Arch Environ Contam Toxicol (2012) 63:391–399
DOI 10.1007/s00244-012-9774-3
Acute and Chronic Effects of Sodium Tungstate on an Aquatic
Invertebrate (Daphnia magna), Green Alga (Pseudokirchneriella
subcapitata), and Zebrafish (Danio rerio)
Leslie N. Clements • Ranulfo Lemus •
Alicia D. Butler • Kate Heim • Matthew R. Rebstock
Carmen Venezia • Michael Pardus
Received: 8 November 2011 / Accepted: 30 April 2012 / Published online: 26 May 2012
Ó Springer Science+Business Media, LLC 2012
Abstract Although aquatic toxicity data exists for tung-
state substances, insufficient data of high quality and
relevancy are available for conducting an adequate
risk assessment. Therefore, a series of acute and chronic
toxicity tests with sodium tungstate (Na2WO4) were con-
ducted on an aquatic invertebrate (Daphnia magna), green
alga (Pseudokirchneriella subcapitata), and zebrafish
(Danio rerio). Collectively, the data from these studies
suggest that sodium tungstate exhibits a relatively low
toxicity to these taxa under these test conditions. All
studies were conducted in the same laboratory under good
laboratory practice standards using Organisation for Eco-
nomic Co-operation and Development guidelines with the
same stock of test material and the same analytical meth-
ods. All results are reported as mg W/L. The following
toxicity values were based on mean measured concentra-
tions. For D. magna, the 21 day test no-observable effect
concentration (NOEC) was 25.9 mg W/L, and the 48-h
median effective concentration (EC50) from the acute test
was [95.5 mg W/L (the highest concentration tested).
The P. subcapitata test yielded an ErC50 of 31 mg W/L. A
38-day test with zebrafish resulted in an NOEC
C5.74 mg W/L with no effects at any concentration. The
L. N. Clements  R. Lemus (&)  A. D. Butler  K. Heim
ARCADIS, Cincinnati, OH, USA
e-mail: ranulfo.lemus-olalde@arcadis-us.com
M. R. Rebstock
ABC Laboratories, Inc., Columbia, MO, USA
C. Venezia
GTP, Towanda, PA, USA
M. Pardus
ARCADIS, Pittsburgh, PA, USA
•
96-h LC50 from the acute test with zebrafish was
[106 mg W/L. The results of the current acute study for
daphnids and fish are consistent with published literature,
whereas the algae results are different from previously
reported values. Transformation/dissolution (T/D) studies,
which were conducted according to United Nations Glob-
ally Harmonized System of Classification and Labelling of
Chemicals protocol, confirmed that the WO-2 4 anion
accounted for most of the tungsten in solution. For classi-
fication purposes, the algae ecotoxity reference value was
then compared with T/D data and would not classify
Na2WO4 as an aquatic toxicant under the European Union
Classification, Labelling and Packaging scheme.
To maintain access to and expand markets within the
European Union (EU), industrial producers, importers,
users, and distributors of all chemical substances have an
obligation under the Registration, Evaluation, Authorisa-
tion and Restriction of Chemicals (REACH) regulation
(EC 2006) to meet reporting requirements on the human
and environmental health effects of their substances.
As part of REACH, producers and/or importers must
classify the substance according to the Classification,
Labelling and Packaging (CLP) regulation (EC 2008)
guidelines, which is based on the United Nations (UN;
Globally Harmonized System of Classification and Label-
ling of Chemicals [GHS]; UN 2009) as well as provide for
the classification of substances with respect to several
human health and aquatic end points. For metals and
sparingly soluble metal compounds, derivation of hazard
classification levels with respect to the aquatic environment
involves a comparison of acute and chronic ecotoxicity
reference values (ERVs) with dissolved metal concen-
trations released at specified intervals under standard
123
392
transformation/dissolution (T/D) conditions (UN 2009)
into standard ecotoxicity testing media. In the GHS, the
acute ERV to be used for comparison with released metal
concentrations is the lowest of the available ecotoxicity
values among fish, crustacea, and aquatic plants (UN
2009).
Both the UN GHS and the EU CLP guidelines provide for
the independent application of the acute and chronic aquatic
classification categories. To meet the requirements for both
hazard classification schemes, it is therefore necessary to
have acute and chronic ERVs for all three levels of taxa that
have been measured by standardized ecotoxicity testing
protocols, such as those developed by the American Society
for Testing and Materials or the Organisation for Economic
Co-operation and Development (OECD).
Overall, tungsten (W) may be considered a data-poor
substance, in the sense that there are no validated ecotox-
icity data for dissolved W in the open literature. The rel-
ative scarcity of investigations of W ecotoxicity data may
reflect a lower priority given to W testing relative to other
metals with greater perceived toxicity and more ubiquitous
environmental releases.
Few studies have shown stable water-quality conditions
and precise analytical confirmation of test substance con-
centrations during ecotoxicity testing of W and do not
provide the robust characterizations of concentrations and
water chemistry necessary to evaluate W and W com-
pounds by current standards. Moreover, no data in the lit-
erature assess the chronic toxicity of W to standard test
organisms, such as daphnia or zebrafish, and only a few on
the acute ecotoxicity of W.
Upitis et al. (1974) reported a decrease in Chlorella
growth at 30 mg W/L, with no effects apparent at
10 mg W/L. Strigul et al. (2009) examined the effects of
sodium tungstate on Daphnia magna and the algae
Pseudokirchneriella subcapitata (formerly known as
Selenastrum capricornutum) in acute-toxicity tests per-
formed according to OECD 202 and 201, respectively. For
median lethal concentration (LC50) values, they derived
24- and 48-h values of 1.7 and 0.34 g W/L, respectively. In
another study, Strigul (2010) exposed guppies (Poecilia
reticulate) to nominal concentrations of Na2WO4 ranging
from 439.5 to 5,274 mg W/L. They reported 4- and 14-day
LC50s of 2,174 and 902 mg W (nominal), respectively but
did not report a NOEC value.
Because W compounds are mobile in soil, groundwater,
and surface water, along with the increased uses of W
involving potentially direct environmental exposure (e.g.,
production and use in W-containing munitions or use of
monotungstates and polytungstates in the mining industry)
and new reporting requirements under REACH and the
