentrations during exposure periods of exposure duration experiment
Dissolved oxygen concentration (mg/L)
Vessel 1
Exposure period (h)
(control)
Vessel 2
Vessel 3
Vessel 4
Vessel 5
Rainbow trout
1
Mean
Range
Mean
Range
Mean
Range
10.3
10.1–10.4
10.7
10.5–10.9
10.4
3.5
3.1–3.8
3.2
2.8–3.5
4.5
2.5
2.5–2.6
2.7
2.4–3.0
3.5
1.5
1.5–1.7
1.6
1.3–2.1
2.5
0.7
0.5–0.9
1.2
0.9–1.5
1.6
6
24
10.1–11.0
4.2–4.9
3.1–3.9
2.0–3.0
1.0–2.2
Roach
1
Mean
Range
Mean
Range
Mean
Range
10.5
10.3–10.7
10.6
10.5–10.7
10.6
2.4
2.3–2.7
2.5
2.3–2.8
2.2
1.6
1.6–1.8
1.5
1.3–1.8
1.3
0.4
0.3–0.7
0.4
0.2–0.6
0.3
—
—
—
—
—
—
6
24
10.4–10.9
2.0–2.4
1.1–1.6
0.1–1.5

Short-term oxygen depletion effects on fish
Environ. Toxicol. Chem. 19, 2000
2939
Fig. 2. Mortality of roach exposed to different concentrations of dis-
solved oxygen for 24 h. ***
0.1%.
ϭ significantly different from control at
significant differences were found for the lowest concentration
at 1-h exposure and the lowest two concentrations at 6- and
24-h exposure. The results suggest there is a narrow threshold
concentration range above which mortality does not occur and
below which mortality rapidly becomes high. From Figure 1,
it can be seen that the threshold concentration increases as
exposure duration increases; mortality was significant at 1.6
mg/L for 6-h exposure but not at 1.5 mg/L for 1-h exposure.
Practical constraints precluded the use of replicate vessels in
the tests. An indication of repeatability can be gained from
comparing the results for different exposure times. Comparison
is only possible for 1.6 mg DO/L exposure. There did appear
to be some differences in sensitivity; in particular, the onset
of mortality was more rapid in the 24-h exposed fish than in
the 1- and 6-h exposed fish.
Roach. Mortality of roach was minimal. One fish died during
1-h exposure at 0.4 mg/L and no fish died during the 6-h
exposure test. Mortality during the 24-h exposure test, plotted
in Figure 2, was restricted to the lowest concentration (mean
concentration 0.3 mg/L). The first mortality was seen after 7.5
h, and by 24 h, 16 out of the 20 fish had died.
Fig. 1. Mortality of rainbow trout exposed to different concentrations
of dissolved oxygen for 1, 6, and 24 h. *
from control at 5%; *** significantly different from control at 0.1%.
ϭ significantly different
ϭ
Note that symbols sometimes overlap.
were successfully maintained within a narrow range during ex-
posure, as indicated by the measured ranges shown in Table 2.
Rainbow trout. Mortality curves for rainbow trout are plot-
ted in Figure 1. Mortality was generally restricted to the ex-
posure period; there was no evidence of significant postex-
posure mortality. The indicated exposure periods in Figure 1
represent the period during which the target concentration was
maintained. For the lowest DO concentrations, some mortal-
ities occurred before this period, i.e., during the period of
lowering the DO concentration.
Exposure frequency
Oxygen concentrations during exposure. Dissolved oxygen
concentrations achieved during the 24-h experiment pulses are
summarized in Table 3, showing that they were close to the
target levels.
Survival. No mortality was observed during the experiment.
All fish appeared to be in good external condition at the end
of the test.
