Content uploaded by Scott Markich
Author content
All content in this area was uploaded by Scott Markich on Jul 22, 2021
Content may be subject to copyright.
© 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC wileyonlinelibrary.com/ETC
Environmental Toxicology and Chemistry—Volume 40, Number 8—pp. 2334–2346, 2021
Received: 25 January 2021
|
Revised: 2 March 2021
|
Accepted: 22 April 2021 2334
Hazard/Risk Assessment
Assessing the Toxicity of Mine‐Water Mixtures and
the Effectiveness of Water Quality Guideline Values in
Protecting Local Aquatic Species
Melanie A. Trenfield,
a,
* Ceiwen J. Pease,
a
Samantha L. Walker,
a
Scott J. Markich,
b
Chris L. Humphrey,
a
Rick A. van Dam,
c
and Andrew J. Harford
a
a
Environmental Research Institute of the Supervising Scientist, Darwin, Northern Territory, Australia
b
Aquatic Solutions International, “Point Break”, North Narrabeen Beach, New South Wales, Australia
c
WQadvice, Adelaide, South Australia, Australia
Abstract: Six tropical freshwater species were used to assess the toxicity of mine waters from a uranium mine adjacent to a
World Heritage area in northern Australia. Key contaminants of potential concern for the mine were U, Mg, Mn, and total
ammonia nitrogen (TAN). Direct toxicity assessments were carried out to assess whether the established site‐specific
guideline values for individual contaminants would be protective with the contaminants occurring as mixtures. Metal
speciation was calculated for contaminants to determine which were the major contributors of toxicity, with 84 to 96% of
Mg predicted in the free‐ion form as Mg
2+
, and 76 to 92% of Mn predicted as Mn
2+
. Uranium, Al, and Cu were predicted to
be strongly bound to fulvic acid. Uranium, Mg, Mn, and Cu were incorporated into concentration addition or independent
action mixture toxicity models to compare the observed toxicity in each of the waters with predicted toxicity. For >90% of
the data, mine‐water toxicity was less than predicted by the concentration addition model. Instances where toxicity
was greater than predicted were accompanied by exceedances of individual metal guideline values in all but one case (i.e., a
Mg concentration within 10% of the guideline value). This indicates that existing individual water quality guideline values for
U, Mg, Mn, and TAN would adequately protect ecosystems downstream of the mine. Environ Toxicol Chem
2021;40:2334–2346. © 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC
Keywords: Metal mixture toxicity; Uranium; Manganese; Magnesium; Copper
INTRODUCTION
Aquatic contaminants generally occur as complex
mixtures in the environment. Despite this, aquatic risk assess-
ments for chemicals are often based on default water quality
guideline values (WQGVs) derived from assessing the
toxicity of individual contaminants to aquatic organisms
(Canadian Council of Ministers of the Environment 2003;
US Environmental Protection Agency 2017; Australian and
New Zealand governments 2018; Proykova et al. 2018). This is
the case with the Ranger uranium mine (whose receiving waters
are in the adjacent United Nations Educational, Scientific and
Cultural Organization's World Heritage‐listed Kakadu National
Park in northern Australia), for which Mg, Mn, U, and total
ammonia nitrogen (TAN) are the contaminants of potential
concern (COPC).
Extensive research during the 40 yr of mine‐site
operation has enabled the derivation of site‐specific WQGVs
for these COPC, which have been incorporated into the
mine‐site's regulatory framework as operational limits (Turner
et al. 2019). These WQGVs aim to provide a high level of
protection (i.e., 99% of species) for the highly valued ecosys-
tems that are downstream of the mine (van Dam et al. 2010;
Harford et al. 2015; van Dam et al. 2017; Mooney et al. 2018).
Water quality guideline values for single contaminants have
provided a high level of confidence that aquatic ecosystems
are protected from the impact of the COPC released during
mine operations. However, freshwater organisms may be ex-
posed to more complex mixtures in the environment following
decommissioning and rehabilitation of the mine. Predicting the
environmental impact of modeled contaminant concentrations
after rehabilitation was a requirement of mine closure plans and
objectives for the site.
This article includes online‐only Supplemental Data.
* Address correspondence to Melanie.trenfield@environment.gov.au
Published online 30 April 2021 in Wiley Online Library
(wileyonlinelibrary.com).
DOI: 10.1002/etc.5103
Mixture toxicity can be predicted based on the knowl-
edge of hazards imposed by individual components, par-
ticularly where direct toxicity assessments (DTAs) of
combined effects are not feasible. Concentration addition
and independent action models, based on the principle of
additivity, are commonly used to predict the toxicity of
mixtures of components with similar or dissimilar modes of
action, respectively (see Supplemental Data, Equations 1
and 2; Barata et al. 2006; Kortenkamp et al. 2009; Nys et al.
2017b). In the natural aquatic environment, metal con-
tamination is usually associated with a mixture of metals
with different modes of action. For U and Mn, the mode of
action appears to be associated with the induction of oxi-
dative stress (Nechay et al. 1980; Ribera et al. 1996; Bare-
scut et al. 2005; Barillet et al. 2007; Periyakaruppan et al.
2007; Conceicao Vieira et al. 2012; Trenfield et al. 2012a).
ForMg,thereisevidencetosuggestthatthemodeof
action involves inhibition of Ca‐dependent processes (Ka-
waii et al. 1999; Gitter et al. 2003; van Dam et al. 2010). For
TAN, postulated effects include an increased energy de-
mand for maintaining ion gradients over the cytoplasmic
membrane (Martinelle and Haggstrom 1993) and sub-
sequent increased oxygen consumption (Smart 1978). Re-
sponses of organisms to metal mixtures can also deviate
from additivity due to interactions among metals that result
in increased (synergistic) or reduced (antagonistic) toxicity
(Norwood et al. 2003; Vijver et al. 2011; Nys et al. 2017a,
2017b; van Regenmortel and De Schamphelaere 2018).
However, it is reported that concentration addition can be
used as a conservative, default first‐tier approach in as-
sessments of mixture toxicity (Kortenkamp et al. 2009).
An important consideration for predicting the toxicity of
contaminant mixtures is the bioavailability or reactivity of the
individual contaminants (van Genderen et al. 2015). The dis-
tribution of a metal among the various physicochemical forms
or species plays an important role in determining the concen-
tration of a metal bound, or complexed, to physiologically
active sites on the cell surface of an organism that results in
uptake and potential toxicity (Markich 2013, 2017a). The con-
centration of the metal‐ligand complex at physiologically active
cell surface sites is generally proportional to the concentration
of the free metal ion (Brown and Markich 2000; Campbell and
Fortin 2013), and is dependent on the reactivity, or binding
affinity, of the free metal ion with (biotic) ligands at the cell
surface. The physicochemical distribution of a metal, including
the concentration of free metal ion, can be predicted for
natural freshwaters using geochemical speciation models
(Santore et al. 2001; Santore et al. 2017; Di Bonito et al. 2018).
The primary characteristics of natural waters that are known
to influence metal speciation and toxicity include dissolved
organic carbon (DOC) concentration, pH, alkalinity, and
hardness (Riethmuller et al. 2001; Gillis et al. 2008; Cheng et al.
