~1flirX<uiJL<i./

f-

ILLINOIS

NATURAL HISTORY

SURVEY

=\

Effects of Cold Temperature on Toxicity

of Ammonia to Rainbow Trout,

Bluegiils, and Fathead Minnows

Contract Report 68-01 -5832/B

Center for Aquatic Ecology

Keturah A. Reinbold

and
Stephen M. Pescitelli

October 1982
Reprinted January 1990

\=

Aquatic Ecology Technical Report
Contract 68-01-5832/B

Illinois Natural History Survey
Aquatic Ecology Technical Report
Contract 68-01 -5832/B

Effects of Cold Temperature on Toxicity of Ammonia
to Rainbow Trout, Bluegills, and Fathead Minnows

by

Keturah A. Reinbold

and
Stephen M. Pescitelli

Center for Aquatic Ecology
linois Natural History Survey
607 E. Peabody Dr.
Champaign, IL 61820

Final Report to U.S. Environmental Protection Agency

Region V, Chicago, Illinois

Walter Redmon, Project Officer

October 1 982
Reprinted January 1990

ABSTRACT

The acute toxicity of un-ionized ammonia to rainbow trout (Salmo gairdneri), blue-
gill (Lepomis macrochirus), and fathead minnow (Pimephales promelas) was determined
under flow-through conditions at low temperatures (3-5°C) typical of winter conditions and
at higher temperatures typical of summer conditions for each species. The purpose was to
determine whether ammonia toxicity differs under different seasonal temperature condi-
tions. The 96-h LC50 values from replicate tests with rainbow trout averaged 0.47 mg/L
un-ionized ammonia nitrogen (NH3-N) at 3-5°C and 0.76 mg/L NH3-N at 13-15°C. For
bluegill, LC50 values averaged 0.32 and 1.35 mg/L NH3-N at 4-5°C and 24-25°C, respec-
tively, while at the same temperature ranges LC50 values for fathead minnow were 0.60
and 1.17 mg/L NH3-N, respectively. Thus, across the temperature span experienced from
summer to winter for each species, bluegill appeared to be the most sensitive of the three
species to the effect of low temperature on ammonia toxicity.

This report was submitted in fulfillment of Contract No. 68-01-5832 by the Illinois
Natural History Survey under the sponsorship of the Environmental Protection Agency.
Work was completed in December 1981.

CONTENTS

Page

Acknowledgments iv

I. Introduction i

II. Conclusions i

III. Recommendations 2

IV. Materials and Methods 2

Test organisms 2

Dilution water 3

Exposure systems 5

Analytical procedures 5

Test procedures 6

Data analysis 6

V. Results 6

Rainbow trout - 8

Bluegill 9

Fathead minnow 9

vi. discussion 10

References 12

Appendix A 13
Tables

1 . Chemical characteristics of dilution water 4

2. Test conditions and age, length, and weight of test fish 7

3 . The 96-h LC50 values for un-ionized and total ammonia nitrogen
for rainbow trout, bluegill, and fathead minnow at winter and

summer temperatures 7

4 . Ratios of warm-temperature versus cold-temperature 96-h LC50

values for ammonia for rainbow trout, bluegill, and fathead minnow 10

ui

ACKNOWLEDGMENTS

We are grateful to Dr. Richard Sparks for providing valuable suggestions during
this investigation and for reviewing the final report. We are grateful to Mr. Dennis Leonard
of Anaconda Company, Denver, CO, and to Mr. Ron Phillips of Becker Industries, Inc.,
Newport, OR, for providing the clinoptilolite used to remove background ammonia from
the dilution water. We thank Dr. Rosemarie Russo, USEPA, Duluth, MN, for providing
the computer calculation of LC50 values. During the study we were fortunate to have the
laboratory assistance of Diane Lynn, Karen Scrogum, Gaylene and Angela Boerger, and
Karin Kuhl, who performed many of the tasks necessary in conducting the tests. Many of
the water quality and metal analyses were performed under the direction of Drs. Allison R.
Brigham and Suzanne Wood, respectively. Ms. Jana Waite retyped, formatted, and edited
the 1990 version of this report.

We are indebted to the following individuals for supplying rainbow trout used as
test organisms: Mr. Rodney Homer, Illinois Department of Conservation, and Mr.
Gordon Proctor, Shepherd of the Hills Hatchery, Missouri Department of Conservation.

Our sincere appreciation is extended to Mr. Walter Redmon, Project Officer,
USEPA Region V, Chicago, IL, for his patience and constructive advice during the study.

IV

I. INTRODUCTION

Ammonia is an important pollutant in natural waters both as a toxicant and as an
oxygen-demanding material. It commonly occurs in municipal and industrial waste
discharges and in runoff from agricultural feed lots. Aqueous solutions of ammonia are
very toxic to fishes under certain environmental conditions, and toxicity is primarily attri-
buted to the un-ionized form. The percentage of total ammonia occurring in the un-ionized
form depends on pH, temperature, and to a lesser extent, ionic strength.

Frequency of occurrence and the high cost of treatment for adequate removal have
made the issue of water quality criteria for ammonia very difficult. It is essential to develop
a broad data base of ammonia toxicity to fish at representative water quality and environ-
mental conditions.

