EPA-600/3-76-050
May 1976
MORTALITY, SALTWATER ADAPTATION AND REPRODUCTION
OF FISH DURING GAS SUPERSATURATION
by
Gerald R. Bouck
Allen V. Nebeker
Donald G. Stevens
Western Fish Toxicology Station*
Environmental Research Laboratory-Duluth
Corvallis, Oregon 97330
(*Western Fish Toxicology Station is now attached
to the Corvallis Environmental Research Laboratory,
Corvallis, Oregon 97330)
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
ENVIRONMENTAL RESEARCH LABORATORY
DULUTH, MINNESOTA 55804
-------
EPA-600/3-76-050
May 1976
MORTALITY, SALTWATER ADAPTATION AND REPRODUCTION
OF FISH DURING GAS SUPERSATURATION
by
Gerald R. Bouck
Allen V. Nebeker
Donald G. Stevens
Western Fish Toxicology Station*
Environmental Research Laboratory-Duluth
Corvallis, Oregon 97330
(*Western Fish Toxicology Station is now attached
to the Corvallis Environmental Research Laboratory,
Con/all is, Oregon 97330)
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
ENVIRONMENTAL RESEARCH LABORATORY
DULUTH, MINNESOTA
-------
DISCLAIMER
This report has been reviewed by the Environmental Research
Laboratory-Duluth, U.S. Environmental Protection Agency, and approved
for publication. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.
ii
-------
ABSTRACT
Tests were conducted using continuous exposure in shallow water
at levels of total dissolved gas pressure ranging from 110-140% of
barometric pressure (hyperbaric pressure = 103-410 g/cm^). Both times
to 20% and to median mortality were determined on several life stages
of Pacific salmonids (Oncorhynchus and Salmo) and largemouth bass
(Micropterus salmoidesj^Mean times to^0% mortality at 115% total
gas saturation were 309, 154, and 125 hours for adults, smolts and
parr. At 120% saturation mean times to 201 mortality were 48, 41,
and 53 hours for adults, smolts, and parr. At 125% saturation, mean
times to 20% mortality decreased to 18, 17, and 24 hours for adults,
smolts, and parr. Factors which influence time to death included
genera, life stage, acclimation temperature, activity level, sex,
and body size. Mortality curves were typically skewed to the right.
Gross pathology of gas bubble disease was described relative to these
experiments. High gas levels that killed 50% of three species of
salmon smolts had no apparent effect on the ability of survivors to
tolerate an immediate transfer into seawater (30 ppt Cl). Long-term
(3-month) continuous exposure of adult spring chinook salmon to 110%
saturation had no readily apparent adverse impact on the fertilization
and hatching of their eggs.
-------
CONTENTS
Page
Abstract 111
List of Figures vii
List of Tables viil
Acknowledgments 1x
Section
I INTRODUCTION 1
II CONCLUSIONS AND RECOMMENDATIONS 4
III METHODS AND MATERIALS 6
Supersaturation Exposure Facilities 6
Water Quality 6
Biological Methods 15
Design of Experiments 15
Statistical Procedures 18
IV RESULTS 19
General Tolerances of Salmonids and Non-Salmonids 19
Sample Responses to Supersaturation 19
Pathobiology and Times to Mortality 26
Tolerance to 130% Saturation 26
Tolerance to 125% Saturation 30
Tolerance to 120% Saturation 30
Tolerance to 115% Saturation 31
Tolerance to 110% Saturation 32
Estimated Times to 20 Percent Mortality 32
Rank Order of Tolerance to Supersaturation 35
-------
Page
Factors Affecting Mortality by Supersaturation 35
Test Temperature 35
Swimming Activity 39
Sex 39
Life Stage 42
Body Size and Condition Coefficient 42
Other Factors Influencing Tolerance to 45
Supersaturation
Tolerance to Seawater After Exposure to 45
Supersaturated Water
Effects of Chronic Supersaturation on Reproduction 47
V DISCUSSION 48
VI REFERENCES 52
VI
-------
FIGURES
No. Page
1 Diagram of facilities for exposing fish to 7
supersaturated water
2 Supersaturation exposure facilities 8
3 Diagram of modified Weiss Saturometer 14
4 Time to death of fasting winter steel head trout parr 23
(Salmo gairdneri) at 125% saturation and 10 C in water
65 cm deep
5 Mortality response of rainbow trout (Salmo gairdneri) 24
to 120% saturation
6 Means and 90% confidence belts of times to 20% mortality 33
at various life stages of salmonids and levels of super-
saturation
7 Effects of test temperature on the time to median 37
mortality from supersaturation
8 Effects of swimming activity on the mortality of 40
largemouth bass (Micropterus salmoides) in water
supersaturated to 140%
9 Effect of percent saturation and sex on mortality of 41
adult coho salmon (Oncorhynchus kisutch) at 10 C in
tanks'65 cm deep
10 Effect of percent saturation and life stage on mortality 43
of coho salmon (Oncorhynchus kisutch) at 10 C in tanks
65 cm deep
11 Cumulative mortality of salmonid smolts during exposure 46
to 120% air supersaturation and immediate transition
to 30 o/oo seawater
Vll
-------
TABLES
No. Page
1 Examples of natural and man-made sources of 2
supersaturated water
2 A summary of chemical and physical parameters in the 10
water supply of Western Fish Toxicology Station 1972-1973
3 Nominal and measured test levels of total gas saturation 12
4 Conversion table and equivalent units for expressing 16
gas saturation data
5 Fish species tested 17
6 Summary statistics of time to death at three levels of 20
supersaturafion by twenty population samples of rainbow
trout, Pacific salmon, and largemouth bass
7 Hours of exposure to reach 20% mortality among fishes 27
exposed to supersaturation in shallow water at 10 C
8 Estimated hours to median mortality (ETsg) among fishes 29
exposed to supersaturation in shallow water at 10 C
9 Statistical analysis of pooled observations on hours to 34
20% mortality among three life stages of five species of
salmonids
10 Rank order of tolerance between fish tested at 120% total 36
gas saturation (203 g/cm2 hyperbaric)
11 Summary of the relationship between acclimation temperature 38
and time to median mortality in supersaturated waters
12 Correlation analyses of time to death at various levels 44
of supersaturation to body length, weight, and condition
coefficient (K) for adult spring Chinook (Oncorhynchus
tshawytscha)
vm
-------
ACKNOWLEDGEMENTS
We gratefully acknolwedge the assistance of the following
agencies: Fish Commission of Oregon, Wildlife Commission of
Oregon, National Marine Fisheries Service, U.S. Army Corps of
Engineers, and the Columbia River Basin Interagency Fisheries
Technical Advisory Committee.
IX
-------
SECTION I
INTRODUCTION
Scientists have studied supersaturation and its pathologic result,
Gas Bubble Disease (GBD), for over TOO years with recurrent cycles of
interest. Within the last decade interest has been renewed because
man's activities have supersaturated large volumes of surface water,
making them potentially lethal to fish (Ebel, 1969; Bouck, 1972).
Causes of excess dissolved gas pressure (Table 1) include operations
at hydroelectric or other dams, heating of water, and eutrophication
(photosynthesis). Since most means of diminishing supersaturation
could involve large financial costs, there must be a firm data base
for establishing construction priorities or the necessary levels of
water treatment and its achievable benefits.
An additional impetus for renewed research has been the develop-
ment of improved techniques such as gas partition chromatography and
the Weiss Saturometer which permit rapid analysis of total and
individual dissolved gases. One might now resolve long standing
questions regarding the biological impact of individual gases and
easily monitor supersaturation in the aquatic environment.
The condition causing gas bubble disease has several names including
supersaturation, air supersaturation, and nitrogen supersaturation. The
term "air supersaturation" (Table 1) is used here because it correctly
implies the typical problem, namely that the sum of the pressures of
the dissolved components of air exceeds the combined compensatory
pressures of the atmosphere, water, and tissues. Boyle's Law prescribes
that gases will not cavitate regardless of how supersaturated a given
gas may be, unless the total dissolved gas pressure (tension) exceeds
the compensatory pressures.
The primary goal of this study was to obtain data on the acute
lethality of supersaturated water to Pacific salmonids and a predator
species (largemouth bass). These data were needed to evaluate what
criterion of saturation (relative to barometric pressure!/) would
adequately protect fish and aquatic invertebrates. Shallow water and
continuous exposure were selected for test conditions because they
I/ Dissolved gas(es) levels conventionally are described as a percentage
of some reference value, typically barometric pressure which varies
continuously due to weather and between locations due to altitude.
Continued use of this convention adds a potentially significant error
to the data. This is because the relative value of 110% saturation can
represent a (surface) uncompensated absolute force ranging from about
77.2 Kdynes at 7200 ft. above sea level to about 101.2 Kdgnes at sea
level. However, abandoning this convention might seriously confuse
engineers and biologists who are not associated with this research.
The conversion of relative saturation levels to absolute physical units
is presented in Table 4.
-------
TABLE 1. EXAMPLES OF NATURAL AND MAN-MADE SOURCES OF SUPERSATURATED WATER
Type No. I
AIR ENTRAPMENT
Type No. II
THERMAL
Type No. Ill
ORGANIC
A. Natural 1.
2.
Water falls with deep
plunge basins.
Turbulence in high
velocity streams.
Cold water recharging an
aquifer or lake, then
geothermally warmed, then
returned to surface.
Hot weather, often preceded
by cold rain.
Eutrophic levels of
photosynthesis in
conjunction with high
levels of temperature
and solar radiation.
B. Man-made 1.
2.
3.
Flood gates or other 1.
spillways at dams which
entrain air bubbles and
carry them to depths.
Injection of air to prevent 2.
"water hammer" in turbines,
sluiceways, or to reaerate
reservoir water.
Venturi action at pipe
joints, or pumps sucking
air.
Heating water to cool steam-
electric stations or other
industrial processes.
Heating water in fish culture
stations to achieve optimal
growth.
Same as above except may
occur in polluted water
to a much greater degree.
-------
represented the maximum environmental stress, eliminated significant
hydrostatic compensation, and allowed a more direct comparison of
tolerances between species. Collateral goals were to identify factors
which might diminish tolerance to gas bubble disease, to determine
residual effects on salt water adaptation, and to estimate if reproduc-
tion was readily and adversely affected by supersaturation.
The literature on Gas Bubble Disease has been reviewed recently
by Harvey (1975), Wolke, Bouck and Stroud (1975), Bouck (1973), Weitkamp
and Katz (1973), and Rucker (1972). These publications contain
additional resource material and more detailed descriptions of Gas
Bubble Disease.
-------
SECTION I!
CONCLUSIONS AND RECOMMENDATIONS
1. Some mortality occurred among sensitive fish at dissolved gas
levels of 110-115% of barometric pressure when they were restricted
to shallow water. These and higher gas levels may be safe for wild
fish if they sound to compensatory depths.
2. The mean hours to 20% mortality and two-tailed 90% confidence
limits are:
Supersaturation Level
Salmonids*
Adul
Smol
ts
ts
Juveniles
1151
309 (t
154
125 (±
1
14)
**
1
22)
48
41
53.5
120%
(± 8.7)
(t 10.1)
(± 23.8)
18.
17.
23.
1
3
2
6
25%
(i
(*
(t
3.
5.
4.