2010; Trenfield et al. 2012a, 2012b; Markich 2013; Oliveira‐
Filho et al. 2014). These factors influence metal
speciation either through metal complexation, changes to the
dissociation of a metal, or by competing with metals for
binding sites at cell surfaces. The features of the natural water
used in the present study generally enhance the free metal
concentration, that is, low‐to‐moderate DOC concentration
(Markich et al. 2000; Trenfield et al. 2011, 2012a, 2012b), slight
acidity (Hyne et al. 1992; Franklin et al. 2000; Wilde et al. 2006;
Cheng et al. 2010), very low alkalinity and hardness
(Riethmuller et al. 2001; Charles et al. 2002; Markich 2013), and
low Mg to Ca mass ratios (van Dam et al. 2010). Thus, the
calculation of metal speciation occurring in these mine‐water
mixtures was considered essential to understanding the
primary contaminants contributing to toxicity.
We hypothesized that toxicity data from DTAs could be
used to identify contaminant interactions and, hence, as-
sess the suitability of site‐specificWQGVsforindividual
contaminants in protecting local aquatic biota from ex-
posures to mixtures of these contaminants. The goal of the
present study was to compare DTA toxicity data with site‐
specific WQGVs for individual contaminants of concern. The
main objectives were to: 1) determine the key contaminants
and contributors to toxicity using hazard quotients (HQs)
and speciation modeling, and 2) determine the extent of
contaminant interactions using the concentration addition
or independent action model predictions of the toxicity of
mine waters (particularly at ≤10% effect concentrations
[EC10]). Determining the specific metal interactions occur-
ring in the mine waters was outside the scope of the present
study.
MATERIALS AND METHODS
Test organisms
The 6 tropical freshwater species used throughout
the present study (Table 1) are local to Kakadu National
Park, which surrounds the Ranger uranium mine lease.
Cultures have been maintained at the Environmental Re-
search Institute of the Supervising Scientist for the past
18 yr (with periodic renewal from new field collections when
necessary). Culturing and test method details are described
in Trenfield et al. (2020).
TABLE 1: Suite of local tropical freshwater species used in chronic toxicity tests
Test organism Duration Test endpoint Key reference
Chlorella sp. (green alga) 72 h Cell division rate Pease et al. (2016a)
Lemna aequinoctialis (duckweed) 96 h Growth (surface area) Pease et al. (2016b)
Moinodaphnia macleayi (cladoceran) 5–6 d Reproduction (3 brood) Hyne et al. (1996)
Hydra viridissima (green hydra) 96 h Population growth Riethmuller et al. (2003)
Amerianna cumingi (gastropod) 96 h Reproduction (egg number) Houston et al. (2007)
Mogurnda mogurnda (northern trout gudgeon) 7 d Growth (length) Pease et al. (2021)
Toxicity of contaminant mixtures from a uranium mine—Environmental Toxicology and Chemistry, 2021;40:2334–2346 2335
wileyonlinelibrary.com/ETC © 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC
Test design
Five mine‐water types from the Ranger mine site (Figure 1) were
identified as having representative compositions of mine‐water
contaminants that may egress from the mine site after decom-
missioning (Table 2). The differences among the chemistries of
these site waters and the similarities among multiple collections
from the same site were analyzed using a principal component
analysis (PRIMER Ver 7; Clarke and Gorley 2015; and Supplemental
Data, Figure S1). Details on the 5 mine‐water types follow below.
Process water (low pH). This mine water type is repre-
sentative of a contaminant source of tailings and brines de-
posited in a long‐term storage facility (Pit 3). Water was
collected from the tailings storage facility (site code TDWW)
and not pH adjusted (~pH 3), that is, indicating conditions
where potential metal reactivity would be highest.
Process water (pH adjusted). Water also collected
from TDWW was first diluted to various strengths
FIGURE 1: Sampling locations of site waters from the Ranger uranium mine used in direct toxicity assessments. Sites are marked in yellow.
SIS2 =seepage interception system 2; TDWW =tailings dam wastewater; PJ =Pit 1 decant tower; RP2 =retention pond 2.
2336 Environmental Toxicology and Chemistry, 2021;40:2334–2346—M.A. Trenfield et al.
© 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC wileyonlinelibrary.com/ETC
(treatments) and each treatment water was adjusted to pH 6.
This was considered representative of process water that
may be neutralized as it migrates through groundwater
aquifers.
Tailings porewater. Technically a mix of tailings
porewater, groundwater, and rainwater, its composition
changes seasonally. Characterizing a contaminant source
of tailings deposited in a long‐term storage facility (Pit 1),
this water was collected from a process decant
structure (PJ).
Contaminated surface water. This mine‐water type is in-
dicative of contaminated surface waters that will occur on the
decommissioned site. It was collected from a retention
pond (RP2).
Contaminated shallow groundwater. Water that has
leached from the waste rock of the constructed landform
into the shallow aquifers constitutes what will be the major
source of contaminated water from the decommissioned site. It
was collected from a seepage interception system (SIS2) that
was a shallow groundwater source contaminated by water
flowing through the waste rock wall of the tailings storage
facility.
Two separate toxicity tests were carried out for each species
in each mine‐water type, except for Mogurnda mogurnda in
RP2 water; this was tested only once due to no effect being
observed at 100% strength. As the result of a similar lack of
sensitivity to SIS2 water, 50% effect concentration (EC50) could
not be calculated for exposures with Chlorella sp., Lemna ae-
quinoctialis or M. mogurnda. However, an opportunity arose to
use reverse osmosis for the SIS2 water to concentrate the
groundwater. See Preparation of mine‐water test solutions
section immediately following for more information. This con-
centrated water was then used in a single toxicity test for
Chlorella sp., L. aequinoctialis, and M. mogurnda to generate a
complete concentration–response curve and derive EC50s for
those species.
Preparation of mine‐water test solutions
Glassware or test containers were prepared according to a
washing routine described in Trenfield et al. (2020). Each
toxicity test was performed using a range of dilutions of the
mine water, utilizing Magela Creek water diluent collected
upstream of the mine from Bowerbird Billabong located ap-
proximately 20 km southeast of Jabiru Airport (Figure 2).
Water from this location was used as the diluent (rather than
water from Gulungul Creek located west of the mine) because
this enabled data to be compared with individual contaminant
exposures that were conducted in Magela Creek water.
Treatments were prepared 1 or 2 d before test commence-
ment. After dilution, the test waters were not pH adjusted
except for the series of tests with TDWW adjusted to pH 6
(using 0.05 M NaOH).
TABLE 2: Chemistry of full‐strength site waters and Magela Creek water diluent at the time of collection
a
Contaminants of concern (mg/L)
Water type Site EC (mS) pH
DOC
(mg/L)
Alkalinity (mg/
L CaCO
3
)Hardness
b
(mg/L CaCO
3
)TAN U Mn Mg Mg:Ca ratio
Diluent Magela Creek
c
0.02 ±0.006 6.4 ±0.5 3.1 ±1.5 3.6 ±1.4 3.9 ±1.2 0.02 ±0.02 0.00002 0.002 ±0.001 0.7 ±0.3 4.6:1
Process TDWW 33.5 ±0.1 3.0 ±0 9.5 ±0.1 <1 35 234 ±1016 750 ±50 31.8 ±0.8 2350 ±50 8250 ±250 14:1
TDWW pH 6 31.6 ±0.8 2.9 ±0.1 9.5 ±0.1 <1 33 400 ±1204 734 ±21 26 ±1 2450 ±50 7850 ±250 16:1
Tailings porewater PJ 19.3 ±0.1 5.7 ±0.1 0.5 ±0.2 <1 10 501 ±5971 509 ±9 5.1 ±1.4 700 ±450 2550 ±1450 7:1
Treated pond RP2 2.4 ±0.1 8.4 ±0.2 3.8 ±0.2 110 1375 ±25 0.9 ±0.3 2.08 ±0.08 8.8 ±6.4 334 ±6 4:1
Groundwater SIS2 2.5 ±0.4 6.4 ±0.3 1.1 ±0.1 90 1846 ±372 1.0 ±0.1 0.27 ±0.08 4750 ±150 385 ±75 4:1
a
Median ±range of 2 collections per site water.
b
Hardness by calculation. American Public Health Association 1995: 2.497 ×[Ca] +4.118 ×[Mg].
c
Values represent the median of the control treatments across all tests.