CLP, more accurate determination of W ecotoxicity is
timely.
123
Arch Environ Contam Toxicol (2012) 63:391–399
Materials and Methods
Test Substance Characterization
All toxicity tests were conducted by ABC Laboratories,
Columbia, MO. Sodium tungstate powder was obtained
from Wolfram Bergbau-und Hütten-GmbH, Austria (batch
no. 20080422/XNAW [dated May 19, 2008]). Sodium
tungstate (Na2WO4—1.2H2O) purity was 99.9 % with
WO3 content of 73.6 % (equivalent to 58.6 % as W). Full
trace-metals analyses confirmed the high purity of the
powder, which had a water content of 6.4 % and a pH of
8.3. All data are presented as mg W/L. A 1-g/L W refer-
ence standard (Ricca Chemical Co., Arlington, TX) in 2 %
HNO3 and 1 % HF, and subsequently diluted in 2 %
HNO3, was used as standard for inductively coupled
plasma-mass spectrometry (ICP-MS) analysis of test
solutions.
Dilution Water and Environmental Monitoring
The dilution water used for all tests with fish and Daphnia
was naturally hard well water diluted with demineralized
(by way of reverse osmosis) well water to a total hardness
of 130–160 mg CaCO3/L. The water for use with the
Daphnia tests was biologically aged (held in a tank con-
taining aquatic organisms) and passed through a sediment
filter and ultraviolet sterilizer before use. Total hardness
and alkalinity were measured using titrimetric methods
adapted from standard methods (American Public Health
Association 1998). Conductivity was measured with a
WTW Model Cond 330i conductivity/salinity meter.
Temperature and pH were measured with a WTW Model
pH 330i meter. Dissolved oxygen was measured with a
WTW Oxi 330i dissolved oxygen meter. The temperature
of the water bath in which the test vessels were held was
continuously monitored during the tests using a data-logger
and thermistor probe.
Toxicity Test Methods
Acute and chronic studies with D. magna were conducted
according to OECD guidelines and under good laboratory
practice (GLP) standards. An acute-static 48-h toxicity test
was performed according to OECD Guideline 202 (2004),
and a 21-day chronic-static renewal toxicity test was con-
ducted according to OECD 211 (1998); both tests were
conducted with D. magna neonates obtained from an
in-house culture. The test chambers were maintained at
*20 °C in a temperature-controlled water bath, and fluo-
rescent lighting was maintained at a 16-h daylight photo-
period with light intensity of *445 lux. Acute test
concentrations used were 0, 7.6, 14.6, 29.3, 58.6, and
Arch Environ Contam Toxicol (2012) 63:391–399
117.2 mg W/L, and chronic concentrations used were 0,
3.8, 7.6, 14.6, 29.3, and 58.6 mg W/L. Acute and chronic
test concentrations were selected based on either range-
finding tests or results of previous toxicity tests.
Daily observations were recorded on numbers of sur-
viving adults, occurrence of abnormalities, and production
of neonates. At chronic test termination, the length (head to
base of spine) of each surviving adult was measured.
Concentrations of W were measured at test initiation and at
48 h for the acute test. Chronic concentrations were mea-
sured in fresh test solutions and spent solutions corre-
sponding to the fresh test solutions.
The 72-h algal growth—inhibition tests with P. subcap-
itata (University of Texas at Austin) was conducted with
Na2WO4 according to OECD Guideline 201 (2006). The
test medium was a freshwater algal nutrient medium
(pH 7.5) prepared according to American Society for
Testing and Materials (1997).
Nominal concentrations from the first algal study were
0.23, 0.50, 1.11, 2.4, 5.33, and 11.7 mg W/L and for the
second algal study were 0, 7.6, 14.6, 29.3, 58.6, and
117.2 mg W/L. Samples of the test solutions from each
control and test substance treatment were collected at 0 and
72 h for analysis of W concentrations.
Flasks were incubated at 24 ± 2 °C under continuous
cool-white fluorescent lighting (7,923–8,290 lux). The
flasks were swirled on an orbital shaker table at 100 rpm
throughout the test.
Danio rerio acute static 96-h toxicity test was performed
according to OECD Guideline 203 (1992a), and an early
life stage (ELS) test conducted under flow-through condi-
tions and according to OECD Method 210 (1992b) was
conducted with zebrafish (Osage Catfisheries, Osage
Beach, MO). These fish were fed ad libitum with salmon
starter (Rangen, Buhl, ID), flake food (World Wide
Aquatics, Arvin, CA), and brine shrimp (Brine Shrimp
Direct, Ogden, UT) daily. Food was withheld 1 day before
test initiation, and fish were not fed during the test.
For the acute test, the fish used to initiate the test were
*3 months old. The control fish were measured at test
termination and ranged from 28 to 31 mm in total length
(mean ± SD = 30 ± 1.2 mm) and from 0.194 to 0.408 g
blotted wet weight (mean ± SD = 0.276 ± 0.0664 g). In
the acute test, the instantaneous biomass loading rate of the
control replicate was 0.18 g fish tissue/L test solution.
Final test concentrations used were 0, 7.6, 14.6, 29.3, 58.6,
and 117.2 mg W/L. Concentrations were selected for this
test based on the results of previous toxicity tests.
The 21-L test chambers with 10 fish present in each test
chamber were placed in a temperature-controlled water
bath with a target temperature range of 23 ± 1 °C. Fluo-
rescent lighting was maintained on a 16-h daylight photo-
period with 30-min each simulated dawn and dusk periods.
393
Observations of acute mortality and sublethal responses
were recorded every 24 h. Samples of the test solutions
from each control and test substance treatment were col-
lected at 0 and 96 h for analysis of W concentrations.
In the chronic fish ELS study, newly fertilized eggs were
obtained from an in-house culture. A 2-L proportional
equal solvent diluter (Mount and Brungs 1967) with an
FMI MicroPpetter metering pump delivered test solutions
to glass aquaria (5-L solution volume) and replicated four
times per treatment and control. Average solution flow
rate was two cycles/h ([5 volume additions/day). Nomi-
nal concentrations of 0, 0.37, 0.76, 1.46, 2.93, and
5.86 mg W/L were selected for this test based on the
results of a range-finding test. The highest nominal treat-
ment level equaled the maximum test concentration of
10 mg/L as Na2WO4 as required by OECD guideline 210
(1992b). During each diluter cycle, 40-mL volumes of the
diluter stock were introduced to the diluter system by way
of the metering pump, and the delivery volume was con-
firmed every 3 days. Developing embryos were incubated
in 9-cm glass cups with NitexÒ screen bottoms.
The test was initiated when at least 20 embryos were