It has been proposed that allowable limits for ammonia discharges be increased
during winter, based on decreased oxygen demand from nitrification at lower tempera-
tures. On the other hand, there is some evidence that toxicity of un-ionized ammonia to
fish may be increased at lower temperatures. Brown (1968) found that toxicity of ammonia
to rainbow trout is almost twice as high at 3°C as at 10°C. Data for channel catfish show a
similar trend, although at higher temperatures than are pertinent to winter conditions. For
fingerling channel catfish, Colt & Tchobanoglous (1976) found that acute toxicity of un-
ionized ammonia was lower at 30°C than at 22°C; Roseboom & Richey (1977) also found
that this species is less sensitive to ammonia at 28°C than at 22°C in 96-h tests.

This information from the literature is insufficient to ascertain differences in effects
over a range of temperatures representative of all seasons. Winter temperature data have
been reported only for rainbow trout, and many streams do not contain salmonids. Catfish
data apply only to summer conditions.

The research reported here was undertaken to begin to fill the gap in information on
the relative toxicity of ammonia under cold-temperature conditions. The objective of this
study was to determine the toxicity of ammonia to three fish species under winter tempera-
ture conditions versus under summer conditions for each species.

II. CONCLUSIONS

Rainbow trout, bluegill, and fathead minnow are 1.2-5.2 times more sensitive to
un-ionized ammonia at low temperatures (3-5°C) typical of winter conditions than at
higher temperatures typical of summer conditions for each species (13-15°C for
rainbow trout and 24-25°C for the non-salmonids).

Across the temperature span experienced from summer to winter for each species,
bluegill appeared to be the most sensitive of the three species to the effect of low
temperature on ammonia toxicity.

III. RECOMMENDATIONS

The increased toxicity of un-ionized ammonia to fish at low temperatures should be
considered in setting water quality standards. In particular, this increase in
ammonia toxicity at low temperatures should be considered when evaluating
requests to allow increased ammonia discharges during winter on the basis that, at
low temperatures, oxygen demand from ammonia is decreased and the percentage
of total ammonia in the un-ionized form is decreased. Results of this study show
that, although the proportional fraction of total ammonia in the un-ionized form is
less, toxicity to fish can be greater at low temperatures.

IV. MATERIALS AND METHODS

TEST ORGANISMS

Toxicity of ammonia was measured for juveniles of three fish species: rainbow
trout (Salmo gairdneri), bluegill (Lepomis macrochirus), and fathead minnow (Pimephales
promelas).

Fish used in toxicity tests were obtained from hatcheries or from cultures main-
tained in outdoor ponds at the Illinois Natural History Survey (INHS), Champaign. All
species were gradually acclimated in the laboratory to the temperature and photoperiod at
which the toxicity test was subsequently conducted. During acclimation, changes in photo-
period were usually <15 min/d and never more than 30 min/d. Temperature changes were
limited to 3°C in any 24-h period unless otherwise stated.

Rainbow trout were obtained from two sources. Trout used in test I were hatched
in facilities of the Illinois Department of Conservation at Havana on 22 December 1980
from eggs (Jalpo River strain) obtained from the state fish hatchery at Arlee, MT. Fish
were transported to INHS on 19 March 1981 in approximately 375 L of well water in
which they had been reared. Transport water was gradually replaced with laboratory
dilution water in a flow-through system with 95% replacement of water in 30 h (Sprague
1973) and were acclimated in the laboratory for 1 month before testing.

These trout were reared at 13-14°C in Havana and the transport water was 12°C
when they reached the laboratory. They were held at 12-12.9°C for 2 wk, after which they
were separated into two groups. One group was held at 12.1-13°C in the laboratory where
the warm-temperature test was conducted while the other group was moved into an
environmental chamber where the cold- temperature test was conducted. In the environ-
mental chamber, water temperature was gradually reduced to 5°Cover 18 d before the test.
These fish were tested at an age of 4 months after hatch. Photoperiods for warm- and cold-
temperature groups were gradually adjusted to 16 and 1 1 h, respectively.

The fish weighed 5-6 g each on 19 March. They had been fed Glencoe Mills ration
at the hatchery and were fed Glencoe Mills No. 4 crumble in the laboratory. They were fed
seven times daily beginning 19 March; feedings were gradually reduced to three times
daily over the next 4 wk as body weight increased. Total weight of food per individual per
day was 5% of body weight.

For tests II and III, rainbow trout were obtained from the Missouri Department of
Conservation's Shepherd of the Hills hatchery at Table Rock Dam near Branson, MO. The
fish hatched in late April and were shipped by air to Champaign on 13 May. Total transit
time was approximately 6 h. After arrival at the laboratory, dilution water was gradually
added to the transport water until the fish were in predominantly dilution water after 4 h.

These trout weighed approximately 0.5 g each on 13 May. They were fed Glencoe
Mills open formula diet at the hatchery and were continued on the same diet during acclima-
tion; they were fed seven times daily at first with a gradual reduction to five times daily
after 2 wk.

These trout were hatched and reared at 7°C. On arrival at INHS the transport water
was 0.5°C. The temperature was gradually raised to 5.8°C over the following 4 h. Fish
were then placed in a 200-L aquarium in an environmental chamber with a constant flow of
dilution water and were held at 5.4-6.3°C for the next 6 d. Approximately half of the fish
were then transferred to the laboratory where the warm-temperature test was conducted.
Initial water temperature was 7°C. Over the next 8 d, the water temperature for the warm-
temperature test group was gradually raised to 13°C. Photoperiods were set to 16 and 12 h
for warm- and cold-temperature test groups, respectively.