2)
2)
4)
*(10 C, shallow water)
**(= single sample only)
3. Response {time to death) curves are skewed to the right and contain
at least three phases which include: (1) a sublethal period; (2) a
period of log-linear mortality; and (3) a period of protracted survival.
These skewed curves indicate that: (1) as much as 75% of a laboratory
population of fish can be killed in about half the time required to kill
the total population; (2) probably few factors influence the mortality;
and (3) the median is probably the best measure of central tendency in
these data.
4. Times to death varied among the tests, genera and species in some
cases. Variability was greater in long-term exposure tests than in
short, acute exposures.
5. Several additional factors were significantly correlated with time
to death. These included life stage, physiological condition, body
size, activity, behavior, and water temperature.
6. Largemouth bass survived prolonged exposure at 120% of barometric
pressure (203 g/cm2 hyperbaric) which killed salmon and trout. This
level caused external signs of gas bubble disease on bass but did not
prevent them from preying on young salmonids. Thus supersaturation may
exert ecological succession pressure against salmonids and favor bass.
-------
7. The external lesions of gas bubble disease such as dermal emphysema
or exophthalmus varied between species and became most severe when
conditions were marginally lethal; such conditions allowed the longest
survival, hence the longest period for the pathologic process to develop.
In acutely lethal cases with short exposure time, early mortality
frequently showed no external signs of gas bubble disease. Therefore,
the presence of air emboli in the blood along with identification of
supersaturated water must be used to confirm acutely lethal gas bubble
disease.
8. Supersaturation levels which killed 50% of the smolts in a species
did not prevent the survivors from tolerating a direct transfer to 30 °/oo
seawater where they survived for five days.
9. Appearance, fertility and hatching of eggs were not adversely
affected in spring chinook which had been held at 110% for three months
(IOC, 65 cm of water) and artifically spawned.
10. Actual conditions encountered by fish in supersaturated water
should be clearly defined, such as intermittent exposure and water depth.
11. A zero supersaturation limit is recommended for salmonid fish
hatcheries; any supersaturation typically indicates a more basic problem
that is capable of reaching harmful levels especially when the water is
heated or pumped.
-------
SECTION III
METHODS AND MATERIALS
SUPERSATURATION EXPOSURE FACILITIES
Each exposure chamber operated independently and consisted of
a large fiberglass tank about 3.75 m in diameter by 1 meter deep.
A frame supported rubber-coated nylon sides (1.2 m high) and 2.5 cm
mesh netting over the top. These tanks were located outdoors, but
insulated with 5 cm of urethane foam and located under a roof to
shield them from direct sunlight (Figures 1 and 2).
Well water was aerated to near saturation before being pumped
at 1900 kilo dynes/cm^ through each exposure system. A small amount
of this water was recycled to the aeration tank to promote aeration.
Compressed air was injected at a constant rate into the pressurized
water immediately prior to passing thru pipe ells (Figure 1) where
turbulence promoted the solution of the gases. The water passed
into a retention tank and the undissolved air was removed by drawing
air off the top; water was drawn off the bottom and delivered to the
test tank. The level of total dissolved gas pressure (supersaturation)
was proportional to the volume of air being injected into the system.
As tested, this system could deliver 75 1/min of water with a total
dissolved gas pressure easily as high as 150 percent of barometric
pressure. Water was introduced into the exposure chamber via a
subsurface port to reduce gas losses to the air. Valves were omitted
wherever possible to avoid sharp drops in pressure which remove gases
from solution.
Additional exposure chambers were located within a circular cage
made of wood, coated with epoxy paint, and covered with nylon bobinet
(Figure 2 G). Each cage, about 1.3 m in diameter and 1 m high, was
equipped with hinged doors and sub-divided into quadrants making four
additional exposure chambers. The cage assembly sat over and was
held in place by the tank stand-pipe. Water currents within the cage
promoted good mixing of water, but velocities were slower in the cage
than in the open tank.
WATER QUALITY
Water for these experiments was drawn from a well located adjacent
to the Willamette River. Water from this source has been used for fish
culture for five years, and its general characteristics are listed in
Table 2. Hardness and alkalinity were typically low (20 ppm as CaC03)
and ranged between pH 7.0-7.4 after aeration. Water temperature was
kept at 10 C except where it was altered as an experimental variable.
-------
FIGURE 1. DIAGRAM OF FACILITIES FOR EXPOSING FISH TO SUPERSATURATED WATER.
SUPERSATURATION GENERATOR
air line
Air Flow Meter
Water
Flow Meter
-Turbulence Loop^
(airtwater)
J|^-Air Escape Vent
Retention
Tank
T^
Exposure
Tank
-------
ATest tanks with roof to shield out
direct sunlight.
B--Aeration tank with incoming water
supply and continuous temperature
control system.
C--Test tank with saturometer in
preparation for gas analysis.
Figure 2. Supersaturatfon exposure facilities
8
-------
D (left), E (right)Supersaturation generator system (see Figure 1).
FAir and/or compressed gas delivery
system of Supersaturation generator.
G--Four-chambered nylon bobinet cages
within the test (exposure) tanks.
Young sockeye salmon can be seen in
the chambers.
Figure 2 (continued). Supersaturation exposure facilities
-------
TABLE 2. A SUMMARY OF CHEMICAL AND PHYSICAL PARAMETERS IN THE WATER
SUPPLY OF WESTERN FISH TOXICOLOGY STATION 1972-1973V.
Parameter-/
Temperature (°C)
pH
QUANTITY
Dissolved Oxygen
Total Alkalinity
Total Hardness
Nitrites
Nitrates
Ammonia
Calcium
Magnesium
Sodium
Potassium
Sulfate
Dissolved Solids
Suspended Solids
Chloride
Min
7.0
6.70
(mg/1 ,
2.8
18
21
<.001
.138
<.001
5.8
1.4
4.1
0.48
2.0
34
<1.0
3.0
Max
18.2
7.00
ppm)
9.0
26
25
<.001
.330
.048
13.7
2.4
5.4
0.70
4.0
68
4.0
9.0
Mean
13.1
6.80
4.9
22
23
<.001
.176
.014
7.3
1.7
4.6
0.57
2.9
52
1.7
6.1
2/
Parameter-
Heavy Metal
Cadmium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Nickel
Zinc
Mercury
Min
QUANTITY
s
0.0
1.2
0.5
1.0
55
3.0
1.0
1.0
1.0
<0.5
Max
(yg/1, ppb)
0.3
5.0
3.0
9.0
360
5.0
22.0
4.0
24.0
2.0
Mean
0.1
1.3
1.4
3.5
no
4.7
2.9
1.9
6.4
0.7
I/ Unpublished Data. Donald F. Samuel son
2/
' Number of samples for all parameters ranged between 23-26.
Maximum values occurred during winter months (Dec-Feb) when
saturated ground water dominated the river water supplying
our well.
10
-------
Water quality in the experimental and control tanks was monitored
continuously for temperature; usually once a day for pH, alkalinity and
oxygen; and several times each day for total dissolved gas pressure.
Chemical analyses were done according to "Standard Methods for the
Examination of Water and Wastewater" (Anon., 1971).
Total dissolved gas pressure (TDGP) was measured with a dissolved
gas tensionmeterl/ (Table 3). Fickeisen, Schneider and Montgomery
(1974) reported the accuracy of this instrument to be comparable to
the results obtained by partition chroma tography.
Figure 3 depicts a dissolved gas tensionometer consisting of
four components: (1) a molecular sieve which excludes liquid water but
allows gases to reach pressure equilibrium with their dissolved phases
(400 ft. coil of 0.025" O.D. X 0.012" I.D. dimethyl silicone rubber
tube); (2) a gauge for measuring gas pressure or tension; (3) nylon
tubing (gas tight) to connect components (1) and (2); and (4) a framework
to support and protect the rubber tubing. The sensor is placed under
water, and the gas pressure in the tubing equilibrates with the gas
pressures in the water, including water vapor pressure. The gauge
directly measures the resulting total gas pressure, indicating a
hyperbaric or positive pressure in supersaturated water and a hypo-
baric or negative pressure in water that is not fully saturated.
The measurement of dissolved gas pressure is based on the principle
that the membrane is selectively permeable to gases but not to liquids.
A dissolved gas crosses the membrane and in due time reaches a pressure
equilibrium with its dissolved gas phase. The gauge sums the pressures
of the individual gases including that of water vapor, but always
relative to ambient barometric pressure. Therefore, this instrument
measures the total pressure difference (AP) between barometric pressure
and the sum of dissolved gas pressures including water vapor pressure.
The relative total dissolved saturation is calculated by the
formula:
VP x 100j where
BP
BP = barometric pressure in mmHg;
AP = differential gas pressure in mmHg (from saturometer);
VP = vapor pressure of water in mmHg.
Thus "percent saturation" is a term which relates to the local
barometric pressure, which in turn changes with altitude and weather
conditions. In our case, barometric pressure averaged about 755 mmHg
I/ Developed by Dr. Ray Weiss, Scripts Institute of Oceanography,
LaJolla, California, and initially called a "saturometer".
11
-------
TABLE 3. NOMINAL AMD MEASURED TEST LEVELS OF TOTAL GAS SUPERSATURATION.
Test
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Test
Chamber
1
1
1
1
1
1
2
1
2
1
1
2
1
2
1
3
2
1
2
1
1
1
2
3
Nominal
Total Gas
Saturation (%]
125
120
125
120
115
125
120
120
115
125
125
120
140
130
120
115
140
140
130
135
140
120
115
no
Measured Total Gas Saturation (%'
1 N
16
59
31
44
42
43
53
37
39
45
43
42
20
30
24
34
8
16
22
7
10
10
21
19
Mean
125.2
120.1
125.5
120.2
115.6
125.0
120.1
120.3
114.9
124.7
124.8
120.1
139.3
129.7
120.6
114.7
140.6
140.7
130.1
136.0
141.2
120.0
115.2
109.8
Range
123.6
116.6
122.7
115.2
114.0
123.5
118.2
118.3
113.0
123.2
122.6
118.7
137.2
128.2
117.7
112.6
139.9
138.4
127.5
133.3
138.9
117.4
113.5
109.1
- 127.6
- 123.7
- 127.8
- 123.5
- 125.0
- 127.7
- 122.2
- 122.6
- 116.3
- 126.3
- 126.3
- 121.0
- 141.0
- 130.8
- 122.0
- 116.5
- 141.0
- 142.4
- 133.4
- 138.5
- 143.2
- 121.1
- 116.9
- 110.6
17 1 120 7 119.8 119.1 - 120.5
12
-------
TABLE 3. NOMINAL AND MEASURED TEST LEVELS OF TOTAL GAS SUPERSATURATION (CONTINUED),
Nominal Measured Total Gas Saturation (%)
Test Test Total Gas
No. Chamber Saturation (%) N Mean Range
18 1 125 7 125.5 123.9 - 127.2
19 1 130 17 130.4 126.4 - 133.0
2 125 17 125.0 123.3 - 127.2
3 120 43 119.8 115.1 - 123.4
20 1 120 35 120.2 117.1 - 121.9
2 115 59 115.0 111.9 - 117.3
3 110 157 109.8 107.0 - 116.5
13
-------
SATUROMETER
Handle -+>
Nylon Tubing
Pressure
Guage
^ Valve
Gas Permeable Membrane
(Dimethylsilicone tubing)
Protective Support
Figure 3. Diagram of modified Weiss Saturometer.