EC =electrical conductivity; DOC =dissolved organic carbon; TAN =total ammonia nitrogen; TDWW =tailings dam wastewater; PJ =process decant structure; RP2 =retention pond 2; SIS2 =seepage interception
system 2.
Toxicity of contaminant mixtures from a uranium mine—Environmental Toxicology and Chemistry, 2021;40:2334–2346 2337
wileyonlinelibrary.com/ETC © 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC
Reverse osmosis concentration of shallow
groundwater
A third collection of SIS2 groundwater was concentrated
using reverse osmosis (Compact L300; RO water Australia) ac-
cording to the recommendations in Serkiz and Perdue (1990).
Approximately 20 L of water were used to generate 5 L of
concentrated water (~170% strength of the original SIS2 water,
based on the increase in Mg and Mn content). The resulting
concentrate was not 4 times the strength of the original water
because a small proportion of the water was lost to waste
streams (rather than recirculated) as part of the reverse osmosis
process. The concentrate was collected in a 10‐L, high‐density
FIGURE 2: Location of diluent collection, Bowerbird Billabong is upstream of the Ranger uranium mine, approximately 20 km southeast of Jabiru
Airport (GDA94 datum expressed as decimal degrees, Latitude:‐12.772069329, Longitude: 133.039681119).
TABLE 3: Measured inorganic composition of site waters at the strength (%) at which the most sensitive EC10 (10% effect concentration) was
observed
a
Water type
Process water Pond water Groundwater
TDWW TDWW pH 6 PJ RP2 SIS2
Chemical parameter Limit of reporting (0.002%) (0.006%) (0.03%) (1%) (0.9%)
pH 0.1 6.4 6.5 6.7 6.6 6.6
DOC (mg/L) 1.0 2.8 1.2 1.9 2.7 2.4
Alkalinity (mg/L CaCO
3
)
b
1.0 3.6 3.6 3.6 3.6 3.6
Ca (mg/L) 0.1 0.08 0.03 0.08 0.05 0.78
K (mg/L) 0.1 0.002 0.008 0.018 0.01 0.16
Mg (mg/L) 0.1 0.44 0.75 2.6 0.17 4
Na (mg/L) 0.1 0.01 0.01 0.01 0.07 0.39
SO
4
(mg/L) 0.1 0.8 2.25 1.14 0.66 13
Cl (mg/L) 0.1 0.0018 0.008 0.018 0.01 0.16
Al (µg/L) 0.1 11.8 4.6 0.27 0.6 0.036
Cu (µg/L) 0.01 0.24 0.37 0.004 0.01 0.006
Mn (µg/L) 0.01 48 136 243 0.41 49
U(µg/L) 0.001 0.6 1.1 1.43 1.56 1.6
Zn (µg/L) 0.1 0.1 0.33 0.05 0.03 0.07
a
Values represent an individual measurement at a given strength of water (values used for metal speciation modeling).
b
Due to the high dilutions at which the EC10 occurred, the median alkalinity of Magela Creek water is listed, rather than the alkalinity of the site waters.
See Table 2 footnote for abbreviation definitions.
2338 Environmental Toxicology and Chemistry, 2021;40:2334–2346—M.A. Trenfield et al.
© 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC wileyonlinelibrary.com/ETC
polyethylene container and stored at 4 °C until testing
commenced.
Physicochemical analyses
Electrical conductivity, dissolved oxygen, and pH were
measured for each treatment at the beginning of each test, and
measured again for both new water and old water whenever test
solutions were exchanged (WTW Multiline Level 1 and Multiline
P4 inolab meter). Incubator temperatures were measured at
5‐min intervals using a wireless monitoring system (Testo Saveris).
At the start of each test, a Milli‐Q blank, procedural blank, the
Magela Creek water control, and the highest strength treatment
of mine water for each test were subsampled, filtered (<0.45 µm),
and acidified (1% 15.9 M HNO
3
;MerckARgrade).Thesamples
wereanalyzedforAl,Cd,Co,Cr,Cu,Fe,Mn,Ni,Pb,Se,U,and
Zn using inductively coupled plasma mass spectroscopy (ICP‐MS;
Agilent 7900) and for Ca, Mg, and S (converted to SO
4
)usingICP‐
atomic emission spectroscopy (AES; Agilent 5110). The remaining
treatments were analyzed for Mg, Mn, U, Ca, and SO
4
only.
Dissolved organic carbon and TAN were measured for each
treatment. Dissolved organic carbon analysis was conducted
using high‐temperature combustion (American Public Health As-
sociation Method 5310B, TOC‐5000A; Shimadzu). Total ammonia
nitrogen was measured using spectrophotometry (US Environ-
mental Protection Agency colorimetric Method 350.1; US Envi-
ronmental Protection Agency 1993). Nitrate (NO
3
–
)and
phosphate (PO
43–
) were measured for algal and lemna control
waters that were subsampled at test commencement and frozen
for analyses. Nitrate was analyzed using the vanadium chloride
method and PO
43–
was analyzed using the ascorbic acid/mo-
lybdate method (American Public Health Association Method
4500‐P E; discrete analyzer). The alkalinity of each site water was
determined using titration (American Public Health Association
Methods 2320A and 2320B).
At the end of each test, the blanks and the Magela Creek
water control (subsampling old water from the control replicate
containers) were analyzed for Al, Cd, Co, Cr, Cu, Fe, Mn, Ni, Pb,
Se, U, and Zn using ICP‐MS (Agilent 7900) and for Ca, Mg, and S
(converted to SO
4
)usingICP‐AES (Agilent 5110). All remaining
treatments were analyzed (using old water sampled from the
replicate containers) for Mg, Mn, U, Ca, and SO
4
only. Total
ammonia nitrogen was also measured for each treatment at the
end of each test. The start and end of test concentrations of Mg,
Mn, and U and TAN were then averaged for each treatment and
the mean value used in any subsequent analyses.
The limit of reporting has been listed for each analyte in
Table 3 and results were considered acceptable if the relative
percent difference of duplicate samples was <20%, the spike
recoveries were from 80 to 120%, and the value for reference
standards was between 95 and 105% of the certified value.
Deriving effect concentrations
For each organism and mine water, the response of the
replicates for each treatment was averaged, and the mean,
expressed as a function (%) of the control response, was then
plotted against the percentage strength of mine water
(Sigmaplot Ver 14.0). Nonlinear regressions (with the model
selected for goodness‐of‐fit based on the lowest standard
error [SE] of estimate), were then fitted to generate
concentration–response curves. Effect concentrations (EC10s
and EC50s) and their 95% confidence limits were derived from
the concentration–response curves.
Determining the degree of hazard for
contaminants
Hazard quotients were calculated to identify those con-
taminants likely to be contributing to the toxicity of a mine
water (Lemly 1996; Kortenkamp et al. 2009). Hazard quotients
were calculated by comparing the contaminant concentrations
in the mixture against their respective WQGVs. The HQs were
calculated for diluted mine waters (the dilution at which an
EC10 was observed) rather than undiluted site waters because
of the high unlikelihood that mine waters would be released
undiluted. In addition to the 4 original COPC, HQs were also
calculated for Cu, Al, and Zn because these metals were
measured at elevated concentrations in some of the mine
waters.