Juvenile bluegill (no hybrids included) and fathead minnow were collected by
seining ponds at INHS and were acclimated in the laboratory. Timing and duration of
funding for this study made it necessary to collect these two species during the summer
and to acclimate them to low temperatures prior to cold-temperature tests. .Fish were
brought into the laboratory in pond water. Initial water temperatures were 27-28°C and
were gradually reduced to 24°C over a 2-h period. Bluegill and fathead minnow were each
divided into two groups and acclimated separately for cold- and warm- temperature tests.
Warm-temperature groups were held at 24-25°C while cold-temperature groups were accli-
mated to 4-5°C

For fathead minnow test I, the water temperature for the cold-temperature test group
was reduced from 27 to 4.3°C over 9 d. On two occasions, difficulties in adjusting the
temperature control apparatus caused the temperature change to exceed 3°C in 24 h— a
change of 4.7°C (22.6 to 17.9°C) and 4.5°C (14.5 to 10°C). Fish for bluegill test I and for
fathead minnow test II were gradually acclimated from 27 to 4°C over 5 wk. A temperature
change of more than 3°C in 24 h occurred once for each species. The decrease was 3.8°C
for bluegill and 3.4°C for fathead minnow. The cold-temperature group for bluegill test II
was acclimated from 24 to 4.8°C over 19 d. A change of 5.5°C occurred during one 24-h
period and 3.6°C on another day.

Photoperiods were adjusted to 16 and 12 h for warm- and cold-temperature test
groups, respectively. Both fish species were fed newly hatched brine shrimp nauplii
(Anemia) hatched in the laboratory from eggs available commercially (San Francisco Bay
Brand, Metaframe, Inc.).

DILUTION WATER

Dilution water was taken from municipal wells at depths of 220-370 ft in the
Mahomet-Teays aquifer near Champaign-Urbana. Water was passed through two in-line
charcoal filters to remove chlorine, through clinoptilolite when necessary to remove back-
ground ammonia, and through an ultraviolet sterilizer to eliminate microorganisms. It was

then delivered through PVC pipe to stainless steel holding tanks (670 and 340 L for warm-
and cold-temperature tests, respectively). Sodium thiosulfate was metered into each hold-
ing tank to remove any trace of chlorine that might remain after charcoal filtration. In addi-
tion, a dilute solution of hydrochloric acid was metered into the smaller holding tank to
reduce the pH to that in the larger holding tank. Characteristics of the dilution water are
listed in Table 1.

Table 1 . Chemical characteristics of dilution water.
All values are in mg/L unless otherwise stated.

Total alkalinity (as CaCC-3)

10.7(8.95-11.33)

Hardness (as CaCC>3)

74 (69-76)

Conductivity (umhos/cm at 25°C)

324 (309-332)

Nitrate-N

<0.03

Nitrite-N

<0.01

Soluble orthophosphate

<0.01

Total residue

173 (165-183)

Chloride

13.7 (8.1-17.35)

Sulfate

9.7 (7.6-13.5)

Total dissolved solids (as NaCI)

240 (229-247)

COD

4.9 (1.8-9.35)

Cyanide

<0.001

Al

<0.031

As

0.029

Ca

13.1

Cd

<0.019

Co

<0.004

Cr

<0.014

Cu

<0.009

Fe

0.013

Hg

<0.00007

K

2.34

Mn

<0.007

Na

38.2

Ni

<0.011

P

<0.062

Pb

<0.032

Se

<0.026

Si

3.04

Zn

<0.01

Residual chlorine

<3

Dilution water was aerated in the holding tanks. The water temperature for the
warm-temperature tests was controlled in the tank by a thermistor in conjunction with two
solenoid valves that allowed hot or cold water to pass through a water jacket surrounding
the tank. For cold-temperature tests, the holding tank and lid were surrounded with
fiberglass and foam insulation and the temperature was controlled with a portable cooling
unit (Blue M Electric Company).

EXPOSURE SYSTEMS

For each test, dilution water was pumped through PVC pipe from the holding tank
to a 0.5-L proportional diluter, modified from Mount & Brungs (1967) and Lemke et al.
(1978); the diluter was used to deliver a logarithmic series of five ammonia concentrations
and a control through mixing chambers to two replicate test aquaria. Filling of the valve
bucket tripped a microswitch which, in conjunction with an electronic timer, controlled a
solenoid valve in the water supply line to provide a flow of 0.25 L of water to each test
chamber every 6 min for cold- temperature tests (every 3 min for trout test I) and every 3
min for warm-temperature tests. Holding tanks and diluters were cleaned prior to each test
to remove any bacterial build-up.

Reagent-grade ammonium chloride was used as the toxicant. Stock solutions were
prepared in glass-distilled water and delivered to diluters from a Mariotte bottle. The pH of
the stock solution was adjusted to that of the dilution water with a sodium hydroxide
solution.

