Full text of DOI:10.1002/etc.5620210217

Environmental Toxicology and Chemistry, Vol. 21, No. 2, pp. 347–352, 2002
2002 SETAC
Printed in the USA
᭧
0730-7268/02 $9.00
ϩ .00
THE EFFECT OF CALCIUM AND MAGNESIUM RATIOS ON THE TOXICITY OF
COPPER TO FIVE AQUATIC SPECIES IN FRESHWATER
R
AMI B. NADDY,* WILLIAM A. STUBBLEFIELD, JEFFREY R. MAY, SCOTT A. TUCKER, and J. RUSSELL
ENSR, 4303 West LaPorte Avenue, Fort Collins, Colorado 80521, USA
HOCKETT
(
Received 28 March 2001; Accepted 26 July 2001)
Abstract—While it is generally accepted that water hardness affects copper toxicity, the major ions that contribute to water hardness
(calcium [Ca] and magnesium [Mg]) may affect copper toxicity differently. This is important because the Ca:Mg ratio in standard
laboratory-reconstituted waters often differs from the ratio in natural surface waters. Copper toxicity was assessed for five different
aquatic species: rainbow trout (RBT), fathead minnow (FHM), Ceriodaphnia dubia, Daphnia magna, and an amphipod (Gammarus
sp.) under different Ca:Mg ratios (4:0, 3:1, 1:1, 1:3, and 1:4 mass basis) at a common hardness (180 mg/L as CaCO3) and alkalinity
(120 mg/L as CaCO3). Copper toxicity increased at lower Ca:Mg ratios for RBT but increased at higher Ca:Mg ratios for D. magna.
Fathead minnows (Ͻ24 h old) were more sensitive to copper in 1:1 Ca:Mg waters compared to 3:1 Ca:Mg waters. The toxicity of
copper did not vary under different Ca:Mg ratios for Gammarus sp., C. dubia, and 28-d-old FHM. The effect of Ca:Mg ratios on
copper toxicity changed for D. magna in softer water (90 mg/L as CaCO3) compared with hard water studies.
Keywords—Acute exposure
Copper
Calcium
Magnesium
Water hardness
INTRODUCTION
Water hardness, alkalinity, pH, and natural organic matter
cause of their sensitivity to copper in the copper criteria da-
tabase [1,8] and their general importance in aquatic toxicity
testing [7]. An additional set of Ca:Mg studies (4:0, 3:1, and
1:1) was also conducted with D. magna in laboratory-recon-
stituted moderately hard water (90 mg/L as CaCO3).
are known to affect the toxicity of copper to aquatic organisms
[1–3]. It has also been demonstrated that other factors, spe-
cifically the relative concentrations of calcium and magnesium
[2,4] as well as other ions, such as sodium [2], can influence
copper toxicity. Total water hardness is used as a surrogate to
estimate ameliorating affects for metal toxicity [1]. However,
the relative ratios of Ca and Mg (the major contributors to
water hardness) can vary greatly from one water body to an-
other (R.J. Erickson, unpublished data) and may not affect
metal toxicity equally. For example, Ca was more protective
of copper or cadmium toxicity than Mg to salmonids [4,5].
Therefore, it is important to understand not only how metal
toxicity is influenced by water hardness but, more important,
how the relative concentrations of Ca and Mg can affect the
toxicity of metals in freshwaters. Understanding this relation-
ship is especially critical because standard reconstituted lab-
oratory dilution waters used for toxicity tests do not have Ca:
Mg ratios similar to natural waters. Reconstituted laboratory
dilution waters are comprised of equal quantities of calcium
sulfate (as CaSO4·2H2O) and magnesium sulfate (as anhydrous
MgSO4) [6,7], resulting in a Ca:Mg ratio of approximately 1.1:
1 (0.7:1 on a molar basis), whereas the Ca:Mg ratio in natural
waters ranges from 1.6:1 to 8:1 (1:1–5:1 on a molar basis)
with an average of 3.3:1 (2:1 on molar basis) (R.J. Erickson,
unpublished data).
