1. 804
Environmental Toxicology and Chemistry, Vol. 20, No. 4, pp. 804–809, 2001
᭧ 2001 SETAC
Printed in the USA
0730-7268/01 $9.00 ϩ .00
SEA URCHIN FERTILIZATION ASSAY: AN EVALUATION OF ASSUMPTIONS
RELATED TO SAMPLE SALINITY ADJUSTMENT AND USE OF NATURAL AND
SYNTHETIC MARINE WATERS FOR TESTING
EMILIA JONCZYK,† GUY GILRON,*‡ and BARRY ZAJDLIK§
†BEAK International, 14 Abacus Road, Brampton, Ontario L6T 5B7, Canada
‡ESG International, 361 Southgate Drive, Guelph, Ontario N1G 3M5, Canada
§B. Zajdlik & Associates, R.R. 3, Rockwood, Ontario N0B 2K0, Canada
(Received 15 February 2000; Accepted 2 August 2000)
Abstract—Most industrial effluents discharged into the marine coastal environment are freshwater in nature and therefore require
manipulation prior to testing with marine organisms. The sea urchin fertilization test is a common marine bioassay used for routine
environmental monitoring, investigative evaluations, and/or regulatory testing of effluents and sediment pore waters. The existing
Canadian and U.S. Environmental Protection Agencies test procedures using sea urchin (and sand dollar) gametes allow for sample
salinity adjustment using either brine or dry salts. Moreover, these procedures also allow for the use of either natural or synthetic
marine water for culturing/holding test organisms and for full-scale testing. At present, it is unclear to what extent these variables
affect test results for whole effluents. The test methods simply state that there are no data available and that the use of artificial
dry sea salts should be considered provisional. We conducted a series of concurrent experiments aimed at comparing the two
different treatments of sample salinity adjustment and the use of natural versus synthetic seawater in order to test these assumptions
and evaluate effects on the estimated end points generated by the sea urchin fertilization sublethal toxicity test. Results from these
experiments indicated that there is no significant difference in test end points when dry salts or brine are used for sample salinity
adjustment. Similarly, results obtained from parallel (split-sample) industrial effluent tests with natural and artificial seawater suggest
that both dilution waters produce similar test results. However, data obtained from concurrent tests with the reference toxicant,
copper sulfate, showed higher variability and greater sensitivity when using natural seawater as control/dilution water.
Keywords—Salinity Toxicity Sea urchin fertilization assay Artificial seawater Natural seawater
INTRODUCTION
The sea urchin fertilization (or sperm cell) toxicity test has
been used to evaluate whole effluent, sediment pore water, and
receiving water quality for over 15 years in numerous marine
environments worldwide [1–8]. More recently, the test has
been included in test batteries evaluating effects on marine
invertebrates [9,10] and toxicity identification evaluation [11–
13] and is being considered in a test battery for assessing the
toxicity of dredged materials prior to ocean disposal by En-
vironment Canada [14].
Industrial effluents discharged into the marine environment,
such as pulp and paper mill and metal mining effluents, are
usually freshwater in nature. Sublethal toxicity testing of ef-
fluents with marine organisms (e.g., sea urchins, mysid shrimp,
and inland silversides) therefore requires adjustment to the
salinity of the receiving environment in order to be used in
testing [15]. When salinity adjustment is required, options in
existing marine toxicity test methods are available. These are
addition of hypersaline brine or addition of commercially
available dry salts. Also, either natural or synthetic seawater
can be used as control/dilution water in the sea urchin fertil-
ization test.
Existing standard testing procedures for the sea urchin fer-
tilization test [16,(http://www.pyr.ec.gc.ca/ep/enforcement/),17]
allow for the use of either brine (synthetic or derived from
natural seawater) or dry salts (synthetic formulation of ana-
lytical-grade chemicals or prepared from the evaporation of
* To whom correspondence may be addressed
(ggilron@yahoo.com).
natural marine water) for sample salinity adjustment. Also, the
testing procedures currently allow the use of either natural or
synthetic marine water for culturing and holding test organisms
and for full-scale toxicity testing. The U.S. Environmental
Protection Agency test procedures simply state, ‘‘No data from
sea urchin or sand dollar fertilization tests using sea salts or
artificial seawater are available for evaluation at this time, and
their use should be considered provisional’’ [18]. Since test
methods and techniques conducted for regulatory purposes
must be standardized for all laboratories regardless of their
geographical location for valid comparison of results, it is
important to discover whether the two methods for salinity
adjustment and dilution water use (synthetic or natural sea-
water) will produce comparable test results.