The test chambers used for fish were constructed of glass and silicone sealant.
Each aquarium measured 30 x 40 x 20 cm, had an overflow outlet at a height of 25 cm, and
contained a volume of 20 L. Test aquaria were placed in a stratified random arrangement
and were maintained in circulating water baths at the recommended temperature for the
species being tested. For cold-temperature tests, the proportional diluter and the test
chambers were placed in a walk-in environmental chamber (2.75 m L x 2.2 m W x 2.2 m
H) which was maintained at the desired test temperature.

Photoperiod was automatically controlled for all tests using a combination of
incandescent and fluorescent bulbs. For all warm-temperature tests, a 16-h photoperiod
was maintained, including a 30-min gradual brightening and dimming to simulate dawn and
dusk. In keeping with natural seasonal conditions, a shorter photoperiod was maintained
during cold- temperature tests. An 11-h day was used during rainbow trout test I and a 12-
h day for all other cold-temperature tests.

ANALYTICAL PROCEDURES

Water quality parameters were measured using standard methods (American Public
Health Association et al. 1976, U.S. Environmental Protection Agency 1979). Water
samples were taken from the center of each test chamber. Total ammonia nitrogen concen-
trations were measured at least 4 d/wk and other chemical parameters at least once during
each test.

Total ammonia nitrogen concentrations were determined by the phenate method
(American Public Health Association et al. 1976) using a standard curve prepared by linear
regression. Colorimetric measurements were made with a Coleman 124D double-beam
spectrophotometer. Un-ionized ammonia nitrogen (NH3-N) concentrations were deter-
mined from total ammonia nitrogen, pH, and temperature, using the tables of Thurston et
al. (1979). The pH in each test chamber was determined at least daily with an Orion 701 A
digital pH meter. Dissolved oxygen was measured with an oxygen-specific electrode cali-
brated to titration accuracy (Altex 0260 oxygen analyzer by Beckman). Oxygen measure-
ments were made daily during rainbow trout tests and at least twice each week during other
tests.

Hardness, nitrate nitrogen, nitrite nitrogen, and soluble orthophosphate were deter-
mined using a Technicon Autoanalyzer (U.S. Environmental Protection Agency 1979).
Other water quality parameters, such as alkalinity, conductivity, and COD, were deter-
mined according to analytical procedures described in American Public Health Association
et al. (1976). Analyses of metals in the dilution water were performed by induction-
coupled argon plasma spectrometry (American Society for Testing and Materials 1980).

TEST PROCEDURES

Methodology for these tests generally followed that in "Methods for acute toxicity
tests with fish, macroinvertebrates and amphibians" (U.S. Environmental Protection
Agency 1975). Test organisms were distributed into test chambers one at a time, and with
one exception, were acclimated to the test chamber for 1-2 d while the diluter operated
without toxicant addition. The test was initiated by beginning toxicant addition. For
rainbow trout test HI, the diluter was operated with toxicant addition for 2 d to achieve the
equilibrium toxicant concentration in the test chambers, and the test was initiated when the
fish were added to the test aquaria. When tests with bluegill and fathead minnow were
started, a sufficient quantity of toxicant stock solution was diluted to a volume of 200 mL
with glass-distilled water and added to the test chambers through flow-splitting cells of the
diluter to bring the initial concentrations of ammonia to approximately half of the expected
final concentration. Initial stock solutions were added to prevent acclimation of test organ-
isms to lower concentrations of ammonia before full concentrations were reached.

Fish were fed during the tests to maintain their condition during exposures of up to
2 wk. Feeding was particularly necessary become most fish were small (<1 g body
weight). In addition, changes in feeding behavior were monitored as an indication of
sublethal physiological response to the toxicant. Loss of equilibrium was also monitored.

Mortality was recorded after 1, 3, 6, 12, and 24 h and at least daily thereafter to the
end of the test. Death was determined by lack of gill movement and the lack of response to
gentle prodding.

DATA ANALYSIS

Acute lethal concentration (LC50) values were determined using the trimmed
Spearman- Karber method (Hamilton et al. 1977). When it was necessary to adjust for
mortality in the control, Abbott's formula was used (American Public Health Association et
al. 1976). Toxicity curves (LC50 values versus time) were plotted on log-log graph paper.

V. RESULTS

Test species, conditions during the tests, and mean lengths and weights of test
organisms at the end of each test are listed in Table 2. At least two toxicity tests were
completed with each test species at both summer and winter temperatures. The 96-h LC50
values for both un-ionized and total ammonia nitrogen for all tests are listed in Table 3.

Table 2. Test conditions and age , length, and weight of test fish.

Temperature
(°Q PH

Dissolved oxygen, Number/ Age Mean length Mean weight
ppm (% saturation) chamber (mon.) (mm) (g)

1 5.0
12.8

8.10-8.57
8.02-8.55

Rain
8.0-10.2 (51-88)
4.9- 9.0 (47-85)

bow trout
5
5

4
4

115(103-134)
119 (95-145)

18.1 (12.7-28.4)
20.6 (10.0-32.6)

II 3.0
14.2

8.30-8.59
8.03-8.35

10.9-11.6(86-100)
8.2- 9.4 (76-93)

10
10

1
1

42
45

(32-50)
(35-55)

0.61 (0.23-1.03)
0.86(0.32-1.75)

III 3.3
14.9

8.45-8.76
8.32-8.69

9.1-11.0 (74-95)
7.3- 8.6 (74-87)

10
5

1.5
1.5

44
52

(37-65)
(33-51)