Predicting metal speciation
Metal speciation modeling was based on the chemistry of
each mine water (measured dissolved concentrations; Table 3) at
the strength at which an EC10 was observed using the Wind-
emere Humic Aqueous Model (WHAM; Ver 7.0.5) geochemical
speciation code. Speciation modeling incorporated Al, Cu, Mg,
Mn, U (as U[VI]), and Zn because these were the contaminants
for which HQs were calculated. The inorganic equilibrium con-
stants (aqueous +minerals) used for all speciation modeling
were derived from Markich (2017b), whereas metal binding with
natural dissolved organic matter (DOM) was calculated using the
Humic Ion‐Binding Model VII (Tipping et al. 2011)—asubmodel
of the WHAM. It was assumed that the ratio of active DOM:DOC
was 1:4 and that 100% of the active DOM was fulvic acid. This is
based on DOM containing 50% carbon (by mass) with a 70%
fulvic acid composition (Markich 2017b). The effects of ionic
strength on the inorganic reactions were accounted for using the
extended Debye–Hückel equation. The WHAM speciation
model also accounts for any minerals that may precipitate and/or
any metal binding/adsorption to Fe and/or Al (oxy)hydroxides, if
applicable.
Assessment of the concentration addition and
independent action models
In applying the concentration addition model to each DTA,
toxic units were calculated for U, Mn, Mg, and Cu by dividing
the individual dissolved metal concentration at the observed
EC50 in the mine water by the expected EC50 for that metal
(based on existing toxicity data for the individual contaminants;
Toxicity of contaminant mixtures from a uranium mine—Environmental Toxicology and Chemistry, 2021;40:2334–2346 2339
wileyonlinelibrary.com/ETC © 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC
Supplemental Data, Figures S2–S5). Before the present study,
Cu had not been considered a COPC for the mine. However,
Cu was incorporated into the model due to its elevated con-
centration in process water and the subsequent degree of
hazard it was calculated to present. Copper toxicity data were
subsequently generated for 6 local species in Magela Creek
water to derive expected EC50s. Total ammonia nitrogen was
excluded from the model because it was at background con-
centrations in TDWW and SIS2 water (at the strengths at which
an EC10 was observed). Hence, TAN was considered low risk.
Aluminium and Zn were not included in the models because
there were no individual metal toxicity data for these species.
Single contaminant EC50s for Mn and Mg were sourced from
Harford et al. (2015) and van Dam et al. (2010), respectively, for
all species except M. mogurnda (because the toxicity data for
the latter species were acute). The EC50s for Mg and Mn ex-
posure to M. mogurnda were sourced from chronic toxicity
data published by Pease et al. (2021). To derive an expected
EC50 for U and each species, toxicity data were sourced from
van Dam et al. (2012) for Moinodaphnia macleayi; Trenfield
et al. (2011) for Hydra viridissima; Hogan et al. (2010) for L.
aequinoctialis; Mooney et al. (2016) for Amerianna cumingi;
Pease et al. (2021) for M. mogurnda; and unpublished data (M.
Trenfield, unpublished data) for Chlorella sp. Although there
were published data for Chlorella sp., those values had been
derived using higher nutrient concentrations and starting cell
densities than those in the present study (this is described by
Trenfield et al. [2020]). The sum of toxic units for all metals was
then calculated for each site water to determine whether con-
centration addition applied. The dissolved EC50 values for Mn,
Mg, Cu, and U were corrected for the free metal ion concen-
tration and the first hydroxy complex (e.g., UO
22+
+UO
2
OH
+
),
as a proxy of metal reactivity (or lability) at physiologically ac-
tive sites at the organism cell surface (Brown and Markich,
2000; Trenfield et al. 2011) to assess whether any deviation
from the concentration addition model may be due to differ-
ences in metal reactivity between the testing conditions of the
present study and those of the individual metal testing.
The independent action model was also applied to the site
waters at the strength at which an EC50 was observed, to de-
termine whether concentration addition or independent action
was more accurate at predicting toxicity of the mine waters.
Predicted effects at each dissolved metal concentration
for each species were extrapolated from Supplemental Data,
Figures S2–S5.
The concentration–response plots for individual metals and
their dissolved and chemically reactive fractions (Supplemental
Data, Figures S2–S5) were also used to produce plots of
predicted mixture toxicity versus observed mixture toxicity
(applying the concentration addition model), both in terms of
the dissolved and free metal ion concentration. Effects of
U, Mn, Mg, and Cu were summed for each species at each
direct toxicity assessment dilution. It should be noted that the
concentration addition and independent action models do not
provide a measure of error, and thus deviations from the model
could potentially be a result of stochastic variation in the bio-
logical toxicity tests.
A more complex approach involving multiple linear
regressions was not used because this would have required a
much larger dataset to predict the interaction of the 4 COPC
and the important water quality factors that influence their
bioavailability. The use of a biotic ligand model (BLM) was also
not possible because models for Mg or U are unavailable
(Mebane et al. 2020; Schlekat et al. 2020) and current BLMs do
not predict the effect of COPC mixtures.
RESULTS AND DISCUSSION
Quality control
There were no instances of confounding contamination
across any of the tests based on metal concentrations
measured in the blank and procedural blank samples (data not
shown).
The chemistry of the 2 water collections for each site (3 for
SIS2 water) was similar (Supplemental Data, Figure S1 and
Table 2). Principal component analysis (Supplemental Data,
Figure S1) shows that most of the variation in chemistry among
water types (81%) was accounted for in principal component
axis 1. This axis represented a contaminant concentration
gradient (from left to right) of most‐to least‐contaminated mine
water, in the order of process water, tailings porewater, con-
taminated surface water, and shallow groundwater. The phys-
icochemical variables for each of the waters across all dilutions
are shown in Supplemental Data, Table S1. Due to the broad
range of dilutions tested for each mine water, spanning a large
range of pH and electrical conductivity, minimum and max-
imum values for each of these parameters (rather than a mean
value) have been provided. The range shown for new and old
water demonstrated a minimal shift in pH and electrical con-
ductivity between water renewal (or in the case of Chlorella sp.
and L. aequinoctialis over the entire test duration). In most
cases, pH remained within 0.5 of a unit of the starting pH but in
all cases the shift was ≤1 pH unit. The electrical conductivity
tended to remain within 20% of the starting electrical con-
ductivity; however, for certain species such as A. cumingi and
M. macleayi in the control water of low electrical conductivity,
there were instances where electrical conductivity would almost
double (e.g., increase from 15–27 µS/cm). This was expected,
given that these tests involve the addition of food that remains
in the test solution for 24 h. Dissolved oxygen generally re-
mained between 80 and 115% and temperature ±1°Cofthe
optimal (27 °C for H. viridissima,M. macleayi, and M.
mogurnda; 29 °C for Chlorella sp. and L. aequinoctialis; and
30 °C for A. cumingi). For Chlorella sp. and L. aequinoctialis,
measured nutrient concentrations were near nominal concen-
trations of 0.82 mg/L N and 0.012 mg/L P, and 0.23 mg/L N and
0.03 mg/L P, respectively.
Hazard scoring for contaminants in mine waters
Hazard quotients for the contaminants in the site waters are
shown in Table 4. Calculation of HQs helped to identify the
contaminants contributing to toxicity by providing a system for
2340 Environmental Toxicology and Chemistry, 2021;40:2334–2346—M.A. Trenfield et al.
© 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC wileyonlinelibrary.com/ETC
ranking contaminants for each mine water. In this case, it was
not useful to compare HQs with an arbitrary scale to determine
risk, such as that described by Lemly (1996), because the data
did not conform well to the principle of additivity.