distributed to an egg cup in each of four test chambers at
each dose (at least 80 embryos/treatment group). Aquaria
were held in a temperature-controlled water bath set at a
target temperature of 25 °C and incubated under an elec-
tronically controlled 16:8-h light-to-dark photoperiod with
simulated 30-min dawn and dusk. Embryos were observed
daily, and dead embryos were counted and removed. When
all living embryos had hatched, lengths of time to 95 %
hatch and % hatchability were recorded. Day 0 posthatch
was based on C95 % hatch in the control group (study day
6). One day after 95 % posthatch was achieved (study day
7), live fry and unhatched embryos were transferred from
egg cups to Petri-dish baskets with NitexÒ screen collars
and placed in test chambers.
After the first fry were observed on test day 4, feeding
with a larval fish food supplement and live rotifers
(Brachionus sp.) was initiated at least twice daily and later
supplemented with brine shrimp (Artemia) nauplii. After
32 days of posthatch growth (study day 38), fish were
killed and measured for standard length and blotted wet
weight. The water concentration of the test substance was
measured in test solution samples collected from each
control and treatment solution before initiation and on days
0, 7, 14, 21, 28, 35, and 38 of the definitive test and used to
calculate W concentration in the samples.
Test Substance-Concentration Verification
Samples of the test solutions were collected and analyzed
to determine the concentration of W. Samples were cen-
trifuged for 15 min, and a 0.05-mL volume of the
123
394
supernatant was removed from each 10-mL centrifuged
sample and diluted to 10 mL with 2 % HNO3. Quality-
control fortifications, bracketing the target nominal test
concentrations, were prepared in a similar manner after the
control medium had been fortified with the test substance.
All samples were directly analyzed using a inductively
coupled mass spectrometer (Agilent 7500A ICP-MS;
Babington nebulizer; normal bore torch without shield).
Concentrations of W in the test solutions were determined
from the linear regression of the W standard curve gener-
ated from ICP-MS analysis. Average recovery, calculated
as a percentage of the corresponding nominal concentra-
tion, and minimum quantifiable limit (MQL [lowest ana-
lytical standard concentration 9 dilution factor]) for
Na2WO4 was calculated as follows (Eq. 1):
ð0:200 ng W=mLÞð10:3 mLÞ
MQL ¼ ¼ 0:206 ng W/mL
10 mL
ð1Þ
T/D Study
T/D studies were conducted according to method described
by Skeaff et al. (2008). Briefly, 1-L Schott-DuranÒ jars
containing the weighed amounts of Na2WO4 were agitated
in a refrigerated, temperature-controlled New Brunswick
Annova 43 orbital shaker. The aqueous T/D media were
based on the composition of the OECD 203 medium for
acute fish toxicity testing (OECD 1992a). Tests were
conducted with 100-mg/L loadings in triplicate and
screened at 24-h intervals at pH 8.5 and temperature
20–23 °C.
Tungstate (WO-2 4 ) speciation and total dissolved W was
adapted from the method of Bednar et al. (2007). The T/D
solution samples were split into two streams: one stream
was introduced into a Dionex ICS-3000 high-performance
liquid chromatography (HPLC) column onto which the
WO42- anions were loaded and then eluted with a weak
solution of KOH, and a second stream for total dissolved W
was passed directly into the HPLC eluent. The coupled
solution advanced directly to a Perkin ElmerÒ Elan 6100
simultaneous inductively coupled plasma–mass spectrom-
eter interfaced with a concentric nebulizer, with system
retention times of 1.7 and 4.3 min for W and tungstate,
respectively. The solutions resolved into two separate
peaks, the areas under which were proportional to their
concentrations.
Data Analysis
Statistical analyses of data from all aquatic toxicity tests
were performed using SASÒ software version 9.1 (SASÒ,
Cary, NC). NOECs and LOECs for each test were
123
Arch Environ Contam Toxicol (2012) 63:391–399
determined using one-way analysis of variance (ANOVA),
Fisher’s exact test with Hochberg adjustment, and one-
tailed Dunnett’s test (p B 0.05). Before ANOVA and
Dunnett’s test, Shapiro-Wilk test and Levene’s tests were
conducted to test for normality and homogeneity of vari-
ance, respectively. End points meeting parametric criteria
for normality and homogeneity of variance were examined
using parametric analyses. Others were evaluated using
Dunnett’s test on the ranked data when nontransformed and
transformed data exhibited nonnormality, inequality, or
variance. Dunnett’s test was used in all circumstances in
the ANOVA for algae. ECx estimates for continuous and
count data (e.g., reproduction, growth, growth rate, and
yield) were calculated using a four-parameter logistic
(sigmoid-shaped) model fit to the data with percent inhi-
bition as the dependent variable and concentration as the
independent variable. ECx estimates for dichotomous data
(e.g., death, immobilization) and 95 % confidence limits
(CLs) were calculated using probit method or Trimmed
Spearman–Karber method.
Results
Daphnia magna Toxicity Tests
In the acute test with D. magna, analytical confirmation of
the Na2WO4 test solutions was performed at 0 and 48 h.
The mean measured concentrations were 6.9, 12.6, 24.3,
53.2, and 95.5 mg W/L, which represented 82–92 % of the
nominal concentrations. Results were similar at the
beginning and end of the test with no decrease in test
concentrations. No residues of the test substance were
detected in the controls higher than the MQL. Recoveries
from quality-control samples ranged from 80 to 93 % of
the nominal concentrations at 0 and 48 hours. In the
chronic test with D. magna, time-weighted, mean measured
concentrations were 3.4, 6.7, 12.4, 25.9, and 49.9 mg W/L;
these values were 85–89 % of nominal concentrations.
For the D. magna acute toxicity test, environmental
condition ranges were as follows: temperature 19.3–19.5 °C;
dissolved oxygen 7.6–7.9 mg/L; pH 8.3–8.5; total alkalinity
142 mg CaCO3/L; total hardness 144 mg CaCO3/L; spe-
cific conductivity 347 lS; photoperiod: 16:8-h light to dark
and light intensity: 448 lux.
Test conditions monitored at solution replacement
intervals during the 21-day exposure for D. magna ranged
from temperature 19.2–19.9 °C in fresh and spent solu-
tions; dissolved oxygen 7.4–8.6 mg/L (85–99 % of satu-
ration); pH 8.2–8.5; total alkalinity 144–148 mg CaCO3/L;
total hardness 142–148 mg CaCO3/L; and conductivity
347–418 lS.
Arch Environ Contam Toxicol (2012) 63:391–399 395