14
-------
and varied typically by less than 20 mmHg (2.6% of average). However
a given "total" saturation value does not imply that each component
gas was saturated to that level. For example, a total saturation of
110% can consist entirely of high N2 levels and low 02 levels or vice
versa.
Since the causative agent of gas bubble disease is related (in
part) to the dissolved gas pressure above barometric levels (Bouck,
1973; D'Aoust and Smith, 1974), this is called the AP or hyperbaric
pressure, and it can be measured both accurately and directly with the
saturometer. The AP is the uncompensated hyperbaric pressure that
provides the driving force to form emboli and emphysema by overcoming
the compression forces in blood and other tissues. Hyperbaric pressure
may prove to be a more universal and determinative parameter than
"percent saturation" because it is independent of elevation, barometric
pressure, salinity, or other influences. Table 4 lists equivalent
values for percent saturation, g/cm2, mmHg, and p.s.i.
BIOLOGICAL METHODS
Fish for these experiments were obtained from several sources
listed in Table 5. All juvenile fish were hatchery reared, mainly at
this laboratory, but some adult fish were returnees to hatcheries.
Largemouth bass, sockeye, and possibly spring chinook salmon were wild
fish. Fish usually were placed into the exposure tank and allowed to
adjust to their surroundings for at least 48 hours prior to the onset
of supersaturation. Food was routinely withheld during each experiment,
but some exceptions are so noted. Adult fish were tested in the open
tanks, but smaller fish were usually tested in the bobinet cages
(Figure 2).
Common and scientific names of fishes conform to that of Bailey,
et al_., (1970).
Adult salmon were spawned artificially by the "dry" method after they
ripened sexually (Leitriz, 1959). Eggs were incubated in tray type
incubators. Dead and infertile eggs were removed after pigment was
evident in the embryonic eye and again after hatching was complete.
Temperatures varied seasonally during incubation from 10 C down to
about 6 C.
DESIGN OF EXPERIMENTS
The three objectives were to: (1) determine and compare the exposure
times (in hours) which killed 50% of the specimens at given test
conditions; (2) determine the effects of supersaturation on saltwater
adaptation; and (3) determine the effect of long-term exposure to low
15
-------
TABLE 4. CONVERSION TABLE AND EQUIVALENT UNITS FOR
EXPRESSING SATURATION DATA
Percent
saturation
at 755 mmHg
100
105
no
115
120
125
130
135
140
145
150
Hyperbaric Pressure
g/cn.21/
0
51.4
102.6
153.9
205.3
256.7
307.9
359.2
410.6
461.9
513.6
Torr or mmHg
0
37.8
75.5
113.2
157.0
188.8
226.5
264.2
302.0
350.0
377.5
lb/in2
(psl)
0.00
0.75
1.47
2.21
2.94
3.68
4.41
5.15
5.88
6.62
7.35
Bars2/
0
0.05
0.10
0.15
0.20
0.25
0.30
0.36
0.40
0.46
0.50
' To convert g/cnr to dynes, multiply by 980 cm/sec^ (which is the
approximate acceleration of gravity)
=J Bars are one of two pressure units recommended for use by the
Society for Experimental Biology and Medicine (1972):
1 bar = 10 Pascals = 14.5 lbs/in2 = 0.987 atmospheres
16
-------
TABLE 5. FISH SPECIES TESTED
Species
Coho Salmon
(Oncorhynchus kisutch)
Sockeye Salmon
(Oncorhynchus nerka)
Chinook Salmon
spring migrating variety
(Qncorhynchus tsnawytscha)
Steel head Trout
(Salmo gairdneri )
summer race
winter race
Rainbow Trout
(Salmo gairdneri)
Suckers
(Catostomus sp.)
Largemouth Bass
(Micropterus salmoides)
Test No.
6,7
6,7
6,7
6,7
4,5,7
16,17,18
1,2,3,4,5
20
16
19,20
8
16
1,2,5
7,8
16
11
9
3,4,5,9
20
9,10,12,13,14,15
9,10,12,13,14,15
Life Stage
parr*
post smolts*
jacks
adults
parr*
smolts*
adults
parr*
smolts*
adults
parr
smolts
adults
parr*
smolts*
adults
parr*
yearlings*
adults
juveniles
adults
Source
Oregon Fish Commission Hatchery
Fall Creek, Lincoln Co., Oregon
Bonneville Hatchery - Ore. Fish Comm
Columbia River, Bonneville Ladder
Oregon Fish Commission, Willamette
Hatchery, Oakridge, Oregon
Columbia River Bonneville Ladder
Washougal River Hatchery,
Washington Game Commission
Columbia River, Bonneville Ladder
Oregon Wildlife Commission Hatchery
N, Fk. Alsea R., Benton Co., Oregon
Oregon Game Commission, Roaring
River Hatchery, Linn Co., Oregon
Columbia River at Bonneville
Ponds, Harrisburg, Lane Co., Oregon
* Reared from eggs at WFTS.
-------
level supersaturation on maturation and fertilization of sex products
and subsequent hatching success. Collateral objectives included the
identification of factors influencing mortality and observations of
the pathobiology of gas bubble disease.
Fish were assigned randomly to tanks, and a fish was judged to
be dead when it no longer moved or ventilated. Eight levels of
saturation were selected for testing: 100 (control), 110, 115, 120,
125, 130, 135, and 140% of barometric pressure (hence total gas
saturation). Circumstances never permitted the concurrent testing at
every level, but a control group and at least one previously tested
level were repeated in a series. Replicates or re-runs were conducted
whenever the circumstances permitted, and several life stages or species
were tested concurrently in most cases. The fish were observed
frequently, and times to mortality were plotted on log probit paper for
visual estimation of the time to median mortality (ETso). Raw data
were plotted and used to estimate the relationship between various levels
of supersaturation and respective times to 201 and median mortality.
For the latter purpose all salmonid data were pooled for a given life
stage.
To study tolerance to the combined stresses of supersaturation
followed by saltwater stress, smolts of three species were exposed to
three levels of supersaturation for four days at 10 C or until 50%
mortality occurred. Immediately thereafter, the survivors were placed
in 20 liter jars containing seawater (30 °/oo) and equipped with an
air supply and refrigeration to maintain 10 C t 1°. Survival was
noted at the end of a five-day period.
The effects of sublethal supersaturation upon sexual maturation,
fertility, and hatching were studied using adult spring chinook trapped
in the Columbia River. These were held at 110% from April 23, 1973
until August 3 (103 days), then maintained at normal gas levels for
about three weeks before the fish were artificially spawned.
The general effects of temperature on mortality were studied at
10, 15, and 18 C. The fish were adjusted at a rate i 3 C/day until they
reached the test temperature. Fish were maintained at the test temperature
for at least two days prior to initiating supersaturation.
STATISTICAL PROCEDURES
Fish were assigned to treatments using prepared random number cards.
All statistical procedures conform to those of Snedecor and Cochran
(1973). Statistical significance was assigned whenever the probability
of obtaining a given result by chance alone was <.10.
18
-------
SECTION IV
RESULTS
GENERAL TOLERANCES OF SALMONIDS AND NON-SALMONIDS
Preliminary concurrent exposure of several representative species,
conducted at 130-150% saturation (10 C), established that salmonid
species (Salmo and Oncorhynchus) were highly susceptible to gas bubble
disease. Shiners (Notjopis sp.) and crappies (Pomoxis sp.) also died
rapidly under those conditions. Bluegill (Lepomis macrochirus),
squawfish (PtychocheiTus oregonesis), and warmouth (Lepomis gulosus)
were intermediate in~toTeranee to bullhead (Ictalurus sp7)7 largemouth
bass (Micropterus salmpides), and carp (Cyprinus car pio) which were
extremely hardy under those test conditions.TFese results reveal
differences in tolerance between genera and demonstrate that salmonids
were both highly sensitive and easy to use in these tests.
SAMPLE RESPONSES TO SUPERSATURATION
A general estimation of population response (mortality) to super-
saturation is essential to understanding the overall problem. Specific
information on this subject has not been found in the literature: hence,
we analyzed the data from twenty bioassays (each used either 10 or 20
fish per bioassay) and estimated the response curve. Individual times
to death from experiments with 100% mortality constituted the raw data.
Various measures of central tendency or dispersion were estimated for
population parameters listed in Table 6.
As a preliminary step, pooled times to death data for all life
stages were tested by analysis of variance, and this revealed that
the samples were significantly heterogeneous at the 10% level for a given
level of saturation. Subsequently the data were pooled only within a
given life stage because gross differences in tolerance were apparent
between life stages.
Curves plotted for cumulative mortality and rate of death (Figures
4 and 5) show three phases of population response to supersaturation.
The initial phase consists of a sublethal morbid period during which
emboli form and related dysfunctions approach critical limits for fish
life. The second phase is a period of extensive mortality which appears
linear with the log-jQ of exposure time. The third phase did not occur
in all samples nor at all saturation levels, although its intermittent
occurrence indicates likelihood if larger samples or more replicates
were tested for a longer time. During the third phase the more tolerant
fish died at a greatly reduced rate, typically over an extended period
of time. This protracted survival period gave many fish samples a
definite and significant skewness that rejects the Hypothesis of random
normal distribution. Possibly a fourth phase may exist consisting of
post-exposure mortality or recovery from all effects that carried over
after returning to normal saturation levels.
19
-------
TABLE 6. SUMMARY STATISTICS OF TIME TO DEATH AT THREE LEVELS OF SUPERSATURATION BY TWENTY
POPULATION SAMPLES OF RAINBOW TROUT, PACIFIC SALMON, AND LARGEMOUTH BASS.