0.76(0.41-3.07)
1.47(0.26-1.31)

I 4.0
25.0

8.32-8.47
7.98-8.25

8.7-12.2(73-100)
6.2- 7.0 (74-83)

Bluegill
10
10

1.5
1.5

19
22

(15-25)
(17-27)

0.08(0.04-0.19)
0.11 (0.05-0.24)

II 4.5
24.8

8.06-8.26
7.98-8.20

10.0-11.3 (87-97)
6.5- 7.5 (74-89)

10
10

2
2

28
30

(21-36)
(23-40)

0.25(0.12-0.56)
0.27(0.15-0.70)

I 4.1
23.9

8.21-8.70
7.86-8.18

Fathead minnow
9.5-12.5 (87-96) 20
6.2- 7.9 (73-79) 20

-

15
16

(9-24)
(9-25)

0.03(0.01-0.12)
0.03(0.08-0.11)

II 4.6
25.2

8.13-8.38
8.01-8.32

9.9-11.6 (88-96)
6.1- 6.6 (73-79)

10

10

-

19
21

(14-34)
(16-28)

0.06(0.01-0.35)

0.07(0.03-0.16)

Table 3. The 96-h LC50 values (95% confidence intervals) for un-ionized and total ammonia
nitrogen (mg/L) for rainbow trout, bluegill, and fathead minnow at winter and summer temperatures (°C).

Temperature

PH

Un-ionized ammonia
Mean (95% CI)

Total ammonia
Mean (95% CI)

Rainbow trout I

5.0
12.8

8.10-8.57
8.03-8.55

0.44
0.64

(0.37-0.51)
(0.59-0.69)

13.17(11.25-15.43)
14.29(13.29-15.36)

Rainbow trout II

3.0
14.2

8.30-8.56
8.03-8.29

0.33
0.84

(0.29-0.39)
(0.79-0.89)

10.58 (9.28-12.06)
17.16(16.35-18.00)

Rainbow trout III a

3.3
14.9

8.45-8.76
8.32-8.69

0.63
0.80

(0.54-0.72)
^0.80-0.80)

12.32(10.88-13.96)
8.49 (8.49-8.49)

Bluegill I

4.0
25.0

8.32-8.47
7.98-8.25

0.42
1.58

0.38-0.47)
1.51-1.65)

13.86(12.53-15.32)
25.19(23.66-26.81)

Bluegill II

4.5
24.8

8.06-8.26
7.98-8.20

0.21
1.12

0.20-0.23)
1.01-1.25)

12.49(11.68-13.36)
18.52(16.42-20.89)

Fathead minnow I

4.1
23.9

8.21-8.70
7.86-8.18

0.59
0.97

0.53-0.66)
0.90-1.04)

15.28 (13.96-16.74)
17.60(16.39-18.89)

Fathead minnow II

4.6
25.2

8.13-8.38
8.01-8.32

0.61
1.36 (

0.51-0.74)
1.26-1.47)

25.07(21.17-29.69)
20.90 (18.84-23.19)

aNumber of individuals not equal at the two temperatures; 10/chamber at 3.3°C and 5/chamber at
14.9°C.

The first two tests with rainbow trout were continued beyond 96 h. Total duration
of test I was 14 d but no additional mortality occurred after 96 h. Test II was continued for
8 d and only one fish at each temperature died after 96 h. For these reasons and because
there is a broad data base of LC50 values of toxicants to fish at 96 h for comparison, 96-h
LC50 values are reported in Table 3. LC50 values at additional time intervals are provided
in the appendix in plots of toxicity versus time for each test. There was a tendency for
initial mortality to occur more quickly at the warmer temperature in the first 24 h, but the
lag in onset of mortality at cold temperatures had disappeared by 96 h.

RAINBOW TROUT

Three tests were conducted with rainbow trout — one with individuals from Illinois
(rainbow trout test I in Tables 2 and 3) and two with trout from Missouri. No control
mortality occurred in any of these tests. Although Illinois trout (about 20 g) were much
larger than those from Missouri (about 1 g), 96-h LC50 values were similar at similar
temperatures. At 5 and 3°C, 96-h LC50 values were 0.44 and 0.33 mg/L NH3-N, respec-
tively. LC50 values at 12.8 and 14.2°C were 0.64 and 0.84 for tests I and II, respectively.
In terms of total ammonia nitrogen, 96-h LC50 values at cold and warm temperatures,
respectively, were 13.17 and 14.29 mg/L in test I and 10.58 and 17.16 mg/L in test II. In
a third test with rainbow trout, fish were added to the test chambers after toxicant concen-
trations had reached equilibrium and the test continued for 96 h. The 96-h LC50 values
were 0.63 mg/L NH3-N at 3.3°C and 0.80 mg/L at 14.9°C. Corresponding toxicities as
total ammonia nitrogen were 12.32 and 8.49 mg/L. In test in, 10 fish were tested per
chamber at 3.3°C and 5 per chamber at 14.9°C because there were insufficient fish accli-
mated to that temperature. However, confidence intervals indicate that the test was valid.
In test I at 12.8°C, the minimum dissolved oxygen level was less than that in other tests
with trout (Table 2). Because sensitivity to ammonia in rainbow trout has been shown to
increase as the dissolved oxygen level decreases (Thurston et al. 1981), the lower LC50
value in test I may have been caused by the lower oxygen level. In all three tests, the
toxicity of un-ionized ammonia to rainbow trout was greater at colder temperatures than at
warmer ones.