For each of the three experiments, a Fisher exact probability
test was carried out to identify those exposure regimes where
mortality differed significantly from the control. Statistically
Growth. The mean weights of fish from each treatment, at
days 0 and 75, are shown in Figure 3 and mean lengths are
shown in Figure 4. Analyses of variance indicated no signif-
Table 3. Measured dissolved oxygen concentrations during 24-h experimental pulses
(exposure frequency experiments)
Dissolved oxygen concentration (mg/L)
Vessel 2
(4.0 mg/L,
1/week)
Vessel 3
(5.5 mg/L,
1/week)
Vessel 4
(4.0 mg/L,
2/week)
Vessel 5
(5.5 mg/L,
2/week)
Vessel 1
(control)
Mean
Minimum
Maximum
9.86
8.80
10.70
4.07
3.40
5.70
5.32
4.70
5.80
4.01
3.50
5.80
5.30
4.70
5.90

2940
Environ. Toxicol. Chem. 19, 2000
J. Seager et al.
Fig. 3. Mean weights (with standard deviation) of brown trout before
and after intermittent exposure to low dissolved oxygen (DO; see text
for exposure regimes for each vessel).
icant differences between control and experimental groups in
either body weight or length.
Physiological indices. The results of the physiological mea-
surements are shown in Figure 5. One-way analyses of vari-
ance indicated significant differences between groups for
spleen, liver, and kidney weights (as percentage of body
weight) and for hemoglobin but not for hematocrit. The Dun-
nett’s test showed that, for spleen and kidney weights and
hemoglobin, some of the experimental groups differed signif-
icantly from the control. Inspection of Figure 5 reveals that
spleen weight tended to increase as a percentage of body
weight in the experimental groups, with the two groups ex-
posed at 4 mg/L being significantly elevated in comparison
with the control. In contrast, kidney weight tended to decrease
as a percentage of body weight, and here it was the two groups
exposed to 5.5 mg/L that showed significant differences from
the control. For hemoglobin, there was a tendency to elevation
in the experimental groups, all of which were significantly
greater than the control except the group exposed to 4 mg/L
twice per week.
Fig. 5. Organ weights, as percentage of body weight, and blood pa-
rameters for brown trout after intermittent exposure to low dissolved
oxygen (DO). Values are means with standard deviation. *
icantly different from control at 5%.
ϭ signif-
Gill ventilation rate response to low DO. Mean gill ven-
tilation rates are shown in Table 4. Both groups (control and
worst case exposure) showed almost identical ventilation rates
before exposure and identical responses to exposure in the
form of an increase in ventilation rate of about 20%. The
response was almost immediate. Thus, there was no evidence
of any acclimation, in terms of respiratory efficiency, having
occurred.
Acute exposure. The median period of survival for each
treatment is plotted in Figure 6. There was no evidence that
the low DO exposed fish were more resistant; in fact, the
control fish showed a slightly longer survival period, though
Two-way analyses of variance were also performed on the
experimental data alone (i.e., excluding the controls) to de-
termine the effects of exposure concentration and frequency.
These indicated that, for each of the five variables, there was
a significant effect of DO concentration but not of frequency
of exposure
Gill histopathology. Examination of gill tissues revealed a
number of pathological effects, including hyperplasia, hyper-
trophy, and necrosis, but there was no evidence of any dose–
response effects.
this was only significantly longer (
the 4 mg/L once per week group.
p Ͻ 0.05) than fish from
DISCUSSION
There are few published studies on the effects of short-term
low DO exposure on fish. Graham [12] and Shepard [6] both
investigated the effects of short-term exposure on speckled
trout (Salvelinus fontinalis). Graham [12] subjected speckled
trout to different low DO concentrations at different temper-
atures and found that lethal low DO concentration was tem-
perature dependent, with fish surviving lower DO concentra-
tions at low temperature. However, only one fish was used for
each DO concentration. Shepard [6] performed similar exper-
Table 4. Mean (with 95% confidence limits) ventilation frequencies
of fish before and during a 4 mg/L low dissolved oxygen; CI
confidence interval
ϭ
Mean ventilation frequency
(
Ϯ
95% CI)
Difference
between
Pre-exposure Exposure
Fish group
ventilation (Hz) ventilation (Hz) means
Control (
4 mg/L 2/week (
n
ϭ
8)a
1.66 (
7) 1.69 (
Ϯ
Ϯ
0.17) 2.03 (
0.24) 2.06 (
Ϯ
Ϯ
0.09)
0.27)
0.38
0.37
Fig. 4. Mean length (with standard deviation) of brown trout before
and after intermittent exposure to low dissolved oxygen (DO; see text
for exposure regimes for each vessel).
n
ϭ
a n
ϭ number of fish tested.