Number 90% Coefficient Coefficient
in Range Average Belt Median Standard of of
Sample (hours) (hours) for u (hours) Deviation Variation Skewness
IOC, 130% Supersaturation
1. Spring Chinook salmon (Oncorhynchus tshawytscha) (IFL = 81.0cm; IWt = 7.738kg)
adult female, migrating
a. Time to death* 11 7-12 8.909 ±0.929 8.5 1.700 19.0 +0.160
b. Log1Q - 0.845-1.0792 0.943 +0.044 0.954 0.082 8.7 +0.419
2. Spring Chinook salmon (Onco_rhy_nchus tshawytscha) (IFL = 60cm; IWt = 5.618g)
adult male, migrating
a. Time to death 9 7-11 9.440 ±0.701 10 1.130 12.0 1.486
b. Log1Q - 0.845-1.041 0.972 +0.034 1.00 0.056 5.7 1.497
IOC, 125% Supersaturation; IOC
3. Rainbow trout (Salmo gajrdnerij ("XFL = 15.2cm; IWt = 42.2g)
' parr
p a. Time to death 20 20-48 30.150 +3.276 27 8.474 28.1 1.115
~ b, Log1Q - 1,301-1.681 1.4645 +0.044 1.431 0.114 7.8 0.874
4. Rainbow trout (Salmo gajjdneri) (XFL = 28,18cm; XWt - 259.39)
yearling
a. Time to death 10 17-45 33.70 +5.556 31 9.696 28.8 0.835
b. Log1Q - 1.230-1.653 1.1510 ±0.0792 1.4914 0.1385 12.0 7.3732
5. Winter steelhead trout (Salmo gairdneri) (IFL = 7.8cm; Iwt = 5.6g)
parr
a. Time to death 20 10-224 66.55 +24.34 35 62.966 94.6 1.360
b. Log1Q - 1.000-2.3502 1.6519 ±0.966 1.5524 0.150 24.4 0.559
6. Sockeye salmon (Oncorhynchus nerka) (XFL = 10.2cm; IWt = 12.Og)
parr
a. Time to death 20 13-189 62.150 +18.566 40.0 48.022 77.3 1.509
b. Log1Q - 1.0128-2.2765 1.6551 "+0.1420 1.6021 0.367 22.2 0.615
7. Spring chinook salmon (Oncorhynchus tshawytscha) (IFL = 79.6cm; Iwt = 5.816kg)
adult females
a. Time to death 16 12-22 17.875 ±1.357 18 3.095 17.3 0.605
b, Log1Q - 1.0792-1.3424 1.2456 ±0.0346 1.2553 0.0798 6.4 0.810
8. Coho salmon (Oncorhynchus kisutch) (IFL = 20.4cm; IWt = 87.5g)
post smolts
a. Time to death 20 6-21 12.500 +1.559 12.0 4.032 32.2 0.372
b. Log1Q - 0.7782-1.3222 1.0578 ±0.0556 1.0792 0.1468 13.8 0.437
-------
TABLE 6. SUMMARY STATISTICS OF TIME TO DEATH AT THREE LEVELS OF SUPERSATURATION BY TWENTY
POPULATION SAMPLES OF RAINBOW TROUT, PACIFIC SALMON, AND LARGEMOUTH BASS (CONTINUED).
co
9. Coho salmon (Oncorhynchus kisutch) (XFL
jacks (precocious;
a. Time to death
b. Log10
10. Coho salmon (Qncorhynchus kisutch) (XFL
adult female, gravTa
a. Time to death
b. Log,,,
11. Coho salmon (Oncorhynchus kisutch) (TFL
adult male, ripe
a. Time to death
b. Log]Q
IOC. 120% Supersaturatiqn
Number
in
S_arng]_e_
= 40.3cm; XWt = 795g)
20
12-21
1.0792-1,3222
= 71.1cm; at = 4.173kg)
10
15-28
1.1761-1.4472
66.8cm; XWt = 3,578kg)
10
12-42
1.0792-1.6232
12. Rainbow trout (Salmo gairdneri) (XFL = 15.3cm; XWt = 43.9kg)
parr
a. Time to death 20 32-193
b. Log1Q - 1.5051-2.2856
13. Winter steelhead trout (Sajnro aan£dner[) (XFL = 69.2cm; XWt = 3.563kg)
adult females, gravid
a. Time to death
b. Log]Q
14. Winter steelhead trout (Sal_mo gairdneri) (XFL = 70.2cm; XWt = 3.336kg)
adult males, ripe
a. Time to death
10 59-164
1.7709-2.2148
b-
15
Coho salmon (Oncorhynchus kisutch) (XFL
adult female, migrating, gravid
a. Time to death
b. Iog10
16. Coho salmon (Oncorhynchus kisutch) (XFL
adult female, gravid
a. Time to death
b. Log10
10 57-144
1.7559-2.1584
= 73.0cm; XWt = 4.372kg)
10 31-64
1.4914-1.8062
= 70.4cm; IWt = 4.080kg)
10 42-110
1.6232-2.0414
90X Coefficient Coefficient
Average Belt Median Standard of of
(hours) lorji (hours) Deviation Variation. Skewness
17.200 ±1.276 19 3.3015 19.2 1.635
1.2275 +0.0054 1.2788 0.0865 7.0 1.780
20.900 +1.902 21 3.2812 15.7 0.091
1.3153 +0.0392 1.3222 0.0683 5.2 0.301
23.2 +5.161 20 8.904 38.38 1.078
1.3403 70.1867 1.3010 0.1520 11.34 0.015
70.650 +16.403 51 42.475 60.1 1,388
1.7950 +0.0802 1.7076 0.2078 11.6 1.262
104.50 4-19.316 92 33.323 31.9 1.107
1.9987 +0.0797 1.9638 0.1377 6.8. 0.1377
90.30 +16.599 79 28.6397 31.72 1.1837
1.9370 ~+0.0765 1.8973 0.1324 6.83 0.8998
45.700 +5.178 45 8.9324 19.5 0.235
1.6525 +0.0488 1.6532 0.0843 5.1 0.025
58.80 +11.348 51 19.577 33.2 1.1950
1.7526 ~+0.0693 1.7076 0.1197 6.8 1.1288
-------
to
to
TABLE 6. SUMMARY STATISTICS OF TIME TO DEATH AT THREE LEVELS OF SUPERSATURATION BY TWENTY
POPULATION SAMPLES OF RAINBOW TROUT, PACIFIC SALMON, AND LARGEMOUTH BASS (CONTINUED).
17.
20C
18.
20C
19.
20.
Number
in
Sample
Spring Chinook salmon ( Oncorhy nchus tshawy tscha ) (XFL = 66,
adult female, migrating
a. Time to death 9
b. Log10
, 130% Supersaturation
largemouth bass (Micrppterus salmoides) (YFL = 22,4cm; XWt
a. Time to death 20
b. Log1Q
1401 Supersaturation
Largemouth bass, adults (XFL = 23.1; XWt = 187. 5g)
a. Time to death 20
b. Log10
Largemouth bass, adults (XFL = 24.0cm; Iwt = 209. 6g)
a. Time to death 20
b. Log1Q
Range
(hours)
Average
(hours)
90%
Belt
for u
Median
(hours)
Coef f i cient Coef f i ci ent
Standard of of
Deviation Variation Skewness
,7cm; IWt = 5.208kg)
22-99
1,3424-1.9956
- 163. 4g)
25-69
1,3979-1.8388
10-22
1.000-1.3424
11-19
1.0414-1.2788
54.
1.
45.
1.
16.
1.
12.
1.
44
6676
45
6424
550
2098
45
0895
+16
±°
±4
+0
ll
to
to
to
.735
.1450
.782
.0454
.3924
.0381
.8551
.0696
51
1.7076
45
1.6532
16
1.2041
11
1.0414
30
0
12
0
3
0
2
0
.623
.2653
.369
.1175
.349
.0924
.2117
.0696
56.2
15.9
27.2
7.2
20.2
7.6
17.8
6.4
0.337
0.452
0.109
0.274
0.493
0.183
1.966
2.074
-------
CO
co
QC
O
Q
UJ
Z)
u
100-
90-
80-
70-
60-
1
50-
40-
30-
20-
10-
0
MEDIAN
AVERAGE
I
I
10
CUMULATIVE MORTALITY
Coefficient of skewness= 1.36
MORTALITY PER DAY
:'5£
20
30
40
50
60
70
80
90
h-
-10^
_ o:
o
h5
100
DAYS OF EXPOSURE
UJ
O
CC
111
CL
Figure 4. Time to death of winter steelhead trout parr (Saimo gairdneri)
at 125% super saturation and IOC in water 65 cm deep.
-------
to
PHASE-I | PHASE-IT I
'
>IOO^
h-
cr
o
60H
U40-
0
/
(
MORBIDITY ! / LINEAR \l
i (MORTALITY)'
MEAN
MEDIAN
PHASE IE
(PROTRACTED SURVIVAL)
RATE OF MORTALITY
20 40 60
80 100 120
HOURS
140 160
180 200
Figure 5. Mortality response of rainbow trout (Salmo
gairdneri) to 120% supersaturation.
-------
Skewness was prominent (Table 6) whenever phase III mortality
occurred and relatively high whenever the mortality data were plotted
as hours to death. Although a log transformation of the data generally
diminshed skewness, it still was a prominent factor approaching a high
level of statistical significance (Snedecor and Cochran, 1973). Its
occurrence at high and low levels of supersaturation indicates that the
shortest time to mortality is more sharply defined than the maximum
time to death. Further, skewness usually indicates that a relatively
small number of factors are influencing the resulting population
response, and this may be true for the fish's tolerance to supersaturation,
However, skewness varied between tests and, in the case of adult female
coho salmon (at 120% supersaturation), the coefficient of skewness
increased from 0.235 in late September to 1.195 in late October.
Possibly this reflects the impact of selection via prespawning mortality
prior to capture but, whatever the cause, it influenced the response
curve. Higher levels of skewness were more associated with low levels
of saturation, male fish, and immature juvenile fish than their counter-
parts.
Central tendency was examined by comparing means, median, and
modal values. Confidence limits can be derived for means, a valuable
advantage when statistically comparing different populations or developing
population models. However, mean time to death had required a 100% kill
which neither occurred at 115% supersaturation nor, in many cases, at
120% even after prolonged exposure. This was due in part to the skewed
population response described above. Thus, the median time to death
typically occurred before the more protracted mean time to death.
Median time to death, therefore, not only permits more tests to be
conducted in any given period, but also is an appropriate measure of
central tendency in these tests. However, applying confidence limits to
it is difficult.
Table 6 lists mean times to death varied among levels of super-
saturation, species, and life stages. The width of the 90% (two-tailed)
confidence estimate of the population mean (y) at 120% saturation varied
from a low of about - 6 hours for juvenile rainbow trout to a high of
t 19 hours for adult female steel head (winter). Confidence belts were
generally smaller at higher saturations, and the previous results would
predict extremely broad confidence belts at low levels of supersaturation.
Variability was a function, in part, of the supersaturation level,
hence coefficients of variation (C.V.) were inversely proportional to
the supersaturation level. At 120% saturation C.V. values reached a
high of 60. However, variability usually diminished a great deal when
the data were expressed at logio hours to death, and this was the case
at each saturation level. This emphasizes the need for a log trans-
formation of exposure times in studying gas bubble disease, as well as
toxicants.
25
-------
Similar to the coefficients of variation, standard deviations
were inversely proportional to the saturation level. Frequently the
standard deviations approached or exceeded 35% of the means; however,
these were much smaller when the data had been transformed into log-|Q
values.
PATHOBIOLOGY AND TIMES TO MORTALITY
Toleranceto 130% Saturation
Spring Chinook salmon and largemouth bass were the only species
tested at this high level of saturation. The times required to reach
20% mortality and median mortality at 130% are listed in Tables 7 and 8.
Adult spring chinook began dying within about seven hours, reaching
median mortality in about 8.5 hours and total mortality within 12 hours.
From these data continuous exposure to 130% apparently can kill salmonid
fish in a relatively short time provided that the fish remains in shallow
water. However, if a fish dives to a depth of 3 meters, the hydrostatic
pressure of water would prevent bubbles from forming.
Adult spring chinook at 130% of air saturation indicated approaching
mortality by their behavior. After 4 hours of exposure, morbid chinook
would swim aimlessly, unresponsive to external stimuli. Coughing and
shaking occurred frequently, usually followed by bursts of violent but
generally aimless swimming. Necropsy found few external signs of gas
bubble disease, but emboli usually were present in the blood vessels
of the gills causing discoloration. A massive air-embolus filled the
intravascular space between the capillary beds of the gills and the heart.
Air emboli could be seen in the intercostal arteries and in the dorsal
aorta. After median mortality had been reached, small emphysema (bubbles)
were evident in some fins of some fish, but usually not seen in the tail
of adult spring chinook. Most adult spring chinook salmon sank at death
and did not float during the short period between death and necropsy.