The test acceptability criteria, according to OECD
Guideline 202 (2004), were met for this study based on
mortality and immobility in the control. Based on mean
measured concentrations, the 48-h median effective con-
centration (EC50) value was [95.5 mg W/L (the highest
concentration tested) for immobility. The 48-h NOEC was
determined to be 53.2 mg W/L based on a statistically
significant (p = 0.05) increase in immobility only at the
highest concentration.
The test acceptability criteria according to OECD
Guideline 211 (OECD 1998) were met for the chronic
Daphnia toxicity test study. After 21 days of exposure, no
statistically significant effects on survival or mobility of
D. magna were observed (Table 1). The 21 day NOEC for
these end points was 49.9 mg W/L (mean measured con-
centration), and the definitive 21-day EC50, based on
immobilization of the first generation daphnids, was
[49.9 mg W/L. Adult survival was the least sensitive end
point. Numbers of live young produced by first-generation
daphnids (both total per concentration and mean per indi-
vidual adult) after 21 days were significantly decreased at
the highest mean measured concentration (49.9 mg W/L),
resulting in NOEC, LOEC, and EC50 values for this end
point of 25.9, 49.9, and 46.1 mg W/L (with 95 % CLs of
43.6 and 48.6), respectively (Table 4). Significant increases
in time to first brood release, decreases in mean length at