MATERIALS AND METHODS
Culture methods
The laboratory water used in testing was reconstituted from
American Society of Testing and Materials (Philadelphia, PA,
USA) Type I laboratory water (Milli-Q௡; Millipore, Bedford,
MA, USA). Source water to the Milli-Q system was obtained
from Horsetooth Reservoir surface water (treated by filtration
[sand, 0.2 and 0.05 ␮m] and ultraviolet radiation) or tap water
(Fort Collins, CO, USA). All organisms cultured or held at
the laboratory were maintained either in processed Horsetooth
Reservoir water (HT) or in waters reconstituted from Milli-Q
water. Organisms were not acclimated to the reconstituted wa-
ters in which they were going to be tested, except for D. magna
used in the hard water (180 mg/L) experiments. Daphnia mag-
na used in these experiments were cultured in reconstituted
waters of a similar hardness (1:1 Ca:Mg ratio) and alkalinity
prior to testing.
Rainbow trout used during these studies were obtained from
Lost River Trout Hatchery (Mackey, ID, USA) or Clines Trout
Hatchery (Boulder, CO, USA). Fish were delivered as swim-
up fry and held in flowing HT culture water at approximately
Studies were conducted to determine the effect of different
Ca and Mg ratios (4:0, 3:1, 1:1, 1:3, and 1:4 on a mass basis)
in laboratory reconstituted hard water (180 mg/L as CaCO3)
on the acute toxicity of copper to select aquatic laboratory test
organisms. Studies were conducted with rainbow trout (On-
corhynchus mykiss), fathead minnows (Pimephales promelas,
12
Њ
C (Table 1) until testing (minimum 14 d after receipt). Trout
Salmon Starter; Gard-
were fed trout chow (Zeigler Brothers
௡
ners, PA, USA) ad libitum twice daily during holding. All
trout were juvenile stage (postlarval, swim-up stage), had ab-
sorbed their yolk sacs, and were actively feeding (Ͼ20 d) at
the surface before they were selected for testing. The average
weights of rainbow trout (RBT) used in the studies ranged
from 0.23 to 0.49 g.
Ͻ
24-h- and 28-d-old fish), cladocerans (Ceriodaphnia dubia
and Daphnia magna), and amphipods (Gammarus sp.) be-
* To whom correspondence may be addressed (rnaddy@ensr.com).
Fathead minnow (FHM) fry were obtained from in-house
347

348
Environ. Toxicol. Chem. 21, 2002
R.B. Naddy et al.
Table 1. Water quality characteristics (mean
Ϯ
standard deviation) for laboratory culture water and reconstituted laboratory waters used in
Ca:Mg studies
Total recoverable concn. (mg/L)
Hardness
Alkalinity
Treatmenta
(mg/L as CaCO3) (mg/L as CaCO3) Initial pH
Ca
Mg
1.5
K
Na
2.6
Culture water
Horsetoothb
180 hardness
4:0 (4:0)
3:1 (1.8:1)
1:1 (0.7:1)
1:3 (0.2:1)
1:4 (0.16:1)
0:4 (0:4)
28
Ϯ
4.4
25
Ϯ
1.7
7.4
Ϯ
0.2
8.9
0.8
177
178
180
179
176
180
Ϯ
Ϯ
Ϯ
Ϯ
2.6
2.3
1.8
4.2
115
116
115
114
114
116
Ϯ
Ϯ
Ϯ
Ϯ
3.1
2.3
2.9
6.4
8.2
8.2
8.2
8.3
Ϯ
Ϯ
Ϯ
Ϯ
8.3
8.2
0.2
0.1
0.3
0.0
68.0
44.8
28.5
11.8
Ϯ
Ϯ
Ϯ
Ϯ
4.3
1.1
0.52
0.14
0.02
15.5
25.9
36.0
Ϯ
Ϯ
Ϯ
Ϯ
0.01
0.50
1.6
4.27
Ϯ
Ϯ
Ϯ
Ϯ
0.20
0.21
0.13
0.03
61.9
Ϯ
Ϯ
Ϯ
Ϯ
14
4.35
4.37
4.23
55.6
56.3
53.0
2.3
3.5
1.1
0.21
10.0
—
37.0
—
4.17
—
51.7
—
90 hardness
4:0 (4:0)
3:1 (1.8:1)
1:1 (0.7:1)
90
90
92
56
59
59
8.0
8.0
7.5
38.4
23.8
17.0
0.05
8.52
12.7
2.29
2.15
2.22
29.2
29.6
28.7
a Ca:Mg ratios are presented on a mass basis and on a molar basis parenthetically.
b Total recoverable and dissolved copper concentrations in Horsetooth culture water were 9.0 and 6.0
␮
g/L, respectively.
cultures. Brood stock were maintained in 37.8-L (10-gallon)
aquaria with each aquaria having a male:female ratio of ap-
proximately 2:5. Adult organisms were fed frozen adult brine
shrimp (San Francisco Bay Brand, Newark, CA, USA) ad li-
bitum three times daily. Spawning tiles with eggs were re-
ad libitum daily during holding. The average weight of the
organisms used for testing was 24.5 mg (range 8.71–44.5 mg).