One assumption is that hypersaline brine prepared from dry
salts or natural seawater and the natural water collected from
a clean (uncontaminated) area are adequate for culturing sea
urchins and testing their gametes. A number of advantages and
disadvantages exist for using brine for salinity adjustment or
natural seawater as dilution water. First, brine can be stored
for prolonged periods without compromising seawater integ-
rity. Moreover, brine and natural seawater can contain natural
microelements (e.g., bromine and iodine) required to sustain
healthy organism growth and development. Also, preparation
of brine is cost-effective where seawater is readily available
(laboratories located in coastal regions). Finally, seawater from
the coastal environments in question provides the added ad-
vantage of ecological relevance (the natural seawater used in
the test is from the receiving environment into which the ef-
fluent is discharged).

2. Sea urchin fertilization test evaluation Environ. Toxicol. Chem. 20, 2001 805
A number of disadvantages exist as well for using brine.
For example, the highest exposure concentration can be a max-
imum of only 70%. Moreover, if brine from seawater is used
to adjust salinity, as well as natural seawater used as dilution
water, shipping costs are higher for inland testing laboratories.
In addition, the quality of natural seawater and the brine is
naturally variable over time, and the potential exists for the
occurrence of natural pathogens and predators in the seawater.
These latter two factors could significantly confound test data.
Finally, the potential exists for chemical or biological contam-
ination during the brine preparation process.
An alternative to using brine that obviates some of these
problems is using dry salts dissolved in an effluent sample
used to adjust salinity or dissolved in deionized water. This
assumes that synthetic seawater is adequate for culturing sea
urchins and toxicity testing using their gametes. Using dry
salts for culturing or salinity adjustment is an advantage be-
cause it is possible to test full-strength effluents with dry salts,
dry salts are commercially available with a broad selection of
brand names (e.g., Forty Fathoms௡, Marine Enterprises, Bal-
timore, MD, USA, and Instant Ocean௡, Aquatium Systems,
Mentor, OH, USA), and testing is possible regardless of geo-
graphical location (inland laboratories can conduct testing as
easily as coastal laboratories). Moreover, synthetic water is
standardized, and therefore the possibility of environmental
contamination is diminished. Using synthetic seawater is a
disadvantage because conditioning and monitoring of the water
is required when using dry salts, synergistic effects of dry salts
with potential toxicants in effluents are not well understood
and often are unknown, and the effects may not be reflective
of the receiving environment. The current assumption is that
synthetic marine water will not confound test data. However,
this assumption has, to our knowledge, been tested only with
reference toxicants [19].
This study used an experimental approach that was under-
taken in two phases. Phase 1 investigated the sample salinity
adjustment techniques on the sea urchin fertilization test and
was carried out between 1994 and 1996. This phase tested the
hypothesis that there is no difference in results obtained from
sea urchin fertilization tests for samples salinity-adjusted using
hypersaline brine or dry salts. Phase 2 evaluated the effect of
natural versus synthetic control/dilution water on the sea ur-
chin test and was performed between 1997 and 1998. This
phase tested the hypothesis that there is no difference in sea
urchin fertilization test results conducted with either natural
or synthetic seawater as dilution water.
MATERIALS AND METHODS
Test species
The test species used in this study was the white sea urchin,
Lytechinus pictus. Organisms used in testing were obtained
from a supplier of laboratory organisms (Marinus, Long
Beach, CA, USA). The sea urchins were held in marine aquaria
and fed a diet of romaine lettuce, with an occasional supple-
ment of fish tissue.