Table 1 Chronic (21-day) effects of sodium tungstate on D. magna survival, reproduction, and adult growth
Nominal Time-weighted mean Total first- Total no. Mean no. of live Day of treatment: Mean
concentration concentration in mg W/L generation of live young/surviving adult first brood release length
(mg W/L) (% of nominal concentration) survival (%) young (mma)

Control – 9 (90) 886 98 (CV = 28 %) 8 4.2

3.8 3.35 (88) 9 (90) 1,036 115 8 4.3
7.8 6.74 (89) 10 (100) 1,069 107 8 4.3
14.6 12.4 (85) 10 (100) 1,089 109 8 4.3
29.3 25.9 (88) 10 (100) 934 93 8 4.2
58.6 49.9 (85) 8 (80) 317b 40b 11b 3.7b
a
Mean length of all surviving individual adult daphnid lengths for the treatment
b
Statistically significant decrease (Dunnett’s test, p = 0.05) compared with control

Table 2 Effects of 72-h exposure to W in the initial (low concentration) and second (high concentration) tests examining cell density and growth
rate of P. subcapitata
Nominal Mean concentration in 72-h mean cell 72-h mean yieldb % Change 72-h mean growth % Change in
concentration mg W/L densityb in cells/mL 9 104 in mean rateb in mean growth
(mg/L) (% nominal)a (cells/Ml 9 104) (% CV) yield cells/mL/h (% CV) rate