Manganese was calculated to have the highest HQ for process
water TDWW (0.64) and tailings porewater PJ (3.2). Copper and
Al had the next highest HQs for TDWW. For the groundwater
SIS2, Mg was calculated to have the highest HQ (1.3), followed by
Mn and U. For the contaminated surface water (RP2), U was
considered the major contributor to toxicity. The HQs for TAN
(0.05–0.6) were among the lowest for each mine water.
Focus was then directed toward TDWW and SIS2 mine
waters to study mixture interactivity because the HQs indicated
that these waters could have multiple contaminants con-
tributing to their toxicity.
Metal speciation
The following pattern of decreasing percentage of free
metal ion, Mg
2+
>Mn
2+
>Zn
2+
>Cu
2+
>UO
22+
>Al
3+
, was
generally found across all mine waters (Table 5). Magnesium
and Mn had the highest percentage of free ion available in
solution and the lowest complexation to inorganic ligands (e.g.,
sulfate or carbonate) or natural DOM (using fulvic acid as a
proxy), whereas U and Al (<0.45 µm fraction) were typically
complexed with DOM and thus had the lowest percentage of
free ion available for binding at the cell surface. Specifically, for
TDWW, 90, 91, and 97% of Al, Cu, and U were predicted to be
complexed with natural DOM, respectively, whereas 87 and
82% of Mg and Mn were calculated to occur as Mg
2+
and
Mn
2+
, respectively. Colloidal forms of Al such as amorphous
aluminium hydroxide (Table 5) may also elicit adverse effects to
organisms via indirect pathways (Gillmore et al. 2016).
Although 52% of Zn was predicted to be present as Zn
2+
,it
was not considered to be a hazard based on the low dissolved
fraction of Zn (0.08 µg/L) in TDWW of this strength, and
hence the low HQ (Table 4). For TDWW adjusted to pH 6
(post‐dilution), a lower proportion of metals was predicted to
be complexed by natural DOM compared with the unadjusted
TDWW (Table 5); however, this is likely to be due to the lower
DOM concentrations for those tests, that is, 1.2 mg/L compared
with 2.8 mg/L rather than the slight difference in pH between
these 2 water types (Table 3). Apart from that, the similarity in
the trend across the other waters was likely caused by the
similarity in solution physicochemistry at the high dilutions
where the EC10s were observed (i.e., except for the elevated
COPC, the composition of the site waters resembled that of the
natural Magela Creek water diluent; Table 3). These high
dilutions are considered representative of what could poten-
tially occur with the egress of contaminated shallow ground-
water to receiving waters during times of high rainfall in the wet
season, after the site has been rehabilitated. Thus, whereas
undiluted TDWW was pH 3, the pH (6.4) at which an EC10 was
observed and the subsequent metal speciation at that pH were
considered representative of the metal speciation that would
be present in the natural environment.
Comparing observed toxicity with predicted
toxicity
Overall predictions of mixture toxicity for each mine water at
the EC50 were similar between the concentration addition and
TABLE 4: Hazard quotients (HQs) for contaminants in site waters at the strength at which a 10% effect concentration (EC10) was observed for the
most sensitive species
Contaminants of potential concern
Mn (µg/L) Mg (mg/L) U (µg/L) TAN (mg/L) Al (µg/L) Cu (µg/L) Zn (µg/L)
Guideline value
a
75 3 2.8 0.4 27 0.5 1.5
TDWW ———————
EC10
b
48 0.44 0.6 0.02 11.8 0.24 0.1
HQ
c
0.64 0.15 0.21 0.05 0.43 0.48 0.06
TDWW (adj pH) ———————
EC10
b
136 0.75 1.1 0.05 4.6 0.4 0.3
HQ
c
1.8 0.25 0.4 0.13 0.2 0.8 0.2
SIS2 ———————
EC10
b
49 4 1.6 0.01 0.036 0.006 0.07
HQ
c
0.65 1.3 0.57 0.025 0.001 0.01 0.04
PJ ———————
EC10
b
243 2.6 1.4 0.25 0.27 0.004 0.05
HQ
c
3.2 0.9 0.5 0.6 0.01 0.008 0.03
RP2 ———————
EC10
b
0.4 0.17 1.6 0.001 0.6 0.01 0.03
HQ
c
0.005 0.06 0.57 0.003 0.02 0.02 0.02
a
99% Protection guideline values sourced from Supervising Scientist 2018a, 2018b, 2018c; Australian and New Zealand governments 2018; and Supervising Scientist
unpublished data.
b
Expected waterborne concentration of each contaminant at the strength at which the most sensitive EC10 was observed (either Amerianna cumingi or Hydra
viridissima).
c
Hazard quotient is expected concentration of contaminant ÷ guideline value or allowed concentration, Lemly 1996.
See Table 2 footnote for abbreviation definitions.
Toxicity of contaminant mixtures from a uranium mine—Environmental Toxicology and Chemistry, 2021;40:2334–2346 2341
wileyonlinelibrary.com/ETC © 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC
independent action models (Supplemental Data, Tables S2 and
S3, respectively). In most cases, the mixtures were less toxic
than predicted; this suggests that antagonism was occurring
among contaminants. This was also true for the predictions in
terms of the free‐metal ion fraction (Supplemental Data,
Table S4). With Mg and Mn predicted to be the 2 main
chemically reactive contaminants in these waters (i.e., high
levels of the free ion), it is possible that interaction between
these metals was a major contributor to the antagonism ob-
served, with amelioration of Mn toxicity by Mg potentially oc-
curring (as reported by Peters et al. 2011). For these mine
waters, which are dominated by Mg, it is likely that this would
often be the case; that is, these models (which do not account
for metal–metal interaction) will overestimate toxicity due to
the divalent cations in the mixtures acting as competitors in
their contribution to toxicity (Peters et al. 2011).
Based on the results for these waters, neither the concen-
tration addition nor independent action model would be suit-
able in making accurate predictions of mixture toxicity.
Nevertheless, these results showed the environmental risk
management of these waters; the assumption of additivity, in
most cases, produced a conservative prediction of toxicity. The
application of the concentration addition model at the EC10
(Supplemental Data, Table S5) demonstrated a greater number
of underestimations of toxicity compared with that at the EC50
(Figure 3). However, caution should be applied in interpreting
these results at the EC10 because there would be greater
variability in the sensitivity of the test organisms at this effect
concentration than at the EC50, and more difficulty in making
an accurate comparison between the responses from ex-
posures to individual contaminants with those from exposures
to mixtures.
In applying the concentration addition model (based on
dissolved concentrations of Mg, Mn, U, and Cu) to all process
water (TDWW) and groundwater (SIS2) data, toxicity was over‐
predicted for >90% of the data (Figures 4 and 5). Applying the
model using the free‐ion concentrations of the metals was
important to account for any difference in reactivity between
TABLE 5: Predicted speciation of Al, Cu, Mg, Mn, U, and Zn in site
waters at concentrations (Table 2) at which the most sensitive EC10 was
observed
Metal species
a
TDWW TDWW pH 6 PJ RP2 SIS2
Aluminium
Al(OH)
3
—1.7 1.5 —1.0
Al(OH)
4
–
1.2 3.6 5.0 1.8 2.5
AlFA 90.0 35.8 49.7 97.3 71.9
Al(OH)
3(am)
b
8.0 59.0 42.9 —24.4
Copper
Cu
2+
8.0 13.3 5.6 3.4 6.5
CuOH
+
—1.7 1.1 —1.0
CuSO
4
— — ———
CuFA 90.9 84.5 92.9 95.9 92.0
Magnesium
Mg
2+
87.4 95.0 95.5 83.5 94.7
MgSO
4
————1.8
MgFA 12.4 4.7 4.0 16.5 3.4
Manganese
Mn
2+
82.4 92.4 92.2 76.1 91.6
MnSO
4
————1.9
Mn(SO
4
)
2
2–
— — ———
MnFA 17.3 7.1 7.3 23.7 6.4
Uranium
UO
2
OH +1.2 2.3 1.2 —1.1
UO
2
(OH)
2
—1.0 ———
UO
2
OHSO
4
–
— — ———
UO
2
CO
3
1.8 3.8 2.3 1.2 2.0
UO
2
FA 96.5 92.7 95.6 97.7 96.1
Zinc
Zn
2+
51.6 70.0 56.9 36.0 76.1
ZnSO
4
————1.6
Zn(SO
4
)
2
2–
— — ———
ZnFA 47.9 29.2 42.4 63.6 22.1
a
Each metal species is shown as a percentage of its measured dissolved con-
centration. Species comprising <1% of the total distribution are excluded for
clarity. FA =fulvic acid—a proxy for natural dissolved organic matter.
b
Colloidal amorphous aluminium hydroxide.