Test methods
The biological test method used in the study was the En-
vironment Canada method [16]. This test method offers a num-
ber of test design options. The test design option used in the
study was as follows: 20 min total testing time (10-min sperm
exposure and 10-min exposure of eggs and sperm), followed
by a subsequent addition of fixative. The following briefly
describes the test conditions and procedures. Sea urchins were
injected with 0.5 ml of 0.5 M KCl solution and transferred
onto small Petri dishes. Once spawning occurred and the sex
of the organisms was determined, males were inverted dorsal
side down and allowed to dry spawn. Females were transferred
to small borosilicate beakers (which were rinsed with filtered
seawater prior to use) containing 50 ml of filtered seawater.
Organisms were allowed to spawn for up to a maximum of
10 min, during which time careful observations were made
pertaining to the amount of released gametes, their color, and
the overall behavior of the spawners. Organisms providing
relatively little or dilute gametes were excluded from testing.
Gametes obtained from at least three individuals of each sex
were combined and their densities determined. Densities of
gametes were determined microscopically, using a Sedgewick–
Rafter chamber (Wildlife Supply Company, Saginaw, MI,
USA) and a hemocytometer for eggs and sperm, respectively.
At the beginning of the study, the pretrial testing was con-
ducted in order to determine the sperm-to-egg ratio. Fertiliza-
tion rates between 70 and 98% were achieved. Later, during
the testing phase, we established that to achieve that percentage
fertilization, the usual ratio of sperm to eggs was 20,000 to 1,
respectively. After pretrial testing, definitive parallel tests were
conducted using samples with synthetic or natural seawater as
control and dilution water. Testing was performed at 20 Ϯ 1ЊC
under ambient laboratory light (approx. 200–400 lux). One
hundred milliliters of each test concentration were prepared
and homogenized, then 10 ml were transferred to each of four
replicates, leaving 60 ml for physical/chemical monitoring
(temperature, salinity, pH, and dissolved oxygen). Clean, new
borosilicate scintillation vials were used as test vessels. Once
test solutions were prepared and monitored, gamete densities
were established, and appropriate dilution of eggs and sperm
was made, testing was initiated. Initially, sperm were added
to each test vessel and exposed for a duration of 10 min.
Subsequently, approx. 2,000 eggs were transferred to each test
vessel in a carefully timed procedure matching the addition of
sperm. While eggs were added, the test vessels were swirled
in order to facilitate fertilization. Gametes were exposed for
an additional 10 min before the test was completed by the
addition of 10% buffered formalin to each vessel to terminate
the fertilization process and to preserve samples. Subsequent
to preservation, a subsample containing at least 100 eggs was
removed from each vessel and examined microscopically using
a Sedgewick–Rafter chamber to enumerate fertilized and un-
fertilized eggs, and from this ratio, percentage fertilization was
determined. All samples were scored within 3 to 7 d of testing.
Sample salinity adjustment
Synthetic brine was prepared by dissolving dry salts (In-
stant Ocean) in deionized water (30 g of salt in 1 L of deionized
water) to obtain a salinity of 30 Ϯ 2 ppt. When synthetic brine
in liquid form (salinity range ϭ 90–160 ppt) was used as a
source, the brine was aged (conditioned) for a period of 3 to
7 d, stored in darkness at 4ЊC, and used to dilute the sample
to adjust salinity to the required level. When dry salts (Instant
Ocean) in the crystalline form were used as a source, they
were dissolved in the 100% effluent sample to achieve a spe-
cific salinity, allowed to condition for a minimum of 4 h (max-
imum of 12 h), and stored in darkness at 4ЊC. The conditioning
period was required in order to stabilize and equilibrate the
pH of the solutions.

3. 806 Environ. Toxicol. Chem. 20, 2001 E. Jonczyk et al.
Table 1. Results of sea urchin fertilization toxicity tests using
industrial effluents; salinity adjustment with brine and dry salts
(expressed as % effluent)
Mill
effluent
Brine
IC25a
Salts
IC25
Brine
IC50b
Salts
IC50
1
2
3
4
5
6
7
8
9
10
47.1
0.43
59.6
5.6
1.11
9.5
66.4
7.27
6.38
10.3
63
0.72
39.4
5.04
1.11
9.84
59.9
7.56
3.71
7.98
63.7
0.85
Ͼ67
7.73
1.61
12.5
Ͼ72
9.5
7.85
13.8
80.7
1.12
60.7
8.05
2.25
14.2
77.2
9.84
7.08
10.5
11
12
13
14
15
16
17
18
19
20
1.17
19.5
53
14.6
8.5
5.6
7.5
4.93
68.8
7.5
2.11
30.1
53.6
15.1
8.21
3.9
8.9
0.85
96.3
8.9
1.81
39.3
65.4
11.7
11.4
7.9
11.3
6.55
Ͼ90.1
11.3
2.65
44.5
69.2
18.6
10.8
5.9
12
1.35
Ͼ100
12
a Inhibiting concentration 25%.
b Inhibiting concentration 50%.