High-concentration algal test

Control – 75 – – 0.0695 (0) –
7.6 6.39 (84) 23 – – 0.0532 (4)c -23
14.6 12.8 (88) 8.7 – – 0.0396 (3)c -43
29.3 24.4 (83) 6.0 – – 0.0342 (10)c -51
c
58.6 49.5 (85) 5.4 – – 0.0330 (6) -53
117 98.4 (84) 2.9 – – 0.0241 (11)c -65
Low-concentration algal test
Control – 113 113 (4) – 0.0753 (1) –
0.23 0.205 (90) 110 110 (2) -3 0.0749 (0) -1
0.34 0.476 (96) 113 113 (5) 0 0.0753 (1) 0
1.1 1.05 (94) 90 89.2 (4)c -21 0.0721 (1)c -4
2.4 2.24 (93) 72 71.8 (9)c -36 0.0691 (2)c -8
c
5.3 4.66 (87) 58 57.8 (7) -49 0.0661 (1)c -12
11.7 10.4 (89) 31 30.2 (8)c -73 0.0571 (2)c -24
a
Average of analytical measurements at 0 and 72 h
b
Average among treatment replicates
c
Significant growth decrease compared with control (Dunnett’s test, p = 0.05)

123
396
test termination, and decrease in adult mean length were
consistent at the highest dose. Based on the mean length of
surviving adults, the 21-day NOEC and LOEC values were
25.9 and 49.9 mg W/L (mean measured concentrations),
respectively. The EC50 could not be calculated (Table 4).
Pseudokirchneriella subcapitata Toxicity Tests
In both tests, samples collected at 0 and 72 h were aver-
aged to determine mean measured concentrations. In the
first test, performed at lower concentrations, mean mea-
sured concentrations were as follows: \MQL, 0.2, 0.5, 1,
2.2, 4.7, and 10.4 mg W/L. Recoveries ranged from 87 to
96 % of nominal. Recoveries from quality-control spiked
samples were 92–106 %; recovery in the 0.39-mg/L
(nominal) abiotic control was 101 %. In the second test,
performed at greater concentrations, mean measured con-
centrations were as follows: \MQL, 6.4, 12.8, 24.4, 49.5,
and 98.4 mg W/L. Recoveries ranged from 83 to 88 % of
nominal. Recoveries from quality-control spiked samples
were 84–91 %; recovery in the 7.6 mg W/L (nominal)
abiotic control was 83 %.
Test conditions recorded during both tests remained
within acceptable limits throughout the 72-h exposure period
as follows: temperature 22.8–24.0 °C; pH 7.4–8.4; photo-
period continuous light; and light intensity 7,923–8,479 lux
(for tests 1 and 2).
The test acceptability criteria specified in OECD
Guideline 201 (OECD 2006) were met for both algal
studies. In the initial, low-concentration study, results were
calculated based on mean measured concentrations as lis-
ted in Table 2. Mean cell density in the control after 72
hours was 113 9 104 cells/mL (226 times the initial cell
density). The inhibition-of-growth rate ranged from 0 % at
the second lowest concentration (0.476 mg W/L) to 24 %
at the highest concentration. Significant decreases in mean
yield were observed at W concentrations [0.476 mg W/L.
From these data, the NOEC value was determined to be
0.476 mg W/L based on yield and growth rate. This NOEC
for growth rate, however, must be considered as an artifact
of the unusually low coefficient of variation (CV) that
occurred in this test (0, 1, or 2 % CV at all test concen-
trations). The growth curves at low concentrations and the
control were quite similar, with growth rates being linear
with concentrations and with low variability.
The next highest test concentration to the NOEC, i.e.,
1.05 mg W/L, exhibited only a 4 % inhibition-of-growth
rate, and only 8 % inhibition occurred at 2.24 mg W/L.
These small effects are almost certainly not biologically
significant, despite being statistically significant, due to the
unusual precision of the test. Therefore, the ErC10 found
for growth-rate inhibition (3.37 mg W/L) with 95 % CLs
of 2.89 and 3.86 mg W/L is considered to be a more
123
Arch Environ Contam Toxicol (2012) 63:391–399
accurate reflection for chronic toxicity to P. subcapitata.
The ErC20 and ErC50 values, based on growth rate, were
8.09 and [10.4 mg W/L, respectively (Table 4). Based on
yield, the estimated 72-h EyC10, EyC20, and EyC50 were
0.6, 1.2, and 4.3 mg W/L, respectively, with 95 % CLs of
3.86 and 4.76 mg W/L for the EyC50 value.
The second, high-concentration study was designed to
bracket the ErC50 value, which was estimated in the initial
study to be [10.4 mg W/L. The mean cell density in the
control after 72 h was 75 9 104 cells/mL (150 times the
initial nominal cell density; Table 2). Significant decreases
in mean cell density and growth rate at 72 h were observed
at all doses. The 72-h EC50 based on growth rate was
calculated to be 31.0 mg W/L (95 % CL 23.7–38.3 mg/L),
but the NOEC value (\6.39 mg W/L) could not be pre-
cisely determined (Table 4).
Danio rerio Acute Toxicity and ELS Tests
In the acute test with D. rerio, Na2WO4 concentrations were
measured in test solutions at 0 and 96 h. Mean measured
concentrations in the test solutions during the 96-h study
were 0, 7.03, 13.5, 27, 52.3, and 181 mg W/L, which rep-
resented recoveries of 89–92 % of the nominal concentra-
tions. Recoveries from quality-control fortifications ranged
from 84 to 95 % of the nominal concentrations. In the ELS
study, the mean measured concentrations in the test solu-
tions were 0.4, 0.7, 1.5, 2.8, and 5.75 mg W/L, which
represented 92–103 % of the nominal concentrations.
Recoveries from quality-control fortifications ranged from
80 to 104 %.
In the acute fish test, conditions were as follows: tem-
perature 22.7–23.1 °C; dissolved oxygen 6.7–8.3 mg/L
(82–101 % saturation); pH 8.0–8.3; total alkalinity
150 mg CaCO3/L; total hardness 140 mg CaCO3/L; and
specific conductivity 328 lS. In the ELS test, ranges in test
conditions monitored weekly were as follows: temperature
24.3–25.0 °C; dissolved oxygen 6.7–8.2 mg/L (85–104 %
saturation); pH 8.2–8.3; and total hardness 136–148
mg CaCO3/L. Routine water-quality measurements of
alkalinity, hardness, conductivity, total organic carbon, and
total suspended solids measured in the dilution water dur-
ing this study were within acceptable limits.
The test acceptability criteria specified in OECD Guide-
line 203 (OECD 1992a) were satisfied in the acute test per-
formed with zebrafish because there was no mortality in the
controls. After 96 h of exposure, there was no mortality in
any of the test substance treatments. The estimated 24-, 48-,
72-, and 96-h LC50 values for zebrafish exposed to Na2WO4
were 106 mg W/L, the highest concentration tested. The
96-h NOEC was C106 mg W/L based on the lack of mor-
tality and sublethal effects at this and lower test substance
concentrations. The test acceptability criteria for specified in
Arch Environ Contam Toxicol (2012) 63:391–399 397