These results were reported in greater detail by Stroud, Bouck and
Nebeker (1975).
Largemouth bass, lethargic at 1301 saturation and 10 C, died only
after exposures 10 times longer than those lethal to adult spring chinook
salmon. The longer survival of bass probably accounts for more fully
developed and more frequent incidence of external signs of gas bubble
disease in these fish. After two days of exposure, all the bass were
still alive, but they contained sufficient emphysema to change portions
of their normally transparent fins to a whitish translucent color.
Emphysema usually was not observed under the skin or scales of bass.
Necropsy revealed that portions of the gills which extended to the heart
were filled with air at pressure .sufficient to distend the myocardium.
Largemouth bass floated prior to and after death from gas bubble disease,
possibly because their swim bladder has no opening to their esophagus
for volitional release of excess air. This resulted in tympanites of
the swim bladder.
26
-------
TABLE 7. HOURS OF EXPOSURE TO.REACH 20% MORTALITY AMONG FISHES EXPOSED
TO SUPERSATURATION IN SHALLOW WATER AT 10°C.
% saturation
130
Coho salmon
parr
smol ts
held over smolts
jacks
adult females (gravid)
adult males (ripe)
Spring chinook salmon
parr
smolts
adult females, migrating 7
adult males, migrating 8
Sockeye salmon
parr
smolts
adults, migrating
Steel head Trout, Summer
parr
smol ts
adults, migrating
Steel head Trout, Winter
parr
smol ts
adult females, (gravid )
adult males (rioe)
Rainbow Trout
parr
yearlings
125
37, 22
-
8, 12
14
19
19
_
-
14
15
20
11, 19,
16
<18
_
14, 14,
29
25
19, 38,
20
_
-
-
23, 24
-
30
134
30
41
49
38
56
30
30
50
36
35
50
60
72
40
46
68
68
45
23
120
, 73,
-
, 23,
, 30
, 40
.5, 43
, 40
-
, 30
» 22
, 45
, 27,
, 58
_
-
, 25
, 218
115
>336
-
120, 303
336
122
207
-
>268
430
498
49, 167
>268
140, 650
_
>268
140
160
154
332
263
_
79
no
-
-
-
-
-
-
1440
-
-
>268
_
>268
-
-
>268
-
_
-
2?
-------
TABLE 7- HOURS OF EXPOSURE TO REACH 20^ MORTALITY AMONG FISHES EXPOSED
TO SUPERSATURATION IN SHALLOW WATER AT 10°C.
% saturation
130 125 120 115 110
Largemouth bass
juveniles 115 -0-
adults 65 142
Suckers (Catostomus sp.)
adults - - 226
Tests were conducted with a maximum attainable depth of 65 cm, and 10 fish or
more per test.
28
-------
TABLE 8. ESTIMATED HOURS TO MEDIAN MORTALITY (ET§o) AMONG FISHES EXPOSED TO SUPERSATURATION
IN SHALLOW WATER AT 10°C.
to
CO
Gas levels
SPECIES 130 125 120
Coho salmon (Oncorhynchus kisutch)
a. parr
b. smolts
c. held-over smolts
d. precocious males (jacks)
e. adult females
f. adult males
Spring chinook salmon (Oncorhynchus
tshawytscha)
a. parr
b. smolts
c. adult females, migrating
d. adult males, migrating
Sockeye salmon (Oncorhynchus nerka)
a. parr
b. smolts
c. adults
Steel head Trout, Summer (Sal mo gairdneri)
a. parr
___
___
__ «
8.1
8.9
70, 44
___
11,13
17
20
20
___
___
17.5
16.0
40
20,23,22
<18
19,18
b. smolts
280,120,150;
___
44,34,40 i
60,60,50
44,52 !
80,71 ;
!
I
!
j
1
50 :
45,60 i
65,43 1
i
70,95
60,54,60 i
83 I
i
52 :
c. adults, migrating <; 33 108 i
Steel head Trout, Winter (Salmo gairdneri) 1 ;
a. parr 38,70,35 90,75
b. smolts
>48
c. adult females, gravid 92
d. adult males 80 ;
Rainbow Trout (Salmo gairdneri) I ,
a. parr 25 56
b. yearlings 31 42 :
Largemouth Bass (Micropterus salmoides)
a. juveniles 175 >240 >240 j
b. adults 130
Suckers ( Catostomus sp.)
a. adults
>240 >240 :
>216 ';
(% saturation)
115 no
>336
>240
270,>336
>336
180
265
-
>268
540
500
100,183
>268
>168
>268
>168
270
>268
>336
>336
___
140
-
_ _
___
---
___
>268,>1440
>1440
> 1-440
>268
___
-"-
**"""
""
-------
Tolerance to 125% Saturation
A 125% saturation was lethal to all the salmonid fish in six days
under these test conditions, but largemouth bass and suckers were not
killed. Median mortality was reached in an overall average of 28 hours.
Time to first mortality ranged from 6 hours for hold-over smoltsV to
25 hours for adult winter steel head, but typically occurred after 12
hours of exposure in this shallow water circumstance. Times to median
mortality ranged from 11 hours for hold-over coho smolts to 70 hours
for juvenile coho salmon and to greater than. 240 hours for bass (Table 8).
Tolerance differed very little between the sexes of a given species at
this level, but different life stages frequently showed marked differ-
ences in tolerance, e.g., coho salmon (Table 8). Median survival times
were about twice as long for juveniles of winter steel head as for summer
steel head, indicating that different tolerance may exist due to racial
stocks. However, experimental error may account for these differences
because variability sometimes exceeded 100%.
At least 90% of the fasting largemouth bass survived 125% saturation
for 10 days, but they developed extensive areas of emphysema in their
fins. At this time juvenile salmon were added to the tanks, whereupon
the bass caught and ate them. This led to the conclusion that the bass
were still sufficiently healthy to sustain their predatory inclinations.
Meekin and Turner (1974) reported that sublethal gas bubble disease
inhibited predation by adult squawfish (Ptychocheilus oregonensis).
External signs of gas bubble disease were generally more evident
at 125% than at 130% because the fish survived longer and therefore the
gas had longer to act upon them. The first mortalities at 125% had
minimal external signs of gas bubble disease; typically external signs
were more readily apparent on later mortalities. Signs include gills
mottled by emboli and emphysema in the fins, along the lateral line, and
under the skin. Eye involvement generally was minimal at 125%, possibly
because exposure time was inadequate. Steel head and rainbow trout were
the exception and developed both eye lesions and dermal emphysema, which
formed sooner and more extensively than among other test fish.
Tolerance to 120% Saturation
Time to median mortality ranged from a low of 34 hours for hold-over
coho smolts to as much as 280 hours for coho parr, but no largemouth
bass died at 120% saturation. The first recorded mortality to steel head
parr was in 16 hours. Adult salmonid fishes generally survived three
to four days before reaching 50 percent mortality.
' Smolts not liberated in spring and maintained in tanks until the time
of testing in early winter. Hold-over smolts do not exist in nature and
generally don't survive.
30
-------
Differences in tolerance to gas bubble disease were evident at
120% between adult male and female coho; gravid coho females were
less tolerant among coho and possibly this also was the case for gravid
sockeye salmon. A sex-related difference in tolerance was not noted
in adult spring Chinook, possibly because they were not gravid. Adult
winter steelhead died sooner than adult summer steelhead, again possibly
because the winter fish were ready to spawn.
Most of the previously described signs of gas bubble disease were
observed at 120% saturation where longer survival, hence longer exposure
time, increased the severity of the lesions. Eye damage became evident
at this level of supersaturation, but it occurred in probably less than
10% of the fish and primarily among the various racial stocks of Sal mo
gairdneri. The most frequently observed eye involvement was exophthalmia
from retrobulbular emphysema. This caused the eyeball to protrude well
beyond the orbit and may have diminished the functional status of the
eye. Another form of eye involvement consisted of air bubbles in the
aqueous humor and in the retina causing distorted optics, retinal detach-
ment, and hemorrhage. The latter lesions are considered a serious problem
that can lead to blindness, inability to spawn, and death.
Another significant result of gas bubble disease was the development
of petechial and ecchymotic hemorrhages in various locations including
cutaneous and intra-muscular regions. Presumably these hemorrhages
occurred in the surviving fish, but their impact and extent are unknown.
Also, adult salmon with gas bubble disease frequently developed fungal
infections which spread rapidly across their bodies.
Tolerance to 115% Saturation
Prolonged exposure was required to reach median mortality (Table 8)
at 115% total gas saturation. Total mortality never occurred at this
level during the test period (up to 30 days), hence the average log-|g
time to death could not be computed. The most sensitive fishes were
sockeye parr (ET5Q in 100 hours) and adult female coho (ETsg in 180
hours), and the least sensitive salmonids were juvenile coho salmon
(0 mortality in 336 hours). Intermediate tolerance was indicated by
juvenile steelhead (ETso in 270 hours) and adult spring chinook (ET5Q
in 526 hours). Sex related differences were more pronounced at this level
Symptoms of gas bubble disease were most developed at 115% mainly
due to the prolonged exposure to supersaturation. Exophthalmia became
even more exaggerated and frequent but still involved probably less
than 10% of the animals. Emphysema were abundant in the skin and fins
of essentially all the fish. Fungal infections became very frequent
and covered areas of ulceration. Petechial hemorrhage of hyperemia
often gave the skin a reddish color.
31
-------
Muscular emphysema were noticed in adult spring chinook after two
weeks of exposure at 115%, The emphysema grew in size as exposure time
increased to three weeks until slices of the muscle were so riddled with
holes that they resembled red "swiss cheese." The size of these
emphysema, described in greater detail by Stroud and Nebeker (1975),
reached to slightly greater than 1 cm in diameter.
Tolerance to 110% Saturation
Only a few tests were conducted at 110% total gas saturation and
were generally tolerated by the test fish. One chronic exposure was
conducted on spring chinook salmon adults and smolts. During a 3.5
month period, 10% of the adults died of gas bubble disease and secondary
complications. Similarly, 20% of the spring chinook salmon smolts died
in two months at 110%. In this case about 30% of the smolts had
exophthalmia and at least 50% had other external signs of gas bubble
disease such as emphysema in the skin or fins. Perhaps emphysema
would have been discovered in even more fish had we inspected them via
ultrasonic sound reflection (Mackay and Rubissow, 1971). Necropsy
revealed that most of the dead smolts had emboli in the blood.
Food pellets were offered to the spring chinook smolts to prevent
starvation during the long exposure period. Those smolts with
exophthalmia attempted to and could usually feed, but their actions
suggested impaired vision and, in some cases, blindness.
ESTIMATED TIMES TO 20% MORTALITY (THRESHOLD TOLERANCE)
We established the time to 20% mortality as the threshold for
mortality. Based on the previously described skewed mortality curves,
additional exposure will probably produce disproportionally greater
mortality when the 20% mortality level is achieved. Conversely, remain-
ing well below this exposure time would probably prevent losses due to
gas bubble disease. Therefore we estimated the true population mean
(p) from the observed times to 20% mortality (x) by determining the 90%
(two-tailed) confidence limits of the sample mean (55). Although the
curves are skewed, the data were pooled from all salmonid species of a
given life stage because we have not yet proven that these species were
significantly different, especially in nractical terms.