See Table 2 footnote for abbreviation definitions.
(A)
(B)
FIGURE 3: Sum (∑) of toxic units across mine waters for each
species. (A) at EC50 and (B) at EC10. Mean ±range is shown.
Dashed lines represent where concentration addition applies (∑=1).
Values above the lines indicate where mixtures were less toxic than
predicted. Values below the lines identify where mixtures were more
toxic than predicted. n=5 mine waters for snail, hydra, and cladoceran;
n=3 mine waters for duckweed and fish (2 values >20 were excluded
for clarity). EC50 =50% effect concentration; EC10 =10% effect
concentration.
2342 Environmental Toxicology and Chemistry, 2021;40:2334–2346—M.A. Trenfield et al.
© 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC wileyonlinelibrary.com/ETC
the individual metal testing and the DTAs. This approach did
not appreciably alter the interpretation of mixture effects for
the 2 sites (Figures 4B and 5B) but, in many cases, it did in-
crease the extent of antagonism. This may be partly due to the
differences in DOC concentrations of the Magela Creek water
diluent used for the individual contaminant testing and DTAs
(which were accounted for in the speciation modeling).
The median DOC of the Magela Creek water diluent for the
DTAs was 3.1 mg/L compared with the following DOC range
for the individual metal testing: Cu 2.2 mg/L, Mn 2.3 mg/L,
Mg 4.0 mg/L, and U 1.7 to 5 mg/L.
Instances where synergism occurred at low‐effect
concentrations (the red‐shaded areas of Figures 4 and 5),
represent potential underestimation of toxicity at ≤EC10s. This
underestimation of toxicity, particularly for the cladoceran M.
macleayi, was in some cases >10%. Although these data points
are likely to represent stochastic variation in the biological
toxicity tests, they highlight that there may be instances at low‐
effect concentrations where these species respond more sen-
sitively than expected using a concentration addition model.
However, for all (except one) data points in this area of the
plots, at least one of the COPC was in exceedance of its
guideline value (Supplemental Data, Table S6), indicating that
application of the individual guideline values would still be
protective. For the single data point where there was no ex-
ceedance, the Mg concentration was within 10% of its guide-
line value (Supplemental Data, Table S6). In the case of the
Ranger mine, the use of the individual COPC guideline values
is likely to be protective; however, uncertainty in these
predictions for the most sensitive species could be addressed
with future DTAs that monitor the toxicity of mine‐water egress
from the rehabilitated site. The species that showed greater
sensitivity (M. macleayi,A. cumingi, and Chlorella sp.) should
be prioritized.
(A)
(B)
FIGURE 4: Predicted mixture effect (%) versus observed mixture effect
(%) of TDWW (tailings dam wastewater) based on (A) dissolved metal
concentrations, and (B) free‐ion concentrations. The concentration
addition model was used to predict toxicity based on Mg, Mn, U, and
Cu concentrations. Black lines: observed effects =predicted effects.
Data above black lines: observed toxicity <expected toxicity. Data
below black lines: observed toxicity >expected toxicity. Red shading:
predicted effect is <10% but observed effect is >10%. Blue shading:
predicted effect is <10% but observed effect is <–10% (indicating a
stimulatory effect).
(A)
(B)
FIGURE 5: Predicted mixture effect (%) versus observed mixture effect
(%) of groundwater (SIS2) based on (A) dissolved metal concentrations,
and (B) free‐ion concentrations. Predicted toxicity was calculated using
the concentration addition mixture model incorporating the effect of
Mg, Mn, U, and Cu. SIS2 =seepage interception system 2.
Toxicity of contaminant mixtures from a uranium mine—Environmental Toxicology and Chemistry, 2021;40:2334–2346 2343
wileyonlinelibrary.com/ETC © 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC
Where comparisons were made between observed toxicity
and predicted toxicity based on the primary contaminant only
(Supplemental Data, Figure S6), in TDWW (Mn) and SIS2 (Mg)
a greater proportion of data appeared below the 1:1 line,
compared with the respective plots in Figures 4 and 5 where
the waters were assessed as a mixture. This demonstrated that
for both waters, although Mg and Mn were considered the
primary contributors to toxicity based on their HQs, the toxicity
of these waters would be assessed more conservatively as a
mixture.
Validation of guideline values and comparison
with field data
The general trend of antagonism observed among con-
taminants for the site waters suggests that the local species
used in the present study would be protected if exposed to a
mixture of the COPC at their individual guideline value con-
centrations.
Other supportive evidence for the protectiveness of the
guideline values is suggested through a historical DTA of a
tributary near Ranger mine conducted in 2015 after a high
electrical conductivity event (Trenfield et al. 2017). That study
provided supportive evidence for the protectiveness of the
Mg and Mn guideline values for 3 species exposed to a
mixture that included elevated Mg and Mn concentrations
(350 mg/L and 350 µg/L, respectively). However, U concen-
trations (0.7 µg/L U) were not high enough to test the validity of
the site‐specific guideline value for U (2.8 µg/L).
CONCLUSIONS
The present study has shown the value in assessing the
toxicity of aqueous mixtures of contaminants to determine if
individual site‐specific guideline values provide enough pro-
tection for the local aquatic environment. Through the com-
bined assessment of HQs and metal speciation, Mn was found
to be the major contributor to toxicity for the process waters
and Mg for groundwater. However, for both waters, assess-
ment of toxicity in terms of the combined effect of con-
taminants was warranted. The calculation of HQs was useful in
identifying major contributors to toxicity. Both the concen-
tration addition and independent action models (incorporating
the combined effect of Mg, Mn, U, and Cu) were, in most cases,
a conservative approach to predicting mixture toxicity. Con-
sideration of the chemical reactivity of contaminants enabled
more accurate comparison of the observed toxicity of the site
waters with their predicted toxicity. Over‐prediction by the
models would include unquantifiable error but, despite this,
indicates a general pattern of antagonistic interactions among
contaminants. The low occurrence (<4%) of under‐prediction of
toxicity at low‐effect concentrations was generally associated
with contaminant concentrations that exceeded a site‐specific
guideline value for at least one contaminant. Hence, if adhered
to, the present site‐specific guideline values should be pro-
tective in most cases for these organisms, under the conditions
tested in the present study. Copper was identified as a major
contaminant in process water, highlighting the need for a site‐
specific guideline value to be developed for this metal. Further
investigation into how Mg and Mn interact in extremely soft
waters at both EC10 and EC50 for a range of species is
warranted.
Supplemental Data—The Supplemental Data are available on
the Wiley Online Library at https://doi.org/10.1002/etc.5103.