Short-term oxygen depletion effects on fish
Environ. Toxicol. Chem. 19, 2000
2941
were separated by at most 1 mg/L. For roach, the evidence
was limited due to lack of mortality, but the indication is for
a similar pattern.
These observations suggest there is a critical threshold con-
centration needed to support essential metabolic processes. The
actual threshold concentrations separating survival and mor-
tality in the present study agree well with those found by
Downing and Merkens [7], as illustrated in Table 5.
The present study supports the findings of Shepard [6] and
Downing and Merkens [7] that the critical threshold concen-
tration depends on exposure duration. For very short-term ex-
posure (minutes to hours), fish can survive low DO concen-
trations that would be lethal over longer periods.
In all the above reported studies, the experimental approach
was to place fish in low DO concentrations and record when
they die. Such a method does not allow any assessment of
postexposure mortality. The results of the present study, which
more closely reproduced a short duration exposure, indicate
that there is little or no postexposure effect, even for exposure
to very low DO concentration for short periods.
Fig. 6. Median periods of survival with 95% confidence limits for
brown trout after acute exposure to a low dissolved oxygen (DO)
episode.
For the investigation of exposure frequency in this study,
even the most severe combination of 24-h pulses of 4.0 mg
DO/L applied twice weekly had no significant effect on fish
growth over the 75-d period. The results of other relevant
studies vary widely between tests and between species. The
U.S. Environmental Protection Agency [14] carried out an
extensive literature review and concluded that production im-
pairment for salmonids was reported to range from slight at 6
and 9 mg DO/L to severe at 4 and 7 mg DO/L for adults and
early life stages, respectively. Whitworth [15] found that diel
DO fluctuations significantly reduced growth of brook trout.
The experimental regime was 14 h at DO concentrations re-
duced from 11 mg/L to 5.3, 3.6, 3.5, or 2.0 mg/L. At 2.0 mg/
L, few survived, but at the other concentrations, growth was
severely reduced with no significant difference between
groups. However, Alabaster and Lloyd [3] stated that neither
reduced DO to 5 mg/L nor wide diel fluctuations about this
concentration has much effect on growth of salmonid alevins
and there may be only slight growth reduction at 3 mg/L.
Although growth was not significantly impaired as a con-
sequence of the low DO exposures applied in the present study,
a number of physiological effects were observed that were
attributable more to DO concentration than exposure frequen-
cy. However, although some of the effects were statistically
iments, but with groups of fish, and found that median survival
time was closely correlated with DO concentration and ranged
from about 10 min at 0.9 mg/L to 600 min at 1.7 mg/L.