In addition to mortality in tests using rainbow trout, sublethal effects were ob-
served. These effects were often observed prior to death but also occurred without
mortality. Loss of equilibrium, resulting in swimming on the side or upside down,
occurred at toxicant concentrations at which little or no mortality occurred. Also, failure to
feed when food was available was observed at intermediate toxicant concentrations. All
three fish species were somewhat less active and fed less under low temperature conditions
than at warmer temperatures, as is to be expected for poikilothermic organisms. Any
decreased activity or feeding reported as a sublethal toxic effect of ammonia is in compari-
son to control fish at the same temperature in the same test.

In rainbow trout test I at 5°C, the lowest lethal toxicant concentration was 0.42
mg/L NH3-N, which caused 40% mortality; all surviving fish at that concentration
suffered loss of equilibrium. Reduced feeding, compared with the control, was observed
at 0.23 mg/L NH3-N. In the same test at 12.8°C, only one fish died at a concentration of
0.54 mg/L NH3-N but most survivors stopped feeding.

Sublethal effects during rainbow trout test II were similar to those of the first test.
At all ammonia concentrations that caused some mortality in 8 d, loss of equilibrium and
cessation of feeding were observed in at least some surviving fish. At 3°C, 0.26 mg/L

NH3-N caused 35% mortality and one survivor suffered equilibrium loss. Among the
survivors at that concentration, only one or two ingested food pellets while others either
made no attempt to feed or took pellets into their mouths but spit them out. At 14.2°C, two
fish died at a concentration of 0.63 mg/L NH3-N, and the remaining individuals suffered
severe to slight loss of equilibrium. At the end of 8 d, all survivors showing loss of
equilibrium were transferred to dilution water with no ammonia to determine if they would
recover. Fish from the warm-temperature test were dead the following day, and those from
the cold-temperature test remained in the same condition throughout several days of obser-
vation.

BLUEGILL

Results of comparative ammonia toxicity tests at winter and summer temperatures
with bluegill were similar to those with rainbow trout In bluegill test I at 4°C, mortality
data were corrected for control mortality, because three fish in the control chambers died.
No control mortality occurred in test n.

Fish at 4.0-4.5°C were more sensitive to un-ionized ammonia than were fish at
about 25°C. The 96-h LC50 values at cold and warm temperatures, respectively, were
0.42 and 1.58 mg/L NH3-N in test I and 0.21 and 1.12 mg/L NH3-N in test II (Table 3).
Corresponding 96-h LC50 values for total ammonia nitrogen at cold and warm tempera-
tures, respectively, were 13.86 and 25.19 mg/L in test I and 12.49 and 18.52 mg/L in test
II.

Sublethal effects were also observed during bluegill tests. At 4.5°C, 10% mortality
occurred at the lowest toxicant concentration (0.15 mg/L NH3-N) and an additional 35% of
the test organisms showed a loss of equilibrium. At 25°C, 85 and 10% mortality occurred
at 1.39 and 0.86 mg/L NH3-N, respectively. Remaining individuals at the higher of these
two ammonia concentrations showed a loss of equilibrium and fed little, while survivors at
the lower concentration took longer to respond to food and fed less than the control fish
and fish at lower ammonia concentrations.

FATHEAD MINNOW

Results of ests with fathead minnows were similar to those for the other two fish
species. However, mortality of control fish was 2.5 and 20% at cold temperatures in tests
I and II, respectively, and 0 and 5% at warm temperatures. Mortality data for these tests
were corrected using Abbott's formula.

This species also was more sensitive to NH3-N at cold temperatures. The 96-h
LC50 values for cold and warm temperatures, respectively, were 0.59 and 0.97 mg/L
NH3-N in test I and 0.61 and 1.36 mg/L NH3-N in test II. Corresponding LC50 values
for total ammonia nitrogen were 15.28 and 17.60 mg/L in test I and 25.07 and 20.90 mg/L
in test II.

Effects of ammonia on equilibrium and on feeding were also observed. At 4°C,
10% mortality occurred in 120 h at 0.17 mg/L NH3-N and another 5% showed loss of
equilibrium. At 25°C, 60% mortality occurred at 1.5 mg/L NH3-N; surviving fish venti-
lated at an obviously higher rate than did controls and they did not feed. At 1.0 mg/L NH3-

N, only 5% mortality occurred but survivors were less active and responded more slowly
to food than did controls.

VI. DISCUSSION

Results of this study show that rainbow trout, bluegill, and fathead minnow are
more sensitive to un-ionized ammonia, as determined by mortality and by sublethal effects,
at low temperatures typical of winter conditions than at higher temperatures typical of
summer conditions for each species. Because a lower percentage of total ammonia is in the
un-ionized form at a lower temperature and the un-ionized form is believed to be the pri-
mary cause of toxicity, it would be expected that, if temperature had no effect on a fish's
sensitivity to ammonia, a higher concentration of total ammonia would be required to pro-
duce a toxic effect as temperature decreases and that LC50 values of NH3-N would be
equal at different temperatures. The LC50 values reported here, however, clearly show
that toxicity of NH3-N is greater at colder test temperatures. The ratios of warm-
temperature versus cold-temperature LC50 values (NH3-N) are shown in Table 4. At 96 h,
un-ionized ammonia was 1.2-5.2 times more toxic at cold temperatures.