Table 3 Effects of W on D. rerio in a 32-day (posthatch) early life stage toxicity test
Nominal Mean measured concentration % % Hatch of 32-Day posthatch resultsa
concentration (mg in mg W/L (SD) Nominal embryosa
W/L) Fry Mean standard length Mean wet weight
survival in mm (SD) in mg (SD)
(%)

Control – – 98 92 11.6 (1.2) 23.5 (7.3)

0.37 0.365 (0.0732) 96 100 88 12.3 (1.3) 26.8 (8.9)
0.76 0.703 (0.0629) 92 96 84 11.8 (1.5) 25.2 (8.7)
1.46 1.51 (0.172) 103 96 94 11.6 (1.1) 24.6 (7.7)
2.9 2.78 (0.432) 95 98 95 11.6 (1.5) 23.7 (8.4)
5.86 5.75 (0.607) 98 99 89 11.7 (1.3) 24.5 (8.3)
a
No statistically significant adverse effects were observed (Fisher’s exact test for % hatch and fry survival; p \ 0.05; Dunnett’s test for length
and weight, p \ 0.05)

Table 4 Toxicity of sodium tungstate to D. magna, P. subcapitata, and D. rerio testing

Biological parameter NOEC/EC10 (mg W/L) LOEC (mg W/L) EC50 (95 % CL) (mg W/L)

48-hour D. magna results

Immobility 53.2 (NOEC) 95.5 [95.5
21-Day D. magna results
Adult survivalb 49.9 (NOEC) [49.9 [49.9 (b)
Live young/surviving adultb 25.9 (NOEC) 49.9 46.1(43.6–48.6)
Adult lengthb 25.9 (NOEC) 49.9 [49.9b
72-hour P. subcapitata results
Growth rate (high-concentration test) 6.39 – 31.0 (23.7–38.3)
Growth rate (low-concentration test) 0.476 (NOEC) – [10.4b
3.37 (2.89–3.86) (EC10)
96-hour D. rerio results
Mortality C106 (NOEC) [106 [106
Sublethal observations C106 (NOEC) [106 [106
38-Day (32 days posthatch) D. rerio results
MATC – – –c
Egg hatchability C5.74 (NOEC) [5.74 [5.74
Fry survival C5.74 (NOEC) [5.74 [5.74
Standard length C5.74 (NOEC) [5.74 [5.74
Blotted wet weight C5.74 (NOEC) [5.74 [5.74
Values are mean measured concentrations unless otherwise noted
a
Results calculated based on time-weighted mean measured concentrations
b
Could not be calculated

Guideline 210 (OECD 1992b) were satisfied in the chronic
zebrafish test (see Table 3).
Fry survival ranged from 84 to 95 % with no apparent
dose–response relationship; no significant differences in fry
survival were observed at any concentration (Fisher’s exact
test, p \ 0.05). Based on hatching success, time to hatch, and
fry survival, the mean measured NOEC and LOEC values
were estimated to be C5.74 and[5.74 mg W/L, respectively
(Table 4). The maximum allowable toxicant concentration
(MATC) for egg hatchability could not be calculated due to
the lack of a statistically significant response.
Mean standard length in the control was 11.6 mm (range
11.6–12.3) in the test substance treatments. Mean blotted
wet weight in the control was 0.0235 g/fish (range
0.0237–0.0268 g/fish) in the test substance treatments.
Growth of surviving fry at test termination (day 38
[32 days posthatch]) based on standard length and blotted
wet-weight measurements showed no significant adverse
effects (Dunnett’s test, p \ 0.05) at any concentration
(Table 4). The NOEC and LOEC values for these end
points were C5.74 and [5.74 mg W/L, respectively
(Table 4). The maximum calculated dynamic biomass