Allan et al. [8] investigated the effects of sewage effluent
on rainbow trout. They found that tolerance to low DO de-
creased as the concentration of either carbon dioxide or am-
monia increased. Downing and Merkens [7] exposed eight
different fish species (rainbow trout, perch, roach, mirror carp,
tench, dace, bleak, and chub) to different DO concentrations
and recorded mortality after different time periods. They found
rainbow trout to be considerably less tolerant than other spe-
cies. Over a 3.5-h period, the maximum DO concentration
causing 100% mortality and that allowing 100% survival were
generally both lower than for a 3.5-d period, though for 3.5
and 7 d, there was generally little difference. The difference
between the concentration causing 100% mortality and that
allowing 100% survival was generally small, at around 0.5
mg/L. Similarly, Magaud [13] found that rainbow trout showed
100% mortality at 1.7 mg/L over 2.5 h but 0% mortality at
2.3 mg/L. This is supported by the findings in the present
study, where high survival and high mortality for rainbow trout
Table 5. Dissolved oxygen concentrations separating high mortality and high survival
Dissolved oxygen concentration (mg/L)
Downing and Merkens [7]a
Present study
Temperature 12–15ЊC
Temperature
ϭ
10
Њ
C
Temperature
ϭ
16
Њ
C
ϭ
Species
Duration
100% mortality 100% survival 100% mortality 100% survival
Ͼ
95% mortality
Ͼ
95% survival
Rainbow trout
1 h
3.5 h
6
24 h
3.5 d
1 h
3.5 h
6 h
24 h
3.5 d
—
1.2
—
—
1.3
—
0.2
—
—
—
1.7
—
—
1.9
—
0.4
—
—
—
1.5
—
—
2.4
—
0.3
—
—
—
1.9
—
—
3.0
—
0.6
—
—
0.7
—
1.6
1.6
—
1.5
—
2.7
3.5
—
0.4
—
0.4
1.3
—
Roach
Ͻ
0.4
—
Ͻ
Ͻ
0.4
0.3
0.3
0.6
0.7
0.7
—
a Dissolved oxygen concentrations calculated from reported percent saturations.

2942
Environ. Toxicol. Chem. 19, 2000
J. Seager et al.
significant, they were relatively small, and their significance
for fish health is not clear. The hematocrit and hemoglobin
levels found were within the ranges considered normal for
farmed trout by Blaxhall and Daisley [9].
termittent pollution. Standards aimed at avoiding long-term
effects have been derived, and these are used as design criteria
in the upgrading of storm overflows from sewer systems.
The results of this study present important new information
on the sublethal effects of exposure duration and frequency.
However, there is a need for further validation of the findings,
in particular, assessment of the replicability and repeatability
of the results, and comparative assessment of responses of
different species would be valuable.
A wide range of physiological effects has been reported to
be associated with low DO concentrations, including circu-
latory changes, altered heart rate, reduced blood oxygen sat-
uration, changes in respiratory quotient, and increases in
breathing rate and amplitude. Davis [2] calculated a mean
incipient threshold for physiological effects in freshwater sal-
monids of 6 mg DO/L on the basis of a review of reported
studies. It is assumed that, below this threshold DO level, fish
will be expending excess energy to maintain homeostasis and
that some degree of physiological stress is occurring. Alabaster
and Lloyd [3] suggested, however, that respiratory and car-
diovascular responses to DO changes were not necessarily
indicative of impairment of ecologically important functions.
This would appear to be the case in the present study for the
brown trout exposed to intermittent low DO pulses where,
although there appeared to be some concentration-related
changes in certain physiological parameters such as hemoglo-
bin levels, there were no significant effects on growth and
overall condition.
Acclimation of trout to repeated low DO exposures was
investigated by measuring both gill ventilation rates during
sublethal DO pulses and mortality during acute lethal expo-
sures following the 75-d, repeat-exposure experimental period.
No significant acclimation of fish was observed in these tests,
although certain physiological changes such as increased blood
hemoglobin levels in experimental fish may indicate some de-
gree of physiological adaptation. Acclimation to low DO has
been the subject of many studies, although there has been much
debate about the nature and importance of adaptive mecha-
nisms. Shepard [6] demonstrated that acclimated brook trout
were able to resist a lethal level of hypoxia longer than control
fish. Smith and Heath [16], however, showed that prior accli-
mation decreased resistance to anoxia in rainbow trout. The
severity of hypoxia to which fish are acclimated is likely to
be an important factor here.
The findings of this study have important implications for
the development of environmental quality standards for inter-
mittent pollution. It is clear from the results of this study that
the duration of low DO episodes is a critical factor affecting
the survival of fish during and following pollution events.