Table 4. Ratios of warm-temperature versus
cold-temperature 96-h LC50 values (NH3-N) for
rainbow trout, bluegill, and fathead minnow.

Ratio (warm/cold)

Rainbow trout
1

II
III

1.4
2.6
1.2

Bluegill
1
II

3.7
5.2

Fathead minnow
1
II

1.7
1.9

In some tests, pH values were higher in cold-temperature test chambers than in
warm-temperature chambers. In early tests, difficulties were encountered in maintaining
equal pH levels at the two temperatures. In later tests, however, levels were equalized by
addition of dilute hydrochloric acid to the dilution water used in cold- temperature tests.
Thurston et al. (1981) showed that toxicity of un-ionized ammonia increased at lower pH
values for rainbow trout and fathead minnow. For rainbow trout, 96-h LC50 values were
0.513 mg/L NH3-N at a pH of 7.84 (range of 7.80-7.91) and 0.658 mg/L at a pH of 8.29
(8.24-8.40).

10

In this study, differences in pH at the two temperatures were either of similar
magnitude to the example cited or less. In cases of greatest pH difference1, if the pH at the
low temperature had been the same as that at the higher temperature, the result would
probably have been an even greater toxicity of ammonia at the low temperature. Thus, the
difference in toxicity at the two temperatures would have been greater than that observed
and the ratios in Table 4 would have been larger.

In rainbow trout tests, pH differences at the two test temperatures severe relatively
small. Ratios (warm/cold) of LC50 values (NH3-N) of rainbow trout invests I, II, and III
were 1.4, 2.6, and 1.2, respectively, which compare favorably with Brown (1968) who
reported that toxicity of ammonia to rainbow trout is almost twice as high at 3°C as at 10°C.

In tests with bluegill and fathead minnow, pH differences between the two tempera-
tures were small during the second test with each species. In the first test, however, pH
was lower at the warm temperature. The pH ranges at the two temperatures did not overlap
and mid-points of the ranges differed by 0.28 and 0.26 pH unit for bluegill and fathead
minnow, respectively. These pH differences may, in fact, have affected the results. In
bluegill test I, un-ionized ammonia toxicity at cold temperatures was 3.7 times that at warm
temperatures and 5.2 times in test II (Table 4). Similarly, for fathead minnow toxicity at
cold temperatures was 1.7 and 1.9 times greater than that at warm temperatures in tests I
and II, respectively. Thus, the smaller warm/cold ratios of LC50 values in the initial tests
could have resulted, at least in part, from pH differences.

Dissolved oxygen levels may also have affected results in the first .test with rainbow
trout. Fish used in test I were much larger than those used in the other two tests, making it
more difficult to maintain an optimum dissolved oxygen level. At the warmer temperature,
dissolved oxygen dropped to as low as 5.0 ppm (Table 2), which was lower than that in
other tests with trout. Thurston et al. (1981) showed that, for rainbow trout, LC50 values
for ammonia decrease with any decrease in dissolved oxygen, with a 30% decrease in
tolerance at 5.0 compared to 8.5 ppm dissolved oxygen. This lower oxygen level could
account for the lower LC50 value in test I compared with the other two tests (0.6 versus
0.8 mg/L NH3-N).

Bluegill appears to be more sensitive than fathead minnow to ammonia toxicity at
cold temperatures. Over the same temperature range, the cold temperature caused an in-
crease in toxicity of un-ionized ammonia to bluegill that was more than twice that for
fathead minnow. The ratio of increased toxicity at cold temperatures in this study was
similar for rainbow trout and fathead minnow, but rainbow trout were tested over a
narrower temperature range.

Sublethal toxic effects of ammonia (i.e., loss of equilibrium and reduction in
feeding) were observed in this study at ammonia concentrations causing only low percent-
ages of mortality, or in the case of rainbow trout, no mortality. Under natural conditions,
organisms suffering loss of equilibrium or reduced activity would probably be unable to
avoid predation; those that stop feeding might eventually starve to death. Organisms so
affected probably should be considered "dead." It is possible that fish might recover if they
were no longer exposed to ammonia. However, affected rainbow trout in this study after 8
d of exposure to ammonia did not recover when transferred to clean dilution water. Sub-
sequent mortality of rainbow trout exposed at 14.2°C may have resulted in part from the
stress of handling, however, rather than solely from effects of ammonia.

In summary, the results of this study show that rainbow trout, bluegill, and fathead
minnow are more sensitive to un-ionized ammonia at low temperatures (3-5°C) typical of

11

winter conditions than at higher temperatures typical of summer conditions. Across the
temperature span experienced from summer to winter for each species, bluegill appeared to
be the most sensitive of the three species to the effect of low temperature on toxicity of
ammonia.

REFERENCES

American Public Health Association, American Water Works Association, and Water Pollution Control

Federation. 1976. Standard methods for the examination of water and wastewater, 14th ed.