Dilution water preparation
During the first phase of testing, only synthetic seawater
water with a salinity of 30 Ϯ 2 ppt (obtained by dissolving
commercially available dry salts in deionized water) was used.
The synthetic water was conditioned for a week with vigorous
aeration and then filtered using a Whatman prefilter (Whatman,
Clifton, NJ, USA; ϳ1-␮m pore size) prior to use.
During the second phase of testing, two types of seawater
were used: natural and synthetic. Synthetic water was prepared
by diluting 90-ppt synthetic brine (as described previously)
with deionized water to obtain 30 Ϯ 2 ppt salinity. Natural
seawater was obtained from two sources: Atlantic Ocean sea-
water was supplied by the Environment Canada toxicology
laboratory (Moncton, NB, Canada), and Pacific Ocean sea-
water was provided by the Bamfield Marine Station (Bamfield,
BC, Canada). All control/dilution water was filtered prior to
use and vigorously aerated to achieve dissolved oxygen levels
of 90 to 100% saturation.
Test media and experimental design
Two different test media were used to determine the effects
of adjusting the salinity of effluents and when using natural
versus synthetic seawater. Various industrial effluent samples
were tested, including pulp and paper mill effluents and mining
effluents. Each effluent sample was tested once for each treat-
ment (no replication). A minimum of five and a maximum of
10 test concentrations, a control, and a control with brine,
where appropriate, were tested in quadruplicate. The test con-
centrations were established using a 0.5 dilution of the highest
effluent concentration (100% effluent) with synthetic brine
added. Therefore, less than 100% effluent was tested because
of the addition of brine for salinity adjustment (e.g., when 60
ml of brine was added to 200 ml of freshwater effluent, this
resulted in a highest test concentration of 70%).
In addition to whole industrial effluents, the common ref-
erence toxicant copper sulfate was also tested. A copper stock
solution of 1,000 mg/L of copper was prepared by dissolving
4.4 g of CuSO4·7H2O in 1 L of deionized water. Then, 100
mg/L copper (working stock) solution were prepared by di-
luting 10 ml of the original copper stock (1,000 mg/L) with
90 ml of filtered seawater. Moreover, reference toxicant test
concentrations were prepared by pipetting 1.6 ml of working
stock and diluting it with 198.4 ml of filtered seawater in order
to obtain the highest test concentration (800 ␮g/L). This high-
est test concentration was diluted by a factor of 0.5 with filtered
seawater until the lowest test concentration was achieved. The
copper test concentrations diluted with synthetic seawater were
always the same and ranged from 25 to 800 ␮g/L, while the
copper concentrations prepared with natural (east coast or west
coast) seawater ranged from 0.39 to 800 ␮g/L.
Quality assurance/quality control
In order to reduce any variability among test media, a num-
ber of quality assurance/quality control measures were imple-
mented throughout the study. With regard to test media, the
same effluent sample was tested for each pairwise set of treat-
ments. For test organism quality assurance/quality control, the
same genetic pool of gametes was used, and all tests were
performed using the same sperm-to-egg ratio established in
the pretrial experiments. Furthermore, the same technician
conducted all tests, and the same technician estimated fertil-
ization rate in each side-by-side testing series.