Acknowledgment—The authors thank Environmental staff at
ERA Ranger Mine for assistance with collection of site waters
and A. Holland (La Trobe University) for the use of the reverse
osmosis concentration unit. J. Stauber also provided valuable
guidance on the design and interpretation of the present study,
and D. Parry generously reviewed the manuscript.
Author Contributions Statement—M. Trenfield, A. Harford,
and R. van Dam conceived and designed the experiments. M.
Trenfield, C. Pease, and S. Walker conducted experiments and
chemical analyses. M. Trenfield performed statistical analyses.
M. Trenfield and S. Markich performed speciation modeling.
M. Trenfield wrote the manuscript with guidance from S.
Markich, A. Harford, R. van Dam, and C. Humphrey.
Data Availability Statement—Data, associated metadata, and
calculation tools are available from the corresponding author
(Melanie.trenfield@environment.gov.au).
REFERENCES
American Public Health Association. 1995. Standard Methods for the
Examination of Water and Wastewater, 19th ed. New York, NY, USA.
Australian and New Zealand governments. 2018. Australian and New
Zealand guidelines for fresh and marine water quality. Australian state
and territory governments, Canberra, Australian Capital Territory, Aus-
tralia. [cited 2021 February 22]. Available from: http://www.waterquality.
gov.au/anz‐guidelines/guideline‐values/default/water‐quality‐toxicants/
toxicants/copper‐2000
Barata C, Baird DJ, Nogueira AJA, Soares AMVM, Riva MC. 2006. Toxicity
of binary mixtures of metals and pyrethroid insecticides to Daphnia
magna Straus. Implications for multi‐substance risks assessment. Aquat
Toxicol 78:1–14.
Barescut JC, Gariel JC, Péres JM, Buet A, Barillet S, Camilleri V. 2005.
Changes in oxidative stress parameters in fish as response to direct
uranium exposure. Radioprotection 40:151–155.
Barillet S, Adam C, Palluel O, Devaux A. 2007. Bioaccumulation, oxidative
stress, and neurotoxicity in Danio rerio exposed to different isotopic
compositions of uranium. Environ Toxicol Chem 26:497–505.
Brown PL, Markich SJ. 2000. Evaluation of the free ion activity model of
metal‐organism interaction: Extension of the conceptual model. Aquat
Toxicol 51:177–194.
Campbell PGC, Fortin C. 2013. Biotic ligand model. In Férard JF, Blaise C, eds,
Encyclopaedia of Aquatic Ecotoxicology. Springer, Dordrecht, pp 237–245.
Canadian Council of Ministers of the Environment. 2003. Guidance on the
site‐specific application of water quality guidelines in Canada:
Procedures for deriving numerical water quality objectives. [cited 2021
February 22]. Available from: http://ceqg‐rcqe.ccme.ca/download/en/221
Charles AL, Markich SJ, Stauber JL, De Filippis LF. 2002. The effect of water
hardness on the toxicity of uranium to a tropical freshwater alga
(Chlorella sp.). Aquat Toxicol 60:61–73.
2344 Environmental Toxicology and Chemistry, 2021;40:2334–2346—M.A. Trenfield et al.
© 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC wileyonlinelibrary.com/ETC
Cheng KL, Hogan AC, Parry DL, Markich SJ, Harford AJ, van Dam RA. 2010.
Uranium toxicity and speciation during chronic exposure to the tropical
freshwater fish, Mogurnda mogurnda.Chemosphere 79:547‐554.
Clarke KR, Gorley RN. 2015. PRIMER Ver 7: User Manual/Tutorial. PRIMER‐E,
Plymouth, UK.
Conceicao Vieira M, Torronteras R, Cordoba F, Canalejo A. 2012. Acute
toxicity of manganese in goldfish Carassius auratus is associated with
oxidative stress and organ specific antioxidant responses. Ecotox
Environ Safe 78:212–217.
Di Bonito M, Lofts S, Groenenberg JE. 2018. Models of geochemical spe-
ciation: Structure and applications. In DeVito B, Belkin H, Lima A, eds,
Environmental Geochemistry: Site Characterization, Data Analysis and
Case Histories, 2nd ed. Elsevier, Amsterdam, The Netherlands,
pp 237–306.
Franklin NM, Stauber JL, Markich SJ, Lim RP. 2000. pH dependent toxicity
of copper and uranium to a tropical freshwater alga (Chlorella sp.).
Aquat Toxicol 48:275–289.
Gillis PL, Mitchell RJ, Schwalb AN, McNichols KA, Mackie GL, Wood CM,
Ackerman JD. 2008. Sensitivity of the glochidia (larvae) of freshwater
mussels to copper: Assessing the effect of water hardness and dissolved
organic carbon on the sensitivity of endangered species. Aquat Toxicol
88:137–145.
Gillmore ML, Golding LA, Angel BM, Adams MS, Jolley DF. 2016. Toxicity
of dissolved and precipitated aluminium to marine diatoms. Aquat
Toxicol 174:82–91.
Gitter AH, Oliver D, Thurm U. 2003. Calcium‐and voltage‐dependence
of nematocyst discharge in Hydra vulgaris.J Comp Physiol A 175:
115–122.
Harford A, Mooney T, Trenfield M, van Dam R. 2015. Manganese toxicity to
tropical freshwater species in low hardness water. Environ Sci Technol
34:2856–2863.
Hogan AC, van Dam RA, Houston MA, Harford AJ, Nou S. 2010. Uranium
exposure to the tropical duckweed Lemna aequinoctialis and pulmonate
snail Amerianna cumingi: Fate and Toxicity. Arch Environ Contam Tox-
icol 59:204–215.
Houston MA, Hogan AC, van Dam RA, Nou S. 2007. Procedure for the 96
hour gastropod reproduction toxicity test using Amerianna cumingi.
Internal Report. Supervising Scientist Branch, Darwin, Northern Territory,
Australia.
Hyne RV, Rippon GD, Ellender G. 1992. pH‐dependent uranium toxicity to
freshwater hydra. Sci Total Environ 125:159–173.
Hyne RV, Rippon GD, Hunt SM, Brown GH. 1996. Procedures for the bio-
logical toxicity testing of mine waste waters using freshwater organisms.
Report. Supervising Scientist Branch, Canberra, Australia.
Kawaii S, Yamashita K, Nakai M, Takahashi M, Fusetani N. 1999. Calcium
dependence of settlement and nematocyst discharge in actinulae of the
hydroid Tubularia mesembryanthemum.Biol Bull 196:45–51.
Kortenkamp A, Backhaus T, Faust M. 2009. State of the art report on mix-
ture toxicity. Final report. The School of Pharmacy, University of London,
London, UK.
Lemly AD. 1996. Evaluation of the hazard quotient method for risk assess-
ment of selenium. Ecotox Environ Safe 35:156–162.
Markich SJ. 2013. Water hardness reduces the accumulation and toxicity of
uranium in a freshwater macrophyte (Ceratophyllum demersum). Sci
Total Environ 443:582–589.
Markich SJ. 2017a. Sensitivity of the glochidia (larvae) of freshwater mussels
(Bivalvia: Unionida: Hyriidae) to cadmium, cobalt, copper, lead, nickel
and zinc: Differences between metals, species and exposure time. Sci
Total Environ 601–602:1427–1436.
Markich SJ. 2017b. Thermochemical Data (log K) for Environmentally Rel-
evant Elements, 3rd ed. ASI C1/17. Aquatic Solutions International,
Sydney, Australia.