The 90% confidence limits for average time to 20% mortality are
listed in Table 9 and shown in Figure 6. At the highest stress level
(125% saturation), one can readily predict within narrow limits the
probable time to 20% mortality for a given life stage. Smolts and adults
required about 3/4 of a day and parr required nearly a full day at 125%
saturation to reach 20% mortality. As the stress level diminished to
120% of saturation, time to 20% mortality approached two days of exposure
32
-------
350 n
300-
QC
O
OJ
jjj
Ld
o:
O
250-
200-
150
100
0
100 IO5 110 115 120 125
PERCENTAGE SATURATION
Figure 6. Means and 90% confidence belts of times to
20% mortality at various life stages of
salmonids and levels of supersaturation.
33
-------
TABLE 9. STATISTICAL.ANALYSIS OF POOLED OBSERVATIONS ON HOURS TO
20% MORTALITY AMONG THREE LIFE STAGES OF FIVE SPECIES OF
SALMONIDS.V
125%
mean hours
90% limits of mean
range
n
120%
mean hours
90% limits of mean
range
n
115%
mean hours
90% limits of mean
range
n
Adults
18.33
13.19
14-25
6
48.12
18.69
22-72
12
309.11
1114.28
122-650
9
LIFE STAGE
Smol ts Parr
17.17 23.63
15.24 14.41
11-29 14-38
6 11
41.00 53.50
110.06 +23.83
27-58 25-134
6 8
154 125.30
1122.25
49-67
1 3
I/ Included spring Chinook salmon, coho salmon, sockeye salmon,
summer steel head trout, winter steel head trout, and rainbow trout.
34
-------
and the confidence belt ranged in width from 8 hours to 24 hours. At
the lowest stress level that routinely produced mortality (115% saturation),
mean times to 20% mortality increased to about 13 days (± 5 days) of
exposure for adults and about 5 days (t 5 days). Stress levels at
110% saturation did not routinely produce mortality.
Salmonid fish must typically spend several days at a maximum depth
of less than 60 cm if their populations are to experience as much as
20% mortality at saturation levels of 115%. However, caution is urged
because the confidence limits are approximately 100% of the mean, hence
the response is relatively unpredictable at low stress levels. Evidently
several factors influence tolerance at low levels of saturation, and
some of these are indicated in the sub-section on factors affecting
mortality by supersaturation.
RANK ORDER OF TOLERANCE TO SUPERSATURATION
The life stages of the various test species were placed in a rank
order beginning with the least tolerant and proceeding to the most
tolerant (Table 10). The data base was constructed from the estimated time to
median mortality at 120% total gas saturation (203 g/cm2 hyperbaric).
The authors selected this saturation level because the resulting mortality
was sufficiently protracted to permit the expression of inherent
differences while generally excluding the effects of disease and mal-
nutrition
FACTORS AFFECTING MORTALITY BY SUPERSATURATION (GAS BUBBLE DISEASE)
Test Temperature
Mortality at comparable saturation levels was determined at
temperatures between 10 C and 20 C. Sockeye salmon and largemouth bass
(both adults and juveniles) reveal markedly different response patterns
to supersaturation when tested at different acclimation temperatures
(Figure 7 and Table 11). Adult sockeye were considerably more tolerant
to 120% and 125% saturations when they were acclimated slowly from 10 C
and tested at 18 C. But sockeye salmon and rainbow trout parr showed
a variable response; increased test temperatures increased tolerance
in one case and decreased tolerance in two cases. Both juvenile and
adult largemouth bass showed a decrease in tolerance with increased
test temperature, possibly because they were stressed by temperature
acclimation and by supersaturation.
Obviously, test temperature has an influence on tolerance to
supersaturation, but resulting variability creates uncertainty in
interpreting the impact of acclimation time, temperature preference,
or both. For example, adult sockeye had been acclimated to their
35
-------
TABLE 10. RANK ORDER OF TOLERANCE BETWEEN FISH TESTED AT 120% TOTAL
SAS SATURATION (203 g/cm2 hyperbaric). I/
Species
Held-over coho smolts
Yearling Rainbow Trout
Adult female coho
Spring chinook smolts
Summer steel head smolts
Adult female spring chinook
Adult male spring chinook
Rainbow trout parr
Precocious male coho (jacks)
Sockeye smolts
Adult male coho
Adult male winter steel head
Sockeye parr
Winter steel head parr
Adult sockeye
Adult female steel head
Adult summer steel head
Coho parr
Largemouth bass
J
Tests were conducted at 10°C with
>/
/
Times to median mortality averaged
i/
ET %
tl50
39.3
42.0
48.0
50.0
52.0
52.5
54.0
56.0
56.6
58.0
75.0
80.0
82.5
82.5
83.0
92.0
108.0
183.0
>240.0
a maximum water depth
from Table 8,
Tolerance ~,
ratio -'
1.00
1.06
1.22
1.27
1.32
1.34
1.37
1.42
1.44
1.48
1.90
2.04
2.10
2.10
2.11
2.34
2.75
4.66
>6.00
of 65cm.
-------
2O-.
o
10
50
HOURS TO MEDIAN MORTALITY
adult sockeye (Oncorhynchus nerka)
A^ juvenile sockeye
juvenile largemouth bass
adult largemouth bass (Micropterys solmoides)
o---o rainbow trout (Salmo qairdnen)
Figure 7. Effects of test temperature on the time to
median mortality from supersaturation.
>168
37
-------
TABLE 11. SUMMARY OF THE RELATIONSHIP BETWEEN ACCLIMATION TEMPERATURE
AND TIME TO MEDIAN MORTALITY IN SUPERSATURATED WATERS.
Hours to medianmortality at temperature
10 C 15 C 18 C
Species and Saturation level
1. Sockeye salmon (Oncorhynchus nerka)
a. adults (1) 125218
(2) ~\2Q% 83
b. parr (1) 125% 38
(2) 120% 58
2. Rainbow trout (Salmo gairdneri)
a. parr 125% 29
1201 46
3. Largemouth bass (Micropterus salmoides)
a. adults (1) 140°/ 32
(2) 130% 130
b. juveniles (1) 140% 45
(2) 1301 175
22
21
46
168
16
75
18
92
12
42
10
60
38
-------
test temperatures for about three weeks prior to testing. These fish
showed increased tolerance to supersaturation at higher acclimation
temperatures. This was also the case for one group of juvenile sockeye
and rainbow trout which were well acclimated to warm water. All of
the other salmonid fish in the experiment had little or no time for
temperature acclimation (juvenile sockeye and rainbow trout) and they
showed decreased tolerance to supersaturation,
Bass had been collected by seining farm ponds on November 27, 1972,
and some fish were tested at 10 C beginning December 4, 1972. The bass
were kept at about 10 C until March 19, 1973. Then the fish were warmed
to 20 C in about 4 days and allowed to adjust for an additional 2.5 days
prior to beginning the test. This procedure was repeated several times
in the testing of the bass.
Whether the relatively rapid change from cold to warm temperatures
had influenced the tolerance of bass is unknown, but their response
conforms to the pattern established by salmonids which experienced
minimal temperature acclimation time. However, bass prefer'warm water
and were more active at 20 C than IOC. Possibly the observed differences
in tolerance to gas bubble disease between bass and salmon are based on
more fundamental differences between warm water fishes and cold water
fishes.
Based upon these data, it is obvious that temperature can influence
tolerance to supersaturation. Unfortunately more work will be needed
to clarify the role of temperature.
Swimming Activity
The effects of swimming activity were studied on juvenile sockeye
salmon and bass in supersaturated water; results for largemouth bass
are shown in Figure 8. Bass kept in cages in the tank center experienced
very little current or need to swim and 501 survived for 26 hours. But
bass that were placed concurrently into the outside portion of the tank
(higher velocity) took 25% less time to reach median mortality. Juvenile
sockeye died at about the same rate regardless of their location in the
tank. Increased mortality resulting from activity was seen in other
tests. Those individuals which were extremely active and often jumped
out of the water usually died before those which remained relatively
quiet.
Sex
In some circumstances differences in tolerance to supersaturation
varied between sexes for adult fish (Figure 9). Tolerance was different
between males and females when (1) the rate of mortality was low enough
39
-------
100
ȣ>.
O
80
>- 60
H
cr
O 40
V-
z
yj
020
UJ
ISC
RESTING
16
18 20 22 24
TIME TO DEATH (hours)
26
Figure 8. Effects of swimming activity on the mortality of
largemouth bass (Microjpterus salmoides) in water
supersaturated to 140%.
-------
20 30
40 50 60
TIME (hours)
70 80 90 100
Figure 9. Effect of percent saturation and sex on mortality
of adult coho salmon (Oncorhynchus kisutch) at IOC
in tanks 65 cm deep.
-------
to permit its expression, and (2) if significant collateral differences
also existed. For example, female and male salmonids both died very
rapidly at or above saturation levels of 125% and there was no apparent
difference in tolerance. Also, no apparent difference was evident at
120% or lower when male and female fish experienced comparable conditions,
i.e. migrating spring chinook salmon. However, when the males and
females were not in comparable condition, the result was a difference
in tolerance to supersaturation. For example, male and female winter
steel head showed a difference in tolerance, but this was probably due
to fighting among the males which weakened them. Female coho salmon
were less tolerant than males and we speculate that this difference was
due to the weakened condition of the gravid females (Figure 10).
Largemouth bass did not die except at high levels of supersaturation
which may have precluded the expression of sexual differences. No
differences were noted in the rate or total mortality between male and
female largemouth bass tested at 130% saturation and 20 C.
Life Stage
Figure 10 shows the concurrent times to mortality among juveniles
(parr), post smolts, jacks and adult coho salmon. Median mortality
was reached first by adults, then by jacks, and last by parr at 120%
saturation. At 115% saturation proportionally fewer adults were killed
and essentially no juveniles were killed.
Body Size and Condition Coefficient
Body size and/or condition had a significant influence on time to
death in some instances. Correlation analyses were conducted for log-|g
hours to death on body weight, body length, and condition coefficient
(K) for adult coho. The only significant correlation at 125% saturation
was between the length of female coho and logio hours to death
(r = 0.93); larger fish took longer to die. These fish were essentially
gravid and ready to spawn. Juveniles were not tested.
Adult spring chinook had a significant correlation between log-|g
time to death and body weight (Table 12). While these results have
statictical significance and indicate that larger adult salmon tend to
survive longer in supersaturated w?ter, the overall impact seems
relatively small and questionable. For example, Meekin and Turner (1974)
determined that the size of juvenile Pacific salmon was inversely related
to survival in gas supersaturated water; large juveniles died sooner
than their smaller siblings. Perhaps fat content influences this as
indicated by Boycott and Datnant (1908) and Gersh, Hawkinson, and Rathbun
(1944).
42
-------
smolts jacks adults
125% 125% 125%
9adults juveniles
120% 125%
adult males smolts
120% 120%
_J
cr
o
t-
Ld
O
oc
LU
a.
100
90
80
70
60
50
40
30
20
10
u/x
1 1
jacks
120%
juveniles
120%
20 4O 60
80 !00 120 140 160 180 200
TIME (hours)
Figure 10. Effect of percent saturation and life stage on
mortality of coho salmon (Oncorhynchus kisutch)
at IOC tanks 65 cm deep.