Statistical analyses
The percentage fertilization of sea urchin gametes was com-
pared using samples adjusted with dry salts or brine. Twenty
industrial effluent samples were evaluated in the analysis. The
IC25 and IC50 end points (calculated using linear interpolation
[20]) were compared using the Wilcoxon signed rank test. The
ICp end point is defined as the inhibiting concentration for a
specified percentage effect. It represents a point estimate of
the concentration of test substance that would cause the des-
ignated percent impairment in a biological function, in this
case, fertilization rate.16 In the second phase of the study, the
percentage fertilization of sea urchin gametes using artificial
and natural seawater was compared. Nineteen industrial efflu-
ent samples and eight reference toxicant samples were eval-
uated. The IC25s and IC50s were compared as in the previous
experiment. Using the entire dose response among the treat-
ments, pairwise comparisons of slopes and intercepts, esti-
mated from the dose–response curves, were made under the
assumption of asymptotic normality. This has the advantage
of using the data set in its entirety and obviates the problems
inherent to point estimates [21]. The parameters were esti-
mated using generalized linear models (binomial family and
logit link) for each of the three groups of data (reference tox-
icant, salinity adjustment, and test media).
RESULTS AND DISCUSSION
Analysis of test end-point data
Hypersaline brine versus dry salts comparison. The results
of IC25 and IC50 calculated end points from sea urchin fer-
tilization tests run concurrently on the same effluent are pro-
vided in Table 1. A qualitative comparison of the IC25 and
IC50 end points indicates some substantial differences between
some of the generated values (e.g., effluent 1 [IC25 ϭ 47 vs
63%] or effluent 19 [IC25 ϭ 69 vs 96%] for brine and dry
salts, respectively). However, a Wilcoxon signed rank test on
the paired end points (20 pairs of IC25 data and 20 pairs of

4. Sea urchin fertilization test evaluation Environ. Toxicol. Chem. 20, 2001 807
Table 2. Results of sea urchin fertilization toxicity tests using
industrial effluents; synthetic versus natural seawater as dilution water
(expressed as % effluent)
Mill
effluent
Synthetic
IC25a
Natural
IC25
Synthetic
IC50b
Natural
IC50
1
2
3
4
5
6
7
8
9
10
44.8
23.2
58.7
43.9
0.67
0.42
45
12.1
Ͼ72
20.8
Ͼ70
15.4
Ͼ73
37.4
0.62
0.14
36.4
9.8
Ͼ72
14.9
56
39.1
Ͼ73
55.2
1.05
0.73
56
15.6
Ͼ72
36
Ͼ70
28.1
Ͼ73
52
1.32
0.28
54.4
13.5
Ͼ72
28.3
11
12
13
14
15
16
17
18
19
3
6
17.1
Ͼ67
4.4
4.3
28.7
16.8
4.5
2.4
1.7
14.9
Ͼ67
1.9
1.4
38.7
23.2
1.7
4.1
8
30.1
Ͼ67
6.5
2.4
42.6
28.2
6.7
3.9
4.1
28.3
Ͼ67
3.9
4.9
48.9
29.7
3.6
a Inhibiting concentration 25%.
b Inhibiting concentration 50%.
Table 3. Results of sea urchin fertilization toxicity tests using the
reference toxicant copper sulfate; synthetic versus natural seawater as
dilution water (expressed as ␮g/L Cu)a
Reference
toxicant
experiment
Synthetic
IC25b
Natural
IC25
Synthetic
IC50c
Natural
IC50
1
2
3
4
5
6
7
8
94.9
58.9
206
72.6
72.1
119
96
166
10.8
29.4
148.1
14.3
14.5
33.1
85.1
169
180
90.9
277
120
122
152
135
249
33.4
51.8
243
20.8
20.9
55.7
126
Ͼ200
a Tests 1 to 5 performed using east coast natural seawater; tests 6 to
8 performed using west coast natural seawater.
b Inhibiting concentration 25%.
c Inhibiting concentration 50%.
Table 4. Summary of analyses conducted using the Wilcoxon signed
rank testa
Comparison Test statistic p value
Industrial IC25b
Industrial IC50c
Reference toxicant IC25
Reference toxicant IC50
All data IC25
All data IC50
1.9315
1.664
2.3805
2.3664
3.111
3.0417
0.0534
0.0961
0.0173 A
0.018 A
0.0019 A
0.0024 A
a Tests significant at the ␣ ϭ 0.05 level share the same uppercase
letter.
b Inhibiting concentration 25%.
c Inhibiting concentration 50%.