123
398
loading at test termination was determined to be 0.0188 g/
L/day for all treatments.
Na2WO4 T/D Studies
Na2WO4 was rapidly and completely soluble, achieving
total dissolved W concentrations of *66,935 lg/L in 24-h
screening tests at pH 8.5, which is *20 % greater than cal-
culated for 100 % dissolution. However, the corresponding
WO-2 4 concentration of 75,585 lg/L is as expected for 101 %
dissolution. The ratio between WO-2 4 measured to W mea-
sured was greater than 0.624 (ratio between WO-24 calculated
to W calculated), suggesting that the WO-24 anion accounted
for most of the W in solution (CANMET-MMSL 2010).
Discussion and Conclusion
The standard testing described here was performed to
simulate natural environmental conditions within a pH
range of 7.5–8.5. It is commonly known that the pH of
most freshwater systems ranges from 5 to 8.5. Over a pH
range between 7 and 10, the monomeric W oxyanion is the
only W species if no other ion or complexation agents are
present in the aquatic solution (Strigul 2010). Our T/D
studies confirm that WO-2 4 was the dominant W species in
solution during the tests. For additional details on fate and
transport as well as W speciation, refer to Strigul (2010)
and Koutsospyro et al. (2006).
The ecotoxicity of Na2WO4 to aquatic organisms from
three trophic levels—an aquatic invertebrate (D. magna), a
green alga (P. subcapitata), and zebrafish (D. rerio)—was
examined according to OECD guidelines and GLP. A sum-
mary of these results is in Table 4. The results from the acute
tests with D. magna, P. subcapitata, and D. rerio showed LC/
EC50 values of [95.5, 31.0 (based on growth rate), 4.31
(based on yield), and [106 mg W/L, respectively. The
results from the chronic tests showed NOEC values of 25.9,
3.37 (based on the ErC10), and C5.74 mg W/L for daphnids,
algae, and zebrafish, respectively. Thus, our results indicate
that of the three taxa tested, algae are the most sensitive
species to dissolved Na2WO4 exposures.
To our knowledge, data on the chronic toxicity of
Na2WO4 to an invertebrate or a fish species have not been
reported in the published literature. The Daphnia and fish
data in this study are consistent with the limited published
data available on Na2WO4. Algae seem to be more sensi-
tive to Na2WO4 than the above-mentioned two taxa, and
results reported in the literature are more variable. Col-
lectively, they support the relatively low toxicity of
Na2WO4 to these taxa.
The results of 31.0 mg W/L for green algae from the
current study are comparable with the decrease in
123
Arch Environ Contam Toxicol (2012) 63:391–399
Chlorella growth reported by Upitis et al. (1974), at
30 mg W/L, with no effects apparent at 10 mg W/L. In the
algal growth-inhibition test by Strigul et al. (2009), the
48-h EC50 value of 344 mg W/L is consistent with
the findings in the current study, which found the 48-h
EC50 value for D. magna to be [163 mg W/L.
For algae, our 72-h EC50 value of 4.31 mg W/L is
comparable with the 72-h EC75 value of 2.42 g W/L
reported by Strigul et al. (2009). Strigul et al. (2009) did
not report other effect levels for algae; their EC50 value of
800 mg W/L, inferred from their graphical representation
of yield and the EC50 value, are in conflict with results seen
in the current study, which found 50 % inhibition (based on
yield) at 4.3 mg W/L. In addition, they noted that Na2WO4
at the lowest concentration tested (76.2 mg W/L), stimu-
lated algal growth, which we did not observe. However,
there appear to be methodological and data handling dif-
ferences between the work of Strigul et al. 2009 and the
current study, thus preventing direct comparisons.
Although OECD method 201 was cited as the methodol-
ogy used by Strigul’s group, the investigators reported using
a commercially available kit (Algatoxkit, MicroBioTest Inc,
Gent, Belgium) without describing the extent of similarity
between this kit and the OECD 201 method. It is possible that
differences in the form of W or the solubility could have
resulted from differences in the test media used and thus
could have contributed to the lower toxicity observed in
algae.
The values of 2,174 and 902 mg W (nominal), respec-
tively, for 4- and 14-day LC50 values reported by Strigul
et al. (2010) for guppies (P. reticulate), although an order
of magnitude greater, are supported by the EC50 value
found for acute toxicity to zebrafish of [106 mg W/L in
the current study.
In addition to strengthening the quality and depth of data
available for Na2WO4, the current aquatic toxicity tests
were conducted to support the registration of W and W
compounds under the requirements of the REACH and
CLP legislation. Data generated were used for determining
predicted no-effect concentrations for exposure of aquatic
organisms to soluble W compounds, and to classify
Na2WO4 itself, as well as being used as a read-across to
other W compounds of similar solubility.
Aquatic toxicity classification of inorganic metals and
metal compounds is conducted by comparing T/D data for
the substance generated using the standard protocol (UN
GHS 2009, Annex 10), with toxicity data for the most
soluble metal substance as described in the CLP technical
guidance (Application of Classification Criteria to Metals
and Metal Compounds [section IV.5]) (EC 2008). The
Na2WO4 T/D data are compared with the ERV for the most
sensitive aquatic species for Na2WO4 to determine its CLP
classification. As indicated previously, the Na2WO4 24-h
Arch Environ Contam Toxicol (2012) 63:391–399
T/D data derived at pH 8.5 was found to be equal to
66.9 mg W/L. This T/D value was compared with the ERV
from the acute algae ErC50 value of 31 mg W/L.
This ERV is used for classification by comparing it with
the aquatic toxicity cut-off values for CLP classification.
Because the acute value (49.5 mg Na2WO4/L) is [10 and
B100 mg/L, Na2WO4 classifies this as Acute Category 3 for
aquatic toxicity according to the CLP. However, because the
lowest no-effect chronic value of 5.76 mg Na2WO4/L,
based on the ErC10 (3.37 mg W/L), is [1 mg/L, Na2WO4
does not receive an acute or chronic classification under the
CLP regime.
In summary, Na2WO4 study results mentioned previously
enhance our knowledge on the acute and chronic ecotoxicity
of dissolved W. Acute, and for the first time, chronic effects
of a well-characterized stock of Na2WO4 powder were
evaluated with robust monitoring of water-quality condi-
tions and Na2WO4 concentrations during testing. These data
were subsequently used to derive hazard-classification pro-
posals and risk assessments of inorganic W substances for
REACH and CLP. Taken as a whole, the study results col-
lectively showed a relatively low acute and chronic toxicity
of Na2WO4 to Daphnia and fish. Algae seem to be more
sensitive to Na2WO4 than the above-mentioned two taxa but
in the overall hazard classification context Na2WO4 is not
considered an acute or chronic hazard.
Acknowledgments The authors are grateful for the assistance of
Maureen Niemeier (ARCADIS) in compiling and formatting data as
well as John Aufderheide (ABC Laboratories, Inc.), and Jim Skeaff
(CANMET–MMSL) for their critical reviews of the manuscript.
Funding for the research presented in this manuscript was provided by
the Tungsten REACH Consortium. All studies and results were
developed in support of registration of W substances in accordance
with the EU legislation on Registration, Evaluation, Authorisation
and Restriction of Chemicals (REACH).
References
American Public Health Association (1998) Methods 2320 and 2340.
In: Greenburg AE, Trussell RR, Clesceri LS (eds) Standard
methods for the examination of water and wastewater, 20th edn.
APHA, American Water Works Association, Water Pollution
Control Federation, Washington, DC
American Society for Testing and Materials (1997) Standard guide for
conducting static 96 h toxicity tests with microalgae. ASTM
E1218-97a
Bednar AJ, Mirecki JE, Inouye LS, Winfield LE, Larson SL,
Ringelberg DB (2007) The determination of tungsten, molybde-
num, and phosphorus oxyanions by high performance liquid
chromatography inductively coupled plasma mass spectrome-
tery. Talanta 72:1828–1832
399
CANMET-MMSL (2010) Transformation/dissolution studies on
tungsten compounds, and tungsten metal. Report no. 09-005
(CR)
European Commision (EC) (2006) Regulation No 1907/2006 of the
European Parliament and of the Council of 18 December 2006
concerning the Registration, Evaluation, Authorisation and
Restriction of Chemicals (REACH), establishing a European
Chemicals Agency, amending Directive 1999/45/EC and repeal-
ing Council Regulation (EEC) No 793/93 and Commission
Regulation (EC) No 1488/94 as well as Council Directive
76/769/EEC and Commission Directives 91/155/EEC, 93/67/
EEC, 93/105/EC and 2000/21/EC. Official Journal of the
European Union L 396, 30.12.2006, 1–849
European Commission (EC) Regulation (2008) No 1272/2008 of the
European Parliament and of the Council of 16 December 2008
on classification, labeling and packaging of substances and
mixtures, amending and repealing Directives 67/548/EEC and
1999/45/EC and amending Regulation (EC) No 1907/2006.
Official Journal of the European Union L 353, 31.12.2008,
1–1355
Koutsospyro A, Braida WJ, Christodoulatos C, Dermata D, Strigul
NS (2006) A review of tungsten: from environmental obscurity
to scrutiny. J Hazard Mater 136:1–19
Mount DI, Brungs WA (1967) A simplified dosing apparatus for fish
toxicological studies. Water Res 1:21–29
Organisation for Economic Co-operation and Development (1992a)
OECD guidelines for testing of chemicals. Guideline no. 203:
fish, acute toxicity test
Organisation for Economic Co-operation and Development (1992b)
OECD guidelines for testing of chemicals. Guideline 210: fish,
early life-stage toxicity test
Organisation for Economic Co-operation and Development (1998)
OECD guidelines for testing of chemicals. Guideline 211:
Daphnia magna reproduction test
Organisation for Economic Co-operation and Development (2004)
OECD guidelines for testing of chemicals. Guideline 202:
Daphnia sp., acute immobilisation test
Organisation for Economic Co-operation and Development (2006)
OECD guidelines for testing of chemicals. OECD guideline no.
201: freshwater alga and cyanobacteria, growth inhibition test.
Official Journal of the European Union, 2007
Skeaff JM, Hardy DJ, King P (2008) A new approach to the hazard
classification of alloys based on transformation/dissolution.
Integr Environ Assess Manag 4(1):75–93
Strigul N (2010) Does speciation matter for tungsten ecotoxicology?
Ecotoxicol Environ Safe 73:1099–1113
Strigul N, Galdun C, Vaccari L, Ryan T, Braida W, Christodoulatos C
(2009) Influence of speciation on tungsten toxicity. Desalination
248:869–879
Strigul N, Koutsospyros A, Christodoulatos C (2010) Tungsten
speciation and toxicity: acute toxicity of mono- and poly-
tungstates to fish. Ecotoxicol Environ Safe 73:164–171
United Nations Globally Harmonized System of Classification and
Labelling of Chemicals (2009). Second revised edition. New
York and Geneva. Annex 10: Guidance on transformation/
dissolution of metals and metal compounds.
Upitis V, Nollendorfa A, Pakalne D (1974) Little-investigated trace
elements in Chlorella: tungsten. Vestis Lat Psr Zinat Akad
9(326):34–35
123