-------
TABLE 12. CORRELATION ANALYSES OF TIME TO DEATH AT VARIOUS LEVELS OF
SUPERSATURATION TO BODY LENGTH, WEIGHT, AND CONDITION
COEFFICIENT (K) FOR ADULT SPRING CHINOOK (Oncorhynchus tshawytscha).
Total Gas
Saturation Sex
r
Males (9) a
b
130%
r
Females (11) a
b
r
Males (5) a
b
125%
r
Females (15) a
b
r
Males (9) a
b
120%
r
Females (8) a
b
r
Males (7) a
b
120%
r
Females (9) a
b
115% Females (7) r
Body
Length
(cm)
= +0.653*
= -41.18
= +121.74
= +0.480
= +15.823
= +70.628
= -0.674
= +87.384
= -13.838
» +0.225
= +37.143
= +34.320
= +0.978*
= -130.82
= 111.92
= +0.970*
= -90.52
= 81.03
= -0.642
= +177.9
= -61.09
= -0.650*
= -12.132
= +54.793
= +0.837
Body
Weight
(9)
+0.679*
-22117
+29370
+0.349
-6056
+14928
-0.820
+4563
-1184
+0.137
+1221.
+13593
+0.813*
-12818
10825
+.723*
-3294
+6267
-0.624
25421
-17713
+0.659*
-14240
+12712
+0.487
Condition
Coefficient
(K)
+0.278
+1.163
+0.017
-0.307
+1.672
-0.036
+0.946*
-3.745
+0.007
-0.167
+1.207
-0.009
-0.690*
+5.813
-2.416
+0.688*
+1.137
+0.085
+0.047
+1.236
+0.059
-0.252
+1 . 568
-0.152
+0.520
* Significant at 90% level.
44
-------
Other Factors Influencing Tolerance to Supersaturatlon
At least two other factors may influence tolerance to gas bubble
disease based on our observations. The first is the effect of small
but definite increases in dissolved gas pressure induced by solar
radiation and fluctuations in the control system. Slight increases
in saturation seemed to cause increases in mortality in critically
morbid fish; conversely, small decreases in saturation caused decreases
in mortality. Possibly the same effect was caused by changes in depth
which would alter the amount of hydrostatic pressure to compensate
the emboli or emphysema.
A second important factor that is difficult to quantitate is
behavior. The investigators spent many hours watching the test animals
and agree that behavior can have significant impact upon the results.
For example, adult sockeye salmon are schooling fish and tend to
pursue a common mode or level of activity. Other fish such as adult
steelhead tend to be very individualistic and their results seem to
be more variable. Adult chinook tended to remain at maximum depth in
the tanks; if this is the case in nature, the investigators speculate
that these fish would probably find ample compensation by hydrostatic
pressure, if available.
TOLERANCE TO SEAWATER AFTER EXPOSURE TO SUPERSATURATED WATER
Three species of anadromous salmonids were tested for tolerance
to seawater (30 ppt Cl) after exposure to three levels of super-
saturation (110, 115, and 120%).
Exposure to 120% continued for about 47 hours when the following
levels of mortality had been reached: winter steelhead =15%; sockeye
salmon = 45%; spring chinook salmon * 60%; and summer steelhead = 70%
(Figure 11). Immediately thereafter the fish were transferred to air
equilibrated seawater and no further mortality occurred during a
subsequent 96 hour period that could be related to supersaturation or
seawater.
Other groups of smolts were placed in 115% saturation water for
268 hours. During this time 10% of the sockeye smolts and 20% of the
winter steelhead smolts died of gas bubble disease. None of the summer
steelhead or spring chinook smolts died during the 268 hour exposure
to 115% saturation. However, the survivors of each species had some
emphysema in the skin and fins and a few had protruding eyes from
emphysema around or behind the tye. Signs of gas bubble disease were
most prevalent and severe among the winter and summer steelhead. Yet
these fish survived direct transfer to 30 °/oo seawater and continued
to survive for 124 hours when the experiment was ended.
45
-------
lOOi
I 90
E
FRESH WATER AT
120% OF AIR
SATURATION (10%)
TRANSFERRED ABRUPTLY TO
30%SEAWATER
SUMMER STEELHEAD* (S. gairdneri)
SPRING CHINOOK ' ( 0. fshawyfscha)
SOCKEYE (0. nerka)
WINTER STEELHEAD ( S. gairdneri)
END OF
TEST
1
10 20 30 40 50 60 70 a) 90 100 110 120 130 140 150 160 170
HOURS
' NO MORTALITY IN CONTROL GROUPS
Figure 11. Cumulative mortality of salmonid smolts during exposure
to 120% air super saturation and immediate transition to
30°/ooseawater.
-------
No smolts died that were exposed to "110% saturation for 268
hours followed by abrupt transfer to 30 o/oo seawater.
These results confirm our previous pilot studies on coho salmon
smolts and indicate no latent mortality to smolts from gas bubble
disease after they have entered seawater. Essentially identical
results were reported by Dawley, Monk, Schiewe, and Ossiander (1975).
EFFECTS OF CHRONIC SUPERSATURATION ON REPRODUCTION
The effects of supersaturation on reproduction were investigated
only at 110% saturation because higher levels of supersaturation
produced extensive mortality of adults in shallow tanks. The exposure
of 20 adult spring chinook began April 30, 1973. One female developed
partial paralysis early in July and was sacrificed and examined about
three weeks later. Another fish died at about this time when dissolved
gas levels jumped unexpectedly to 116%, but this fish had a heavy
infection in the gills by fungus and parasitic copepods. The test
exposure ended August 3 when it became apparent that further mortality
would result if we did not treat for the infective fungus and copepods.
Thereafter no mortality occurred in either group until spawning operations
began. Apparently the infections had weakened both the control and test
fish to the extent that heavy mortality occurred each time the fish were
handled and inspected for sexual maturity.
Experimental male spring chinook salmon ripened readily and their
semen was normal in appearance and consistency. Likewise, the eggs of
all experimental females developed to typical size and were normal in
color and general appearnace. However, prespawning mortality from
handling the fish eliminated four of the six females in the experimental
lot; extensive mortality also occurred in the control lot.
Two experimental females survived the stress of supersaturation
and subsequent prespawning handling; these ripened and were spawned
artificially. Their eggs were typical in all respects and approxi-
mately 90% of these eggs hatched. This is well within normal limits
and was comparable to the control fish. The results of this limited
sampling revealed no evidence of adverse effects to the sex products
or to the ability of the resulting embryos to hatch.
47
-------
SECTION V
DISCUSSION
Perhaps some readers may want to extrapolate beyond these data
to natural circumstances in a river, reservoir, or lake. We urge
extreme caution in such endeavors because the research design did not
permit the testing of remedial behavioral responses such as avoidance,
sounding, or other aspects resulting in less than continuous exposure
to supersaturation stress.
Behavioral responses are important because they are known to
influence survival in supersaturated water. Dawley, ejtal., (1975),
reported that juvenile fall chinook sustained less than BY mortality
after 60 days of exposure at 120% supersaturation when a maximum depth
of 2.5 m was permitted; greater than 95% mortality occurred in the same
period when siblings were exposed to 120% supersaturation in water 0.25
m deep. Meekin and Turner (1974) reported similar results for juvenile
fall chinook salmon, juvenile coho salmon, and squawfish, but juvenile
steelhead trout did not survive quite as well.
In 5-hour tests Meekin and Turner (1974) reported that juvenile
chinook salmon avoided supersaturated water and congregated in the
control water. Juvenile coho salmon showed no preference during this
short period, which may in part, be related to their great tolerance
for supersaturation. Juvenile sockeye salmon consistently sound to
maximum depth when placed in supersaturated water hence actively avoid
gas bubble disease (Wesley J. Ebel, personal communication). Thus
avoidance responses could exert an extremely important influence upon
the potential lethality of supersaturation in a river.
Tolerance to gas bubble disease appears to involve different
biological factors at high versus low levels of supersaturation. In
acutely lethal conditions, emboli form in the blood and cause rapid
death. Thus survival at acutely lethal conditions may be influenced
heavily by the ability of individual fish to tolerate changes in
vascular dynamics. But survival at long term sub-acute exposures may
be more dependent upon complex alterations of physiological functions
such as immune responses, infectious agents, or adaptive behavior.
Several other factors were identified that influenced tolerance
to supersaturation, including temperature, activity, life stage, and
body size. One of the most important was the difference in tolerance
between fish families, and to a lesser extent, tolerance varied between
species or races. Another obvious difference was that trout and salmon
which are phyostomous were generally more sensitive at 10 C than bass
48
-------
which are physoclistous. One hypothesis to explain this difference
is based on the observation that adult salmon expel! gas frequently
from their swimbladder and sink when dead. Conversely, bass became
bloated with considerable pressure in their gas bladder and usually
floated before death. Possibly the tolerance of bass to supersaturation
was promoted in part, by a concurrent and compensatory increase of
intrabody pressure. Evidence of high intrabody pressure was indicated
in necropsy examinations. When punctured, the heart and swimbladder
emitted an audible release. High intrabody pressure would keep gases
in solution within vital organs of the area adjacent to the body
cavity and thus protect the fish. However, a sudden release of gas
(from the swimbladder of salmon) would tend to sharply diminish intra-
body pressure and promote cavitation of gas or the growth of accumulated
emboli.
Apparent differences in tolerance to supersaturation were related
to genetic stocks, sex, and life stage, but the attendant variation
renders it impossible to determine appropriate significance. §ex was
a factor in two cases, but only when the respective species approached
spawning condition. Adult male winter steelhead were in spawning
condition when tested and fought territorial battles in the tanks which
may have promoted emboli formation and death or decreased resistance.
Whatever the reasons, male steelhead died much sooner than the females.
Adult summer steelhead in spawning condition did not fight and,
although the sample size was small, there was no readily apparent
difference in tolerance between males and females. The other case of
sex-related tolerance occurred with adult coho salmon, where gravid
females died more rapidly than ripe males. Although the evidence is
not clear, indications are that poor physiological condition is a
greater influence on tolerance than sex pjjr se. Physiological
condition may well be the cause of observecT dTfferences in tolerance
to supersaturation between different life stages.
Osmotic and ionic regulation in seawater provides great stress
to smolts of Pacific salmon and steelhead. If these functions are
impaired to any great extent, the fish generally die within a few
days. In this study the survivors of lethal supersaturation apparently
suffered no significant impairment of osmotic and ionic regulation
because they survived direct transfer to seawater (30 °/oo) and no
mortality resulted within a five-day period thereafter. This result
was confirmed by Dawley, et_ al_., (1975), and by Bouckl/.
Gas composition in these experiments was not determined beyond
the oxygen level which was typically high. High oxygen-nitrogen ratios
have been shown by Rucker (1974) to prolong survival. Apparently helium
I/ Bouck, 6. R. 1972. Unpublished data on coho salmon smolts.
49
-------
has a similar effect because the decompression period of skin divers
can be reduced 50% when half of the nitrogen in the breathing air is
replaced with helium (Workman, 1963). Gas composition was not controlled
in this experiment and its variability may have influenced the resulting
times to death.
Residual effects on sex products were not readily apparent among
Columbia River spring Chinook salmon which had been held for 3.5
months in shallow water saturated to 110% of barometric pressure.