IC50 data) showed no significant differences for either end
point (IC25: Z ϭ Ϫ0.2614, p ϭ 0.7938; IC50: Z ϭ Ϫ1.5121,
p ϭ 0.1305).
Based on the statistical evaluation of paired comparison
testing using brine and dry salts for sample salinity adjustment,
both yield similar test results using the sea urchin fertilization
assay when tested with industrial effluents. Therefore, it is our
view that the dry salts method should be used for adjusting
salinity in these tests. This way, samples can be tested at full
strength (100% effluent) rather than diluted with hypersaline
brine solution, where the highest concentration can be only
70%. This approach would be particularly useful for evaluating
samples of low toxicity, where the reporting of ‘‘greater than’’
values (e.g., Ͼ70%) can be eliminated.
Natural seawater versus synthetic seawater comparison:
Effluents. The results of IC25 and IC50 calculated end points
from sea urchin fertilization tests run concurrently on the same
effluent are provided in Table 2. Similar to the previously
mentioned qualitative analysis, there were also samples in this
data set that showed differences between test results (e.g.,
effluents 1 and 17). Quantitatively over the whole data set, a
comparison of 19 paired observations yielded no significant
differences for both end points, based on the Wilcoxon signed
rank test (IC25: Z ϭ 1.9315, p ϭ 0.0534; IC50: Z ϭ 1.664,
p ϭ 0.0961) (Table 2).
Overall, based on the evaluation of paired comparison test-
ing using natural and synthetic seawater as control and dilution
water, similar test results were obtained using the sea urchin
fertilization assay when tested with industrial effluents. There-
fore, synthetic seawater, prepared from dry salts, can be sat-
isfactorily used for control/dilution water in these tests. An
important implication for this conclusion is that laboratories
do not require access to natural seawater in order to conduct
this test.
Natural seawater versus synthetic seawater comparison:
Reference toxicant. The results from tests using the reference
toxicant, copper sulfate, are provided in Table 3. Qualitatively,
the majority of results would lead one to conclude that sig-
nificant differences between results exist. This is verified by
the significant differences observed in both end points (IC25
and IC50), based on the Wilcoxon signed rank tests (IC25: Z
ϭ 2.3805, p ϭ 0.0173; IC50: Z ϭ 2.3664, p ϭ 0.018) (Table
4).
When pooling all effluent and reference toxicant data, the
paired comparisons also yield significant differences between
IC25 or IC50s (IC25: Z ϭ 3.111, p ϭ 0.0019, IC50: Z ϭ
3.0417, p ϭ 0.0024). It is apparent that the differences ob-
served in the pooled data set are due mainly to differences
detected in the reference toxicant comparison (Table 4).
Based on the evaluation of dilution water types using a
reference toxicant (copper sulfate), there was a significant dif-
ference among treatments. In other words, tests results are
highly variable for copper sulfate, using the two different di-
lution water types. Analyses of the test end points (IC25 and
IC50) generated using reference toxicant data indicate that
fertilization rates are significantly higher when using artificial
water for dilution rather than natural seawater. This conclusion
is the same when the entire dose response is analyzed. In order
to better understand these results, reference toxicant data for
copper sulfate using natural seawater as control/dilution water
were obtained from Environment Canada (Atlantic Region;
Moncton, NB) for comparison to our laboratory data. Our
laboratory’s reference toxicant data with copper sulfate, using
synthetic seawater as dilution water, yielded IC50s ranging
between 206 and 327 ␮g/L as copper, with a coefficient of
variation of 28%. In comparison, the Environment Canada
laboratory’s data with the same reference toxicant, but using
natural seawater as dilution water, yielded IC50s ranging be-

5. 808 Environ. Toxicol. Chem. 20, 2001 E. Jonczyk et al.