American Public Health Association, Washington, DC.
American Society for Testing and Materials. 1980. Annual book of standards. Part 31, Water. American

Society for Testing and Materials, Philadelphia, PA.
Brown, V.M. 1968. The calculation of the acute toxicity of mixtures of poisons to rainbow trout. Water

Res. 2:723-733.
Colt, J., and G. Tchobanoglous. 1976. Evaluation of the short-term toxicity of nitrogenous compounds to

channel catfish, Ictalurus punctatus. Aquaculture 6:209-224.
Hamilton, M.A., R.C. Russo, and R.V. Thurston. 1977. Trimmed Spearman-Karber method for esti-
mating median lethal concentrations in toxicity bioassays. Environ. Sci. Technol. 11:714-719.
Lemke, A.E., W.A. Brungs, and B J. Halligan. 1978. Manual for construction and operation of toxicity-

testing proportional diluters. EPA-600/3-78-072. U.S. Environmental Protection Agency,

Environmental Research Laboratory, Duluth, MN.
Mount, D.I., and W.A. Brungs. 1967. A simplified dosing apparatus for fish toxicity studies. Water Res.

1:21-29.
Roseboom, D.P., and D.L. Richey. 1977. Acute toxicity of residual chlorine and ammonia to some native

Illinois fishes. Report of Investigation 85, Illinois State Water Survey, Urbana.
Sprague, J.B. 1973. The ABC's of pollutant bioassay using fish. Pages 6-30 in Biological methods for

the assessment of water quality. ASTM STP 528, American Society for Testing and Materials,

Baltimore, MD.
Thurston, R.V., G.R. Phillips, R.C. Russo, and S.M. Hinkins. 1981. Increased toxicity of ammonia to

rainbow trout (Salmo gairdneri) resulting from reduced concentrations of dissolved oxygen. Can.

J. Fish. Aquat. Sci. 38:983-988.
Thurston, R.V., R.C. Russo, and K. Emerson. 1979. Aqueous ammonia equilibrium — tabulation of

percent un-ionized ammonia. EPA-600/3-79-091. U.S. Environmental Protection Agency,

Environmental Research Laboratory, Duluth, MN.
Thurston, R.V., R.C. Russo, and G.A. Vinogradov. 1981. Ammonia toxicity to fishes. Effect of pH on

the toxicity of the un-ionized ammonia species. Environ. Sci. Technol. 15:837-840.
U.S. Environmental Protection Agency. 1975. Methods for acute toxicity tests with fish, macro-
invertebrates, and amphibians. EPA-660/3-75-009. National Environmental Research Center,

Corvallis, OR.
U.S. Environmental Protection Agency. 1979. Methods for chemical analysis of water. EPA-600/4-79-

020.

12

Appendix A-1. Ammonia toxicity curves for rainbow trout at3-5°C versus 13-15°C.

200-1

100 1

70 1
50-
40

E 30-

CD

Z 20-

CO

re

^ 10-

>-

CO

o
a-
x

0

TROUT I
5.0°C . o 12.8°C

• o

• o

0.3

0.5

4.1

0.2 0.4 0.6 0.9
CONCENTRATION OF NH3-N (mg/1)

200-1

g 90-

« 70=

£ 50-

"- 40-

CD

- 30-

s 20H

10-

CO
X

1+

0.1

TROUT II

l^Zj

0.

0.3
2 0

<j 14.2°C

o*

0.5

4 0.6

■z r.i

CONCENTRATION OF NHt-N (nig/1)

100
80

60

50
£=49

^30-

o

^20-

CO
O
Q_
X

10-

0

TROUT III

3.3°C/ o 14.9°C

o
o

0.3

015

0.7 I
.6 1.0

0 2 0 4 0
CONCENTRATION OF NH3-N (mg/1)

13

Appendix A-2. Ammonia toxicity curves for bluegill at 4-4.5°C versus 25°C.

^200-1

| 100;

b^ 80

£ 60

£ 50

o 40

*Z 30

20-

^ 10 H

co

UJ 1

BLUEGILL I

4°C

0.1 0.2

0.4

0.6

25°C

'8

^1

1.0

0.3 0.50.7

o
o

2.0

CONCENTRATION OF Nfk-N (mg/1)

100 q

70 1
50-
40-

' 30-
20-

5 10-

I—

o

s:

>-
t

\

BLUEGILL II
4.5°C <^ 24.8°C

.1 0.'2 0!

• o

0.6

1.0

0.3 0 5 0.70.9 2.0
CONCENTRATION OF NH3-N (mg/1) •

14

Appendix A-3. Ammonia toxicity curves for fathead minnow at 4-5°C versus 24-25°C.

200-1

CD

inn

>

S<

8U

>-
1 —

60

u_

sn

u_

w

CD

1 —

3U

oo

3=

20

10-

GO
CD
Q_
X

FATHEAD MINNOW I

•

<^23.9°C
o

•

o

•

o

• o

.1 0,2 1 0,4 o!a r.o 210
0.3 0.5 0.7

CONCENTRATION OF NH3-N (mg/1)

FATHEAD MINNOW II

__l
<x.
t—

CD

lOOn

70:

4.6°C,

•

0
o25.2°C

6-S

u_
u_

CD

1—

50-

40-

30-

• 0
• 0

~

20-

UJ
ST

10-

UJ

en

■=3

1-

0.

CD

a.

X

UJ

1

1
0.3

,o!5

1

0.7

0.2 0.4 0.60.8 2.0

CONCENTRATION OF NH?-N (mg/1)

15