This is more than twice the exposure period that these fish would
experience in passing through the Columbia River. The resulting hyper-
baric pressure is approximately equivalent to what a chinook would
experience at a depth of about 2 meters in water with 130% saturation.
Natural spawning was not possible in this research, but the sex products
had a good appearance when spawned artificially, and both fertilization
and hatching were within normal limits.
Variability was relatively large for the observed times to median
mortality; this was true even within a life stage of a given race of
a given species. The ET5Q value varied by 100% in some cases. If one
adds to this the potential influence of depth, swimming activity, and
acclimation temperature, it is then easy to hypothesize that data from
most salmonids can be treated as if they were from the same population
without regard to race or species. Thus it appears that the sum of the
factors causing increased or decreased tolerance to supersaturation tend
to overwhelm the influence of a single factor except when it achieves
critical importance on a given occasion.
The interpretation of these data relative to the Columbia and Snake
Rivers requires the consideration of many factors. Wild fish live in
a dynamic circumstance and are free to alter their depth which, in turn,
alters their compensation by hydrostatic pressure. Likewise, in nature
intermittent exposure is probably the rule thus permitting higher
tolerance than indicated by these data. Also, the potential exposure
time of fish must be considered. For adult salmon in the Columbia River
typical passage time between the ocean and safe tributary streams is
probably three weeks or less.
There is no question that the three species of Pacific salmon,
racial variants of steel head trout, and even bass, can get into
serious trouble if they remain in shallow water at high saturation
levels for over a day. This has considerable significance in fish
hatcheries and laboratories where exposure to supersaturation may be
for a longer period and in shallow water. What must be determined is
the extent that these data apply to a large and complex river ecosystem.
Supersaturation appears to be exerting ecological selection
pressure in the Columbia River Basin (Ebel, et. al_., 1974), but its
direction and final result are not evident yet. Thus far, most
50
-------
researchers have investigated only the detrimental effects of super-
saturation, describing supersaturation only in terms of morbidity and
mortality. But broader ecological effects are potentially possible
and remain undescribed. For example, a low level of supersaturation
(or intense supersaturation at certain times of the year) might be an
effective tool in eradicating rough fish populations. From this
aspect supersaturation has the distinct advantage of leaving no
residue, harmful or otherwise. Likewise, if supersaturation kills
bass fry and inhibits predation by squawfish, the net result could
benefit salmon. Further, if supersaturation selectively eliminates
small adults from breeding populations, the net result might be
larger salmon therefore better adapted to the Columbia River. This
is also suggested by the studies of Cramer (1974).
While a supersaturation level of 110% was not acutely lethal,
it did produce low level mortality and signs of ill health in adult
and juvenile spring chinook salmon during a 3.5 month laboratory
exposure in shallow water conditions. Such a prolonged exposure
(especially in shallow water) would be unlikely for wild chinook,
but may occur in fish culture stations or possibly in the outfalls of
heated effluents. The latter could be particularly important when
seasonal temperature cycles dropped beyond the preferred temperatures
of local fish species. In such cases, fish congregating in the warmer
water might die or be adversely affected by gas bubble disease.
Therefore, a saturation level of 110% might be acceptable for a river,
but may not be acceptable for heated effluents.
51
-------
SECTION VI
REFERENCES
Anon. 1971. Standard Methods for the Examination of Water, Sewage
and Industrial Wastes. 13th Edition. Amer. Public Health
Assoc., Inc., NY 1350 pp.
Bailey, R. M., J. E. Fitch, E. S. Herald, E. A. Lachner, C. C. Lindsey,
C. R. Bobins, and W. B. Scott. 1970. A List of Common and
Scientific Names of Fishes from the United States and Canada.
Third Edition. Amer. Fish. Soc., Washington, D.C. 150 pp.
Bouck, G. R. 1972. Effects of supersaturation on Columbia River salmon.
Symposium: Ecological Society of America. August 31, 1972,
Minneapolis, MN (mimeo).
Bouck, G. R. 1973. Total dissolved gases (supersaturation). ITK
Water Quality Criteria 1972, a report of the Committee on
Water Quality Criteria, Nat. Acad. Sci., Nat. Acad. Engr.
EPA-R3-73-033-March, 1973.
Boycott, A. E. and G. C. C. Damant. 1908. Experiments on the influence
of fatness on susceptibility to caisson disease. J. Hyg.
8: 445-456.
Cramer, S. P. 1974. The heritability of resistance to gas bubble
disease in Columbia River fall chinook salmon (Oncorhynchus
tshawytscha). M.S. Thesis, Oregon State University 59 pp.
D'Aoust, B. G. and L. S. Smith. 1974. Bends in Fish. Comp. Biochem.
Physio!. 49: 311-321.
Dawley, E., B. Monk, M. Schiewe, and F. Ossiander. 1975. Salmonid
bioassay of supersaturation of dissolved gas in water. National
Marine Fisheries Service, Northwest Fisheries Center, Seattle,
Washington, 98112. Final Report - December, 1974 (to be
published - 1975.
Ebel, W. J. 1969. Supersaturation of nitrogen in the Columbia River
and its effect on salmon and steelhead trout. U.S. Fish and
Wildlife Service, Bur. Comrn. Fish. Bull. 68(1): 1-10.
Ebel, W., H. L. Raymond, G. E. Monan, W. Farr, and G. K. Tononaka. 1974.
Effect of atmospheric gas supersaturation caused by dams on
salmon and steelhead trout of the Snake and Columbia River:
A review of the problem and the progress toward a solution.
National Marine Fisheries Service, Seattle, WA. Ill pages (mimeo).
52
-------
Fickeisen, D. H., M. J. Schneider, and J. C. Montgomery, 1974. A
comparative evaluation of the Weiss Saturometer. Manuscript
submitted for publication by Battelle Northwest Memorial
Laboratory, Richland, WA.
Gersh, I., 6. E. Hawkinson, and E. N. Rathbun. 1944. Tissue and
vascular bubbles after decompression from high pressure
atmospheres correlation of specific gravity with
morphological changes. J. Cell. Comp. Physio!. 24: 35-70.
Harvey, H. H. 1975. Gas disease in fishes --a
and Physics of Agueous Gas Solutions, p.
Electrochemical Society, Princeton, NJ.
Leitritz, E. 1959. Trout and salmon culture.
Dept. of Fish and Games, Sacramento, CA.
review.
450-485.
In: Chemistry
Publ. The
Bull. 107 Calif.
169 pp.
Mackay, R. S. and G. Rubissow. 1971. Detection of bubbles in tissues
and blood, pp. 151-160. _In_: Underwater Physiology, C. J.
Lambertson, Editor. Academic Press, NY.
Meekin, T. K. and B. K. Turner, 1974. Tolerance of salmonid eggs,
juveniles and squawfish to supersaturated nitrogen, pp. 78-126.
In: Nitrogen Supersaturation Investigations in the Mid-Columbia
River. Tech. Rept. 12. Washington Dept. of Fisheries, Olympia,
WA.
Rucker, R. R. 1972. Gas-bubble disease of salmonids: A critical
review. Technical Paper No. 58 of the Bureau of Sport Fisheries
and Wildlife, Seattle, WA. 11 pages.
Rucker, R. R. 1974. Gas-bubble disease of salmonids: Variation in
oxygen-nitrogen ration with constant total gas pressure.
National Marine Fisheries Service, Seattle, WA. 12 pages (mimeo).
Snedecor, G. W. and W. G. Cocbran. 1973. Statistical Methods. 593 pages.
Iowa State University Press, Ames, IA.
Stroud, R. K., G. R. Bouck and A, V. Nebeker. 1975. Pathology of acute
and chronic exposure of salmonid fishes to supersaturated water.
pp. 435-449. jn: Chemistry and Physics of Aqueous Gas
Solutions. Publ. The Electrochemical Society, Princeton, NJ.
Stroud, R. K., and A. V. Nebeker. 1975. A study of the pathogenesis
of gas bubble disease in steelhead trout. Proc. Gas-Bubble
Disease Workshop, Battelle Northwest, Richland, WA.
53
-------
Weitkamp, D. E. and M. Katz. 1973. Resource and Literature Review
of Dissolved Gas Supersaturation in Relation to the Columbia
and Snake River Fishery Resources. Seattle Marine Laboratory,
Seattle, WA. 55 pages.
Wolke, R. A., G. R. Bouck, and R. K. Stroud. 1974. Gas-bubble disease:
A review in relation to modern energy production, p. 239-266.
In: Fisheries and Energy Production, S. B. Saila, Editor.
Lexington Books, Lexington, MA.
Workman, R. D. 1963. Studies of decompression and inert gas-oxygen
mixtures in the U.S. Navy. pp. 22-28. In: Underwater
Physiology Symposium, C. J. Lambertson and L. C. Greenbaum, Jr.
Editors, Pub!. No. 1181, Nat. Acad. Sci., Nat. Res. Council,
Washington, D.C.
54
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/3-76-050
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
MORTALITY, SALTWATER ADAPTATION AND REPRODUCTION
OF FISH DURING GAS SUPERSATURATION
5. REPORT DATE
May 1976 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Gerald R. Bouck, Allen V. Nebeker, and Donald G. Stevens
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Western Fish Toxicology Station*
US Environmental Protection Agency
1350 SE Goodnight Ave
Corvallis, Oregon 97330
10. PROGRAM ELEMENT NO.
U3A608
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Duluth, Minnesota 55804
13. TYPE OF REPORT AND PERIOD COVERED
Final 1970-1972
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
*Now attached to the Corvallis Environmental Research Laboratory, Corvallis, Oregon
16. ABSTRACT
Tests were conducted using continuous exposure in shallow water at levels of total
dissolved gas pressure ranging from 110-1^0% of barometric pressure (hyperbaric
pressure * 103-410 g/cm^). Both times to 20% and to median mortality were determined
on several life stapes of Pacific salraonids (Oneorhynchus and Salmo) and Largernouth
bass (Micropterus salmoides). Mean times to 207, mortality at 115% total gas sat-
uration were 309, 154 and 125 hours for adults, smolts and parr. At 120% saturation
mean times to 207. mortality were 48, 41 and 53 hours for adults, smolts and parr,
At 125 7, saturation, mean times to 20% mortality decreased to 18, 17, and 24 hours
for adults, smolts, and parr. Factors which influenced time to death included
genera, life stage, acclimation temperature, activity level, sex and body size.
Mortality curves were typically skewed to the right. Gross pathology of gas bubble
disease was described relative to these experiments. High gas levels that killed
50% of three species of salmon smolts had no apparent effect on the ability of the
survivors to tolerate an immediate transfer into seawater (30 ppt Cl). Lonp-term
(3-month) continuous exposure of adult spring chinook salmon to 110%, saturation
had no readily apparent adverse impact on the fertilization and hatching of their
eggs.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
temperature, animal physiology, salmon,
trout, environments, eye diseases, gas
dynamics, pathophysilogy, water pollution,
fresh water, supersaturation
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Cas bubble disease
06/c/f
07/b
13/,1/k
18. DISTRIBUTION STATEMENT
Unlimited release
19, SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
65
20. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
,5-U.S. GOVERNMENT PRINTING OFFICE: 1976-657-695/5<)2lt Region No. 5-1 I
55
------- |