Table 5. Summary of analyses of generalized linear model parametersa
Comparison Parameter
Mean value of parameter
Naturalb Syntheticc
Test
statistic
Degrees
of
freedom p value
Salinity adjustment
Test media
Reference toxicant
Intercepts
Slopes
Intercepts
Slopes
Intercepts
Slopes
2.807131
Ϫ0.3516381
2.084383
Ϫ0.2771331
2.364742
Ϫ0.05030137
2.540522
Ϫ0.2599838
2.829844
Ϫ0.2901326
3.351008
Ϫ0.02357855
1.4403
Ϫ1.8563
Ϫ2.3225
0.391
Ϫ2.0029
Ϫ1.7333
19
19
18
18
7
7
0.1661
0.709
0.0321 A
0.7004
0.0852
0.1266
a Tests significant at the ␣ ϭ 0.05 level share the same uppercase letter.
b Salinity adjustment is made using concentrated brine solution.
c Salinity adjustment is made using dry salts.
tween 53.9 and 317 ␮g/L as copper, with a coefficient of var-
iation of 40%. This comparison indicates that copper yields a
higher variability of response in natural seawater as compared
to synthetic seawater. A further comparison between reference
toxicant and effluent responses is required in order to under-
stand the differences in these results. Originally, copper sulfate
was chosen as the reference toxicant in this study because an
adequate database was available for comparison. However, as
copper toxicity is both pH and hardness dependent, its appli-
cation as a reference toxicant in this study was not appropriate.
Moreover, in light of the variability observed in the results of
this test with copper sulfate and its predisposition to precipitate
at high concentrations when combined with seawater, we do
not believe that it is an ideal standard reference toxicant for
this test. Currently, there are other available chemicals that
can be, and have previously been, used as reference toxicants
(e.g., sodium dodecyl sulfate; [18,19]).
Analyses of dose responses
Again, parameters estimated using the entire dose response
rather than a single point estimate were used to estimate the
effects of dilution water. A summary of the estimated gener-
alized linear model dose–response parameters is provided in
Table 5. The intercepts were significantly larger when the test
media were diluted using synthetic seawater rather than natural
seawater (p value ϭ 0.0321), but the dose responses were
parallel. Therefore, end-point estimates such as the IC25 or
IC50 would be significantly larger when diluting samples with
synthetic seawater.
CONCLUSIONS AND RECOMMENDATIONS
Results from these experiments indicated that no significant
difference exists in test end points when dry salts or brine are
used for sample salinity adjustment. Similarly, results obtained
from parallel (split-sample) industrial effluent tests with nat-
ural and artificial seawater suggest that both dilution waters
produce similar test results. However, data obtained from con-
current tests with the reference toxicant, copper sulfate,
showed higher variability and greater sensitivity when using
natural seawater as control/dilution water.
Based on the results of this study, further research is still
required to determine the cause of differences between efflu-
ents and copper sulfate and to determine the effects of natural
versus synthetic seawater on culturing and holding sea urchins.
In either case, the choice of reference toxicants other than
copper sulfate is recommended.
Moreover, since other marine toxicity tests also have the
option of using natural or synthetic seawater for dilution water
(for culturing and testing), similar studies should also be con-
ducted for other marine test species (e.g., Champia parvula,
mysid shrimp, inland silversides, and topsmelt) currently used
in standard regulatory testing of industrial effluents.
Acknowledgement—The authors wish to acknowledge the technical
assistance of M. Bidlofsky and K. Elliot. K. Doe provided additional
reference toxicant data for copper sulfate.
REFERENCES
1. Dinnel PA, Stober QJ, DiJulio DH. 1981. Sea urchin sperm bio-
assay for sewage and chlorinated seawater and its relation to fish
bioassays. Mar Environ Res 5:29–39.
2. Dinnel PA, Stober QJ, Crumley SC, Nakata RE. 1982. Devel-
opment of a sperm cell toxicity test for marine waters. In Pearson
JG, et al., eds, Aquatic Toxicology and Hazard Assessment: Fifth
Conference. STP 766. American Society for Testing and Mate-
rials, Philadelphia, PA, pp 82–98.
3. Dinnel PA, Link JM, Stober QJ. 1987. Improved methodology
for a sea urchin sperm cell bioassay for marine waters. Arch
Environ Contam Toxicol 16:23–32.
4. Dinnel PA, Pagano GG, Oshida PS. 1988. A sea urchin test system
for marine environmental monitoring. In Burke RD et al., eds,
Echinoderm Biology. Balkema, Rotterdam, The Netherlands, pp
611–619.