Reference toxicant tests (positive control test with sodium
chloride) were conducted with each batch of testing organisms
at the Fort Collins Environmental Toxicology Laboratory (Fort
Collins, CO, USA). Organism responses were within historical
moved daily and placed in Fungus Guard௢ (Jungle Labora-
tories, Cibolo, TX, USA) for 24 to 48 h, and then spawning
substrates with eyed eggs were placed in aerated hatching trays
filled with reconstituted moderately hard water (hardness and
alkalinity of 90 and 60 mg/L as CaCO3, respectively). Less-
than-24-h-old minnows were collected and used in three
control limits, which represent Ϯ2 standard deviations from
the historical laboratory mean median lethal concentration
(LC50).
Experimental design
rounds of studies. For 28-d-old fish, approximately 500
Ͻ
24-
Test methods were in accordance with appropriate U.S. EPA
and American Society for Testing and Materials guidance
[6,7,9]. The laboratory water used in Ca:Mg testing studies
was prepared in a similar fashion as described in U.S. EPA
[6,7], except the amounts of calcium (as CaSO4·2H2O), mag-
nesium (as anhydrous MgSO4), and sodium bicarbonate
(NaHCO3) were altered to achieve a target hardness (180 mg/
L as CaCO3), alkalinity (120 mg/L as CaCO3), and Ca:Mg
ratio (4:0, 3:1, 1:1, 1:3, and 1:4) (Table 1). All Ca:Mg ratios
are presented on a mass basis unless otherwise indicated.
All testing was conducted using static (48-h tests) or static
renewal (96-h tests) procedures with a 16:8-h light:dark pho-
h-old FHM were maintained in HT water (similar to adults)
and fed trout chow and brine shrimp nauplii ad libitum three
times daily. Fish (28 d old) used for testing averaged 11.2 mm
long (range 9–13 mm) and weighed 13.4 mg (range 4.86–26.1
mg).
Ceriodaphnia dubia were obtained from in-house cultures
reared following U.S. Environmental Protection Agency (U.S.
EPA) procedures [7]. Monocultures of gravid C. dubia were
isolated daily to ensure that enough neonates could be pro-
duced that would be
Ͻ24 h old and were within 8 h of age.
Organisms were fed 0.2 ml of a yeast:trout chow:cereal leaves
(YTC)/algal suspension [7] daily in moderately hard recon-
toperiod. For 96-h tests, solutions (ϳ90–100%) were renewed
stituted water. Organisms were maintained at 25
culturing.
Daphnia magna (Ͻ24 h old) were also obtained from in-
house cultures for all studies. Approximately 30 adult organ-
isms were cultured in 2 L of reconstituted hard water (hardness
Ϯ
1
Њ
C during
after 48 h of exposure. During test renewal, organisms were
transferred to freshly prepared test solutions. Organisms in all
tests were exposed to six experimental copper concentrations
(using a 50 or 60% dilution series) in each Ca:Mg laboratory
water. Negative control treatments (unspiked dilution water)
were conducted concurrently with each test. The biological
endpoint measured for all acute toxicity tests was death (de-
fined as no visible movement or any response to gentle prod-
ding with a blunt probe).
and alkalinity
and maintained at a temperature of 20
ϳ
180 and 120 mg/L as CaCO3, respectively)
C. Organisms were
Ϯ 1Њ
fed a 3:5 YTC:algal mixture daily at a rate of 8 ml/1 L of
culture water. The YTC was prepared as described in U.S.
EPA [7] except that the trout chow was not digested prior to
use. Adult organisms were transferred to new culture media
Rainbow trout studies were conducted as 96-h static re-
newal tests at 12
rate 0.58 g/L) with two replicates of 10 fish each (Table 2).
Fish were held in HT water at 12 C and were not fed for
Ϯ 1ЊC, in 5 to 9 L of test solution (loading
(
ϳ
3:1 new:old media) three times weekly.
Յ
Amphipods (Gammarus sp.) were obtained from Environ-
Ϯ 1Њ
mental Consulting and Testing (Superior, WI, USA). Organ-
isms (adults of mixed age) were maintained in a mixture of
reconstituted hard water and water in which the organisms
were shipped (hardness and alkalinity of 62 and 61 mg/L as
CaCO3, respectively). Organisms were fed YTC (C. dubia diet)
48 h prior to testing. In the initial set of studies, a 0:4 Ca:Mg
treatment was included; however, control organisms in this
treatment had significant mortality (90% at 48 h). Because of
this response, the 0:4 Ca:Mg treatment was replaced with a
1:4 Ca:Mg treatment.
and Tetramin
௡
(Tetra Sales, Blacksburg, VA, USA) flake food
Fathead minnow studies were conducted as 48-h static (two

Effect of Ca:Mg ratios on copper toxicity
Environ. Toxicol. Chem. 21, 2002
349
Table 2. Experimental design and median lethal copper concentrations (LC50s) measured during Ca:Mg studies
Water hardness/
alkalinity
Test
duration (h)
LC50 (95% CI)
total Cu ( g/L)
% Dissolved
Cu
Species
Organism age
42 d posthatch
Ca:Mg ratioa
␮
Rainbow trout
4:0
3:1
1:1
1:3
0:4
4:0
3:1
1:1
1:3
1:4
4:0
3:1
1:1
4:0
3:1
1:1
3:1
1:1
4:0c
3:1
1:1
4:0
3:1
1:1
4:0
3:1
1:1
4:0
3:1
1:1
4:0
3:1
1:1
4:0
3:1
1:1
4:0
3:1
1:1
4:0
3:1
1:1
4:0
3:1
1:1
172/114
178/116
176/114
176/119
180/116
176/108
180/113
178/114
180/110
176/114
179/116
177/117
180/115
178/115
178/117
180/111
180/117
180/117
176/118
176/116
182/116
176/118
176/116
182/116
178/115
178/117
180/111
178/117
180/117
180/117
180/117
182/117
180/118
176/114
176/111
180/111
180/117
182/117
180/118
90/56
96
67.9 (55.5–82.9)
53.9 (47.1–61.8)
35.5 (31.3–40.4)
18.1 (15.8–20.7)
87
84
85
96
—
89
90
97
93
95
86
90
90
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
b
—
47 d posthatch
45 d posthatch
52.5 (45.2–61.0)
46.2 (38.7–55.1)
30.7 (25.7–36.6)
17.9 (15.3–21.0)
18.1 (15.6–21.0)
37.3 (32.0–43.5)
27.7 (23.3–32.9)
21.2 (18.6–24.3)
—
Fathead minnow
Ͻ
24 h old
48
96
636
(572–706)
(541–1031)
(508–823)
(296–541)
—
747
647
400
837
503
442
502
434
(686–1021)d
(418–607)d
(377–518)
(421–599)
(365–516)
28 d old
Ceriodaphnia dubia
Daphnia magna
Gammarus sp.
Ͻ
24 h old
48
48
96
10.50 (9.62–11.47)
17.47 (13.50–22.60)
16.16 (13.85–18.87)
7.92 (7.62–8.23)
12.94 (11.60–14.43)
14.49 (12.82–16.38)
7.88 (6.61–9.39)
9.14 (5.78–14.45)
7.89 (6.15–10.14)
16.35 (14.27–18.74)
20.97 (17.66–24.90)
57.28 (48.66–67.44)
21.55 (16.94–27.41)
31.77 (26.90–37.51)
42.68 (35.88–50.77)
12.20 (10.27–13.88)
16.93 (13.97–20.52)
11.99 (10.06–14.29)
Ͻ
24 h old
90/59
92/59
176/118
176/118
182/116
Mixed
181
103
133
(94.24–348)
(84.48–125)
(94.7–187)
a Ca:Mg ratio presented on mass basis.
b Control survival was 10% at 48 h. See text for details.
c Consisted of only a control treatment.
d Data are for 48 h; 96-h values are provided in the text.
rounds for
for 24-h- and 28-d-old fish) tests at 25
ml of test solution (loading rate 0.45 g/L) with either two
Ͻ
24-h-old fish) and 96-h static renewal (one round
ditional acclimation. Organisms were not fed during testing.
Daphnia magna studies were carried out in a similar fashion,
except studies were conducted at 20 Ϯ 1ЊC. An additional
Ͻ
Ϯ
1ЊC in 200 or 300
Յ
replicates of 10 fish each (96-h tests) or four replicates of five
fish each (48-h tests). Organisms less than 24 h old were placed
directly into test solutions, whereas 28-d-old organisms were held
in HT culture water at test temperature for 48 h off-feed prior to
test initiation. Fathead minnows were not fed during 48-h acute
studies. However, in the 96-h acute toxicity studies, organisms
series of studies was conducted at a target hardness and al-
kalinity of 90 and 60 mg/L, respectively.
Gammarus sp. studies were 96-h static renewal tests con-
ducted at 17
licates of 10 organisms each. Test solutions were renewed on
day 2. Amphipods were held at 13 to 16 C without feeding
Ϯ 1ЊC in 300 ml of test solution with two rep-
Њ
were fed 0.1 ml (
concentrated newly hatched brine shrimp nauplii suspension per
test chamber 2 h prior to test renewals on day 2.
Ͻ
24-h-old fish) or 0.2 ml (28-d-old fish) of
for 48 h prior to test initiation, and organisms were not fed
during testing.
Test solutions
Ceriodaphnia dubia studies were conducted as 48-h static
tests at 25
Ϯ
1
Њ
C in 50 or 80 ml of test solution with four
Copper was introduced into each test as a reagent-grade
chloride salt (CuCl2·2H2O; Mallinckrodt, Paris, KY, USA).
During RBT testing, up to four samples were collected from
replicates of five organisms each (Table 2). Ceriodaphnia du-
bia were placed directly into test solutions without any ad-

350
Environ. Toxicol. Chem. 21, 2002
R.B. Naddy et al.
each test treatment for analytical confirmation of nominal cop-
per concentrations. Samples were collected and composited
from treatment replicates. For 48-h acute studies, only two
samples were collected from each test treatment, at test ini-
tiation and at test termination or when 100% mortality occurred
in an individual treatment. For 96-h static renewal acute tox-
icity studies, samples were collected at test initiation, prior to
test renewal (old solutions), just after test renewal (FHM and
amphipod studies), and at test termination or when 100% mor-
tality occurred in an individual treatment (trout studies).
Test solutions were analyzed for total recoverable, dis-
solved, or acidified (no filtration or digestion, only acidified)
copper [10]. All samples were acidified to 1% using concen-
trated, trace metals–grade nitric acid (Mallinckrodt). We de-
termined that for our laboratory waters, total recoverable and
acidified copper measurements were very similar (relative
changes in copper toxicity occurred within the mid-Ca:Mg
ratios, and the effect appeared to level off at higher and lower
Ca:Mg ratios (Table 2).
Although RBT were more sensitive to copper in waters
having higher Mg concentrations, we also observed that RBT
could survive (for 96 h) in laboratory waters in which the Ca:
Mg ratio was 4:0 but not when it was 0:4 Ca:Mg. Organisms
tested in the 0:4 Ca:Mg treatment had significant mortality in
the control treatment (90% at 48 h).
Fathead minnow studies
Less-than-24-h-old FHM were most sensitive to copper in
1:1 Ca:Mg waters compared with 3:1 Ca:Mg waters (Table 2).
Studies were initially conducted with 4:0 Ca:Mg waters; how-
ever, control survival of
Ͻ24-h-old FHM was significantly
reduced (45 and 40% survival in the first and the third study,
respectively) in these waters. In the third set of tests, 96-h
standard deviation 2.6
Ϯ 1.8%, n ϭ 12). Dissolved (filtration
through a 0.45- m polyethyl sulfone filter; Whatman, Clifton,
␮
LC50 (95% CI) for
Ca:Mg, 525.9 (461.3–599.6)
(271.1–415.6) g/L. As with the 48-h LC50s, FHM were more
sensitive to copper in 1:1 Ca:Mg waters. No difference was
observed in copper sensitivity for 28-d-old FHM under dif-
Ͻ
24-h-old FHM were as follows: 3:1
NJ, USA) copper concentrations were measured only for RBT
studies. Copper concentrations were determined using induc-
tively coupled plasma atomic emission spectroscopy, graphite
furnace atomic absorption, or direct aspiration flame atomic
absorption spectroscopy.
␮
g/L, and 1:1 Ca:Mg, 335.7
␮
ferent Ca:Mg ratios (Table 2). In addition,
Ͻ24-h-old FHM
were not any more sensitive to copper than 28-d-old FHM
(Table 2).
Data analysis
Results of each of the individual acute tests were used to
calculate time-dependent LC50 values. The LC50 values and
associated 95% confidence limits were calculated using Spear-
man–Karber, trimmed Spearman–Karber, and probit analysis
methods [11]. Test concentrations were log10 transformed prior
to calculation of the LC50. The method selected for reporting
the test results was determined by the characteristics of these
data (the presence or absence of 0 and 100% mortality and
the number of concentrations in which partial mortality had
occurred) [7]. The binomial LC50 estimation method was also
used when appropriate [12]. Trends among LC50 data were
Cladoceran studies
Ceriodaphnia dubia sensitivity to copper was not consis-
tently affected by Ca:Mg ratios, although some trends were
observed among LC50s (Table 2). In the first two sets of stud-
ies, C. dubia were more sensitive to copper in the 4:0 Ca:Mg
treatments than the 3:1 or 1:1 Ca:Mg waters. However, the
last study showed no observable trend in copper sensitivity
among Ca:Mg treatments (Table 2).
Daphnia magna sensitivity to copper was the reverse of
that observed for RBT. Daphnia magna tested in hard water
(hardness 180 mg/L) were more sensitive to copper in 4:0
Ca:Mg treatments and least sensitive to copper in 1:1 Ca:Mg
waters (Table 2). No difference in copper sensitivity was ob-
served among Ca:Mg treatments in moderately hard waters
(hardness 90 mg/L) (Table 2).
compared using least-squares regression analyses (Microsoft
௡
Excel, Version 97 SR-2).
Differences among LC50s were compared using overlap of
95% CI in cases of only two treatments. In cases of three or
more treatments, this was also used as a conservative method
for determining differences among LC50s because of the in-
creased chance for type I error. Because of this, most com-
parisons were done on a relative basis to determine trends in
copper toxicity due to changes in Ca:Mg treatments.
Gammarus sp. studies
Gammarids were equally sensitive to copper in all three
Ca:Mg treatments (Table 2), although the LC50s were slightly
higher in the 4:0 and 1:1 Ca:Mg waters compared to the 3:1
Ca:Mg treatment.
RESULTS
Initial water quality characteristics for water treatments for
test waters were similar among treatments at a given hardness
except for the experimental variables (Table 1). Sodium con-
centrations were slightly higher in the 4:0 waters in 180-mg/
L hardness studies as a result of one round of testing (FHM
[96-h studies] and Gammarus sp.) because the alkalinity had
to be slightly augmented by the addition of sodium bicarbon-
ate.
DISCUSSION
We examined the influence of Ca:Mg ratios on copper tox-
icity by maintaining hardness, alkalinity, pH, and other se-
ϩ
ϩ
Ϫ
lected ions (K , Na , and Cl ) approximately constant. Our
results indicate that the Ca:Mg ratio does not affect the toxicity
of copper to these five laboratory test organisms in the same
way. The sensitivity of RBT,
Ͻ24-h-old FHM, and D. magna
to copper changed under different Ca:Mg ratios, while the
sensitivity of C. dubia, 28-d-old FHM, and Gammarus sp. was
Rainbow trout studies
Median lethal concentrations for RBT decreased with lower
Ca:Mg ratios for each series of studies conducted (Table 2).
Calcium had a greater effect on acute copper toxicity than
magnesium; copper was more toxic in waters with higher mag-
nesium concentrations (but a similar hardness and alkalinity)
than in waters with high calcium concentrations. The greatest
relatively unaffected. Both RBT and Ͻ24-h-old FHM were
less sensitive to copper in waters having higher calcium con-
centrations, while D. magna exhibited greater sensitivity at
higher Ca:Mg ratios. Welsh et al. [4] and Erickson et al. [2]
also observed decreased copper toxicity at higher Ca:Mg ratios
to salmonids and FHM, respectively. The Ca:Mg studies at a

Effect of Ca:Mg ratios on copper toxicity
Environ. Toxicol. Chem. 21, 2002
351
regulator of copper toxicity to salmonids in softer waters (hard-
ness 40 and 90 mg/L). The sensitivity of copper to 24-h-old
FHM was also influenced by higher Ca:Mg ratios in hard wa-
ters ( 146 mg/L), but this effect was not evident in softer
Ͻ
Ͼ
waters [2]. From other experiments conducted in our labora-
tory with RBT (unpublished data), other factors (e.g., alkalinity
and dissolved organic carbon) may become more important in
mitigating copper toxicity in softer waters. Magnesium appears
to be a better inhibitor of copper for D. magna (in hard water)
(Fig. 1b) (p ϭ 0.014), while both Ca and Mg were equally
competitive inhibitors of copper for C. dubia, Gammarus sp.,
and 28-d-old FHM. The difference in response to copper under
different Ca:Mg ratios in hard waters may be due to decreased
gill permeability at higher Ca concentrations [14] but not by
greater Mg concentrations [15]. While this explains why RBT
and
Ͻ24-h-old FHM are more sensitive to copper in 1:1 Ca:Mg
waters compared to 4:0 Ca:Mg waters, it does not explain why
28-d-old FHM were not influenced by different Ca:Mg ratios.
Nor does it explain why the invertebrates were either unaf-
fected or more sensitive at higher Ca:Mg ratios.
Although the absolute toxicity of copper can change for
certain species because of differences in Ca:Mg ratios, the real
question is, How does this impact the ambient water quality
criteria (AWQC) for copper? The acute AWQC or criterion
maximum concentration is calculated as half the final acute
value, which is derived from the four lowest genus mean acute
values [16]. For copper, the four most sensitive genera are the
Cladocerans (Ceriodaphnia and Daphnia), Northern squaw-
fish (Ptychocheilus; common name has been changed to
Northern pikeminnow), and amphipod (Gammarus) [8]. The
copper toxicity for C. dubia and Gammarus sp. was not sig-
nificantly affected by changes in Ca:Mg ratios. Furthermore,
D. magna were affected by differences in Ca:Mg ratio (at least
in hard waters), but they are more sensitive to copper at higher
Ca:Mg ratios. Since many of the Daphnia studies were con-
ducted in natural waters that tend to have a higher Ca:Mg ratio
[4], any difference in toxicity due to changes in Ca:Mg ratio
may be accounted for in the AWQC derivation. However, be-
cause organism sensitivity to copper under different Ca:Mg
ratios may change because of hardness, studies should also be
conducted over a range of water hardnesses to determine po-
tential effects. Based on our studies, it appears that for copper
the potential confounding effects of Ca:Mg ratios to the AWQC
are minimal.
Our studies suggest that the copper sensitivity for some,
but not all, organisms is influenced by differences in the Ca:
Mg ratio. However, until sufficient studies have been con-
ducted to fully evaluate this relationship for multiple species
in waters of varying hardnesses, we also recommend that
calcium and magnesium concentrations be matched with
known Ca:Mg ratios in waters of interest when preparing
laboratory waters for certain testing purposes (e.g., whole
effluent toxicity testing and water-effect ratios) [4]. These
relationships should also be evaluated for other metals, as
the Ca:Mg ratio has been shown to affect the toxicity of
cadmium [5,17] and nickel [18].
Fig. 1. Relationship between copper toxicity and calcium concentra-
tion (surrogate for Ca:Mg ratios because studies were conducted at
the same hardness) with (a) rainbow trout (RBT), fathead minnows
(Ͻ24-h- and 28-d-old FHM), and Gammarus sp. and (b) Cerioda-
phnia dubia and Daphnia magna. Results for RBT, FHM, and Gam-
marus sp. are for 96-h studies, while values for C. dubia and D.
magna are 48-h data. All studies were conducted at a hardness and
alkalinity of 180 and 120 mg/L as CaCO3, respectively.
lower hardness resulted in changes in organism sensitivity to
copper. Daphnia magna tested at 180-mg/L hardness exhibited
differences in copper sensitivity due to Ca:Mg ratios, while
those tested in waters at 90-mg/L hardness did not. Erickson
et al. [2] also observed differences in copper toxicity for
Ͻ24-
h-old FHM under different Ca:Mg ratios at hardnesses of 146
and 246 mg/L as CaCO3 but not at a hardness of 106 mg/L.
However, differences in our study may be due to other factors
(e.g., alkalinity, pH, and Na concentration) besides the Ca:Mg
ratio, as the ionic composition of the waters were different
between the two studies (Table 2).
ϩ
Based on the competitive exclusion of the cupric ion (Cu2 )
ϩ
at Na channels in the fish gill (the mode of action for Cu
ϩ
[13]), we believe Ca2 is a better inhibitor of copper toxicity
ϩ
than Mg2 is for RBT and
Ͻ24-h FHM in hard water. The
Acknowledgement—Primary funding for this project was sponsored
by the Atlantic Richfield Company. Thanks also to Stan Capps, Eric
VanGenderen, Kerry Byrn, and Gina Stern for laboratory assistance
and to Robert Gensemer, Joseph Meyer, and two anonymous reviewers
for constructive comments.
effect of different Ca:Mg ratios (expressed as calcium con-
centration) significantly affected copper toxicity for RBT in
hard waters (Fig. 1a) (
p ϭ 0.0007). Studies conducted by Welsh
et al. [4] suggest that the Ca:Mg ratio is also an important

352
Environ. Toxicol. Chem. 21, 2002
R.B. Naddy et al.
REFERENCES
on determination and use of water-effect ratios for metals. EPA-
823-B-94-001. Office of Water, Washington, DC.
10. U.S. Environmental Protection Agency. 1983. Methods for chem-
ical analysis of water and wastes. EPA/600/4-79/020. Environ-
mental Monitoring and Support Laboratory, Cincinnati, OH.
1. U.S. Environmental Protection Agency. 1985. Ambient water
quality criteria for copper—1984. EPA 440/5-84-031. Office of
Water Regulations and Standards, Washington, DC.
2. Erickson RJ, Benoit DA, Mattson VR, Nelson HP Jr, Leonard
EN. 1996. The effects of water chemistry on the toxicity of copper
to fathead minnows. Environ Toxicol Chem 15:181–193.
3. Schubauer-Berigan MK, Dierkes JR, Monson PD, Ankley GT.
1993. pH-dependent toxicity of Cd, Cu, Ni, Pb, and Zn to Cer-
iodaphnia dubia, Pimephales promelas, Hyalella azteca, and
Lumbriculus variegatus. Environ Toxicol Chem 12:1261–1266.
4. Welsh PG, Lipton J, Chapman GA, Podrabsky TL. 2000. Relative
importance of calcium and magnesium in hardness-based modi-
fication of copper toxicity. Environ Toxicol Chem 19:1624–1631.
5. Carroll JJ, Ellis SJ, Oliver WS. 1979. Influences of total hardness
constituents on the acute toxicity of cadmium to brook trout (Sal-
velinus fontinalis). Bull Environ Contam Toxicol 22:575–581.
6. American Society for Testing and Materials. 1998. Standard guide
for conducting acute toxicity tests with fishes, macroinvertebrates,
and amphibians. E729-96. In Annual Book of ASTM Standards,
Vol 11.04. Philadelphia, PA, pp 218–238.
11. Gulley DD. 1994. TOXSTAT
௡, Ver 3.4. Western Ecosystems
Technology, Cheyenne, WY, USA.
12. Peltier WH, Weber CI. 1985. Methods for the acute toxicity of
effluents to freshwater and marine organisms, 3rd ed. EPA-600/
4-85/013. U.S. Environmental Protection Agency, Environmental
Monitoring and Support Laboratory, Cincinnati, OH.
13. Laure´n DJ, McDonald DG. 1986. Influence of water hardness,
pH, and alkalinity on the mechanisms of copper toxicity in ju-
venile rainbow trout, Salmo gairdneri. Can J Fish Aquat Sci 43:
1488–1496.
14. Hunn JB. 1985. Role of calcium in gill function in freshwater
fishes. Comp Biochem Physiol A 82:543–547.
15. Wendelaar Bonga SE, Lowik CJM, Van Der Meij JCA. 1983.
ϩ
ϩ
Effects of external Mg2 and Ca2 on branchial osmotic water
permeability and prolactin secretion on the teleost fish Saroth-
erodan mossambicus. Gen Comp Endocrinol 52:222–231.
16. U.S. Environmental Protection Agency. 1985. Guidelines for de-
riving numerical national water quality criteria for the protection
of aquatic organisms and their uses. PB85-227049. National Tech-
nical Information Service, Springfield, VA.
7. U.S. Environmental Protection Agency. 1993. Methods for mea-
suring the acute toxicity of effluents and receiving waters to fresh-
water and marine organisms, 4th ed. EPA/600/4-90/027F. Office
of Research and Development, Washington, DC.
17. Jackson BP, Lasier PJ, Miller WP, Winger PW. 2000. Effects of
calcium, magnesium, and sodium on alleviating cadmium toxicity
to Hyalella azteca. Bull Environ Contam Toxicol 64:279–286.
18. Doig L, Liber K. 2001. All water hardnesses are not created equal.
SETAC Globe 2:34–35.
8. U.S. Environmental Protection Agency. 1996. 1995 Updates: Wa-
ter quality criteria documents for the protection of aquatic life in
ambient water. EPA-820-B-96-001. Office of Water, Washington,
DC.
9. U.S. Environmental Protection Agency. 1994. Interim guidance