Markich SJ, Brown PL, Jeffree RA, Lim RP. 2000. Valve movement
responses of Velesunio angasi (Bivalvia: Hyriidae) to manganese and
uranium: An exception to the free ion activity model. Aquat Toxicol
51:155–175.
Martinelle K, Haggstrom L. 1993. Mechanisms of ammonia and ammonium
ion toxicity in animal cells: Transport across cell membranes. J Bio-
technol 30:339–350.
Mebane CA, Chowdhury MJ, De Schamphelaere KA, Lofts S, Paquin PR,
Santore RC, Wood CM. 2020. Metal bioavailability models: Current
status, lessons learned, considerations for regulatory use, and the path
forward. Environ Toxicol Chem 39:60–84.
Mooney TJ, Harford AJ, Trenfield MA, Pease CJ, Hogan AC, van Dam RA.
2016. Increasing uranium exposure durations to the aquatic snail,
Amerianna cumingi, does not result in lower toxicity estimates. Environ
Toxicol Chem 35:2851–2858.
Mooney TJ, Pease C, Trenfield M, van Dam R, Harford AJ. 2018.
Modeling the pH–ammonia toxicity relationship for Hydra viridissima in
soft waters with low ionic concentrations. Environ Toxicol Chem
37:1189–1196.
Nechay BR, Thompson JD, Saunders JP. 1980. Inhibition by uranyl nitrate of
adenosine triphosphatases derived from animal and human tissues.
Toxicol Appl Pharmacol 53:410–419.
Norwood W, Borgmann U, Dixon D, Wallace A. 2003. Effects of metal
mixtures on aquatic biota: A review of observations and methods. Hum
Ecol Risk Assess 9:795–811.
Nys C, Janssen CR, De Schamphelaere KAC. 2017a. Development and
validation of a metal mixture bioavailability model (MMBM) to predict
chronic toxicity of Ni‐Zn‐Pb mixtures to Ceriodaphnia dubia.Environ
Pollut 220:1271–1281.
Nys C, Versieren L, Cordery KI, Blust R, Smolders E, De Schamphelaere
KAC. 2017b. Systematic evaluation of chronic metal‐mixture toxicity to
three species and implications for risk assessment. Environ Sci Technol
51:4615–4623.
Oliveira‐Filho EC, Caixeta NR, Simplicio NCS, Sousa SR, Aragão TP, Muniz
DHF. 2014. Implications of water hardness in ecotoxicological assess-
ments for water quality regulatory purposes: A case study with the
aquatic snail Biomphalaria glabrata (Say, 1818). Braz J Biol 74:175–180.
Pease C, Mooney T, Trenfield M, Costello C, Harford A. 2016a. Updated
procedure for the 72 hour algal growth inhibition toxicity test using
Chlorella sp. Internal Report 645. Supervising Scientist Branch, Darwin,
Northern Territory, Australia.
Pease C, Trenfield M, Cheng K, Harford A, Hogan A, Costello C, Mooney T,
van Dam R. 2016b. Refinement of the reference toxicity test protocol for
the tropical duckweed Lemna aequinoctialis. Internal Report 644. Su-
pervising Scientist Branch, Darwin, Northern Territory, Australia.
Pease CJ, Walker S, Tanneberger C, Mooney TJ, Trenfield MA, van Dam RA,
Harford AJ. 2021. Development of a sub‐lethal chronic toxicity test for
the Northern Trout Gudgeon, Mogurnda mogurnda Environ Toxicol
Chem, and application to uranium, magnesium and manganese., in
press. https://doi.org/10.1002/etc.5005
Periyakaruppan A, Kumar F, Sarkar S, Sharma CS, Ramesh GT. 2007.
Uranium induces oxidative stress in lung epithelial cells. Arch Toxicol
81:389–395.
Peters A, Lofts S, Merrington G, Brown B, Stubblefield W, Harlow K.
2011. Development of biotic ligand models for chronic manganese
toxicity to fish, invertebrates, and algae. Environ Toxicol Chem
30:2407–2415.
Proykova A, Kraetke R, Bertollini R, Borges T, Duarte‐Davidson R,
Panagiotakos D, Slama R, Scott M, Testai E, Vermeire T, Vighi M. 2018.
Memorandum on weight of evidence and uncertainties. Revision
2018. Technical Report. European Commission, Luxembourg City,
Luxembourg.
Ribera D, Labrot F, Tisnerat G, Narbonne JF. 1996. Uranium in the envi-
ronment: Occurrence, transfer, and biological effects. In Ware GW, ed,
Reviews of Environmental Contamination and Toxicology. Springer,
New York, NY, USA, pp 53–89.
Riethmuller N, Camilleri C, Franklin N, Hogan A, King A, Koch A, Markich SJ,
Turley C, van Dam R. 2003. Ecotoxicological testing protocols for Aus-
tralian tropical freshwater ecosystems. Report 173. Supervising Scientist
Branch, Darwin, Northern Territory, Australia.
Riethmuller N, Markich SJ, van Dam RA, Parry D. 2001. Effects of water
hardness and alkalinity on the toxicity of uranium to a tropical freshwater
hydra (Hydra viridissima). Biomarkers 6:45–51.
Santore R, Di Toro D, Paquin PR, Allen HE, Meyer J. 2001. Biotic ligand
model of the acute toxicity of metals. 2. Application to acute copper
toxicity in freshwater fish and Daphnia. Environ Toxicol Chem
20:2397–2402.
Santore R, Ryan A, Kroglund F, Rodriguez P, Stubblefield W, Cardwell A,
Adams W, Nordheim E. 2017. Development and application of a biotic
ligand model for predicting the chronic toxicity of dissolved and pre-
cipitated aluminum to aquatic organisms. Environ Toxicol Chem
37:70–79.
Toxicity of contaminant mixtures from a uranium mine—Environmental Toxicology and Chemistry, 2021;40:2334–2346 2345
wileyonlinelibrary.com/ETC © 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC
Schlekat C, Stubblefield W, Gallagher K. 2020. State of the science on metal
bioavailability modeling: Introduction to the outcome of a society of
environmental toxicology and chemistry technical workshop. Environ
Toxicol Chem 39:42–47.
Serkiz S, Perdue EM. 1990. Isolation of dissolved organic matter from
Suwannee River using reverse osmosis. Water Res 7:911–916.
Smart GR. 1978. Investigations of the toxic mechanisms of ammonia to
fish–gas exchange in rainbow trout (Salmo gairdneri) exposed to acutely
lethal concentrations. J Fish Biol 12:93–104.
Supervising Scientist. 2018a. Uranium and manganese in surface water—
Rehabilitation standard for the Ranger uranium mine (Ver 1). Supervising
Scientist Branch, Darwin, Northern Territory, Australia. [cited 2020 De-
cember 24]. Available from: http://www.environment.gov.au/science/
supervising‐scientist/publications/ss‐rehabilitation‐standards
Supervising Scientist. 2018b. Magnesium in surface water—Rehabilitation
standard for the Ranger uranium mine (Ver 1). Supervising Scientist
Branch, Darwin, Northern Territory, Australia. [cited 2020 December 24].
Available from: http://www.environment.gov.au/science/supervising‐
scientist/publications/ss‐rehabilitation‐standards
Supervising Scientist. 2018c. Ammonia in surface water—Rehabilitation
standard for the Ranger uranium mine (Ver 1). Supervising Scientist
Branch, Darwin, Northern Territory, Australia. [cited 2020 December 24].
Available from: http://www.environment.gov.au/science/supervising‐
scientist/publications/ss‐rehabilitation‐standards
Tipping E, Lofts S, Sonke JE. 2011. Humic ion‐binding model VII: A revised
parameterisation of cation‐binding by