5. Dinnel PA, Stober QJ. 1987. Application of the sea urchin sperm
bioassay to sewage treatment efficiency and toxicity in marine
waters. Mar Environ Res 21:121–133.
6. Kobayashi N. 1991. Marine pollution bioassay by using sea urchin
eggs in the Tanabe Bay, Wakayama Prefecture, Japan, 1970–1987.
Mar Pollut Bull 23:709–713.
7. Kobayashi N. 1995. Bioassay data for marine pollution using
echinoderms. In Cheremisinoff PN, ed, Encyclopaedia of Envi-
ronmental Control Technology, Vol 9. Butterworth-Hienemann,
Oxford, UK, pp 539–609.
8. Pagano G, Corsale G, Esposito A, Dinnel PA, Romana, LA. 1989.
Use of sea urchin sperm and embryo bioassay in testing the sub-
lethal toxicity of realistic pollutant levels. Adv Appl Biotechnol
5:153–163.
9. Carr RS, Chapman DC, Howard CL, Biedenbach J. 1996. Sedi-
ment quality triad assessment survey in the Galveston Bay, Texas,
system. Ecotoxicology 5:1–25.
10. Carr RS, Long ER, Chapman DC, Thursby G, Biedenbach JM,
Windom H, Sloane G, Wolfe DA. 1996. Toxicity assessment stud-
ies of contaminated sediments in Tampa Bay, Florida. Environ
Toxicol Chem 15:1218–1231.
11. Bailey HC, Miller JL, Miller MJ, Dhaliwal BS. 1995. Application
of toxicity identification procedures to the echinoderm fertiliza-
tion assay to identify toxicity in a municipal effluent. Environ
Toxicol Chem 14:1–6.
12. Ho KT, Mitchell K, Zappala M, Burgess RM. 1995. Effects of
brine addition on effluent toxicity and marine toxicity identifi-
cation evaluation manipulations. Environ Toxicol Chem 14:245–
249.
13. Thompson BE, Bay SM, Anderson JW, Laughlin JD, Greenstein
DJ, Tsukada DT. 1989. Chronic effects of contaminated sediments

6. Sea urchin fertilization test evaluation Environ. Toxicol. Chem. 20, 2001 809
on the urchin Lytechinus pictus. Environ Toxicol Chem 8:629–
637.
14. Porebski LM, Osborne JM, Doe KG, Zajdlik BA, Lee D, Pock-
lington P. 1999. Evaluating the techniques for a tiered testing
approach to dredged sediment assessment: A study over a metal
concentration gradient. Environ Toxicol Chem 18:2600–2610.
15. Pillard DA, Dufresne DL, Tietge JE, Evans JM. 1999. Response
of mysid shrimp (Mysidopsis bahia), sheepshead minnow (Cy-
prinodon variegatus), and inland silverside minnow (Menidia
beryllina) to changes in artificial seawater salinity. Environ Tox-
icol Chem 18:430–435.
16. Environment Canada. 1992. Biological test method: Fertilization
assay using echinoids (sea urchins and sand dollars), amended
November 1997. EPS 1/RM/27. North Vancouver, BC.
17. U.S. Environmental Protection Agency. 1994. Short-term meth-
ods for estimating the chronic toxicity of effluents and receiving
water to marine and estuarine organisms. EPA/600/4-91/003.
Washington, DC.
18. U.S. Environmental Protection Agency. 1995. Short-term meth-
ods for estimating the chronic toxicity of effluents and receiving
waters to west coast marine and estuarine organisms. EPA/600/
R-95-136. Washington, DC.
19. Nieheisel TW, Young ME. 1992. Use of three artificial sea salts
to maintain fertile sea urchins (Arbacia punctulata) and to con-
duct fertilization tests with copper and sodium dodecyl sulfate.
Environ Toxicol Chem 11:1179–1183.
20. Norberg-King TJ. 1993. A linear interpolation method for sub-
lethal toxicity: Inhibition concentration (ICp) approach. Technical
Report 03-93. U.S. Environmental Protection Agency, Duluth,
MN.
21. Hoekstra JA, Van Ewijk PH. 1993. Alternatives for the no-ob-
served-effect level. Environ Toxicol Chem 12:187–194.