Time-varying standards are therefore required to incorporate
this dimension. A preliminary approach to the development
of standards that incorporate exposure duration has been pro-
posed by Whitelaw and Solbe´ [17]. These standards were based
on available toxicity data reported in the literature from which
short-term 50% lethal concentration values over periods of up
to 1,000 minutes could be derived. The drawback of this ap-
proach is that the standards were based on DO concentrations
that were lethal to fish, giving rise to uncertainties over whether
the margin of safety was sufficient to adequately protect fish
populations. The results presented here allow the determina-
tion of minimum DO concentrations that result in no mortality
for a given exposure period. This has provided an improved
basis for developing environmental quality standards for in-
Acknowledgement—This work was funded by the National Rivers
Authority. We thank the United Kingdom Urban Pollution Manage-
ment Steering Group.
REFERENCES
1. Doudoroff P, Shumway DL. 1970. Dissolved oxygen requirements
of freshwater fish. FAO Technical Paper 86. Food Agricultural
Organization of the United Nations, Rome, Italy.
2. Davis JC. 1975. Minimal dissolved oxygen requirements of aquat-
ic life with emphasis on Canadian species: A review. J Fish Res
Board Can 32:2295–2332.
3. Alabaster JS, Lloyd R. 1982. Water Quality Criteria for Fresh-
water Fish, 2nd ed. Butterworths, London, UK.
4. Schofield K, Seager J, Merriman RP. 1990. The impact of inten-
sive dairy farming activities on river quality: The Eastern Cleddau
Catchment Study. J Inst Water Environ Manage 4:176–186.
5. Seager J, Maltby L. 1989. Assessing the impact of episodic pol-
lution. Hydrobiologia 188/189:633–640.
6. Shepard MP. 1955. Resistance and tolerance of young speckled
trout (Salvenlinus fontinalis) to oxygen lack, with special refer-
ence to low oxygen acclimation. J Fish Res Board Can 12:387–
446.
7. Downing KM, Merkens JC. 1957. The influence of temperature
on the survival of several species of fish in low tensions of dis-
solved oxygen. Ann Appl Biol 45:261–267.
8. Allan IRH, Herbert DWM, Alabaster JS. 1958. A field and lab-
oratory investigation of fish in a sewage effluent. MAFF Fishery
Investigations, Series I, Vol 6, No 2. Ministry of Agriculture,
Fisheries, and Food, London, UK.
9. Blaxhall PC, Daisley KW. 1973. Routine haematological methods
for use with fish blood. J Fish Biol 5:771–781.
10. Litchfield JT Jr. 1949. A method for rapid graphic solution of
time percent effect curves. J Pharmacol Exp Ther 96:99–113.
11. Evans GP, Johnson D, Withell C. 1986. Development of the WRc
MK III Fish Monitor: Descriptions of the system and its response
to some commonly encountered pollutants. WRc Technical Report
TR233. Water Research Centre, Medmenham, Buckinghamshire,
UK.
12. Graham JM. 1949. Some effects of temperature and oxygen pres-
sure on the metabolism and activity of the speckled trout, Sal-
velinus fontinalis. Can J Res D 27:270–288.
13. Magaud H. 1993. Effet le´tal d’une hypoxie en pre´sence
d’ammoniaque sur les truitelles arc-en-ciel. Rapport technique.
Cemagref, Lyon, France, p 31.
14. U.S. Environmental Protection Agency. 1986. Ambient water
quality criteria for dissolved oxygen. EPA 440/5-86-003. National
Technical Information Service, Springfield, VA.
15. Whitworth WR. 1968. Effects of diurnal fluctuations of dissolved
oxygen on the growth of brook trout. J Fish Res Board Can 25:
579–584.
16. Smith MJ, Heath AG. 1980. Responses to acute anoxia and pro-
longed hypoxia by rainbow trout (Salmo gairdneri) and mirror
carp (Cyprinus carpio) red and white muscle: Use of conventional
and modified metabolic pathways. Comp Biochem Physiol B 66:
267.
17. Whitelaw K, Solbe´ JF de LG. 1989. River catchment manage-
ment: An approach to the derivation of quality standards for farm
pollution and storm sewage discharges. Water Sci Technol 21:
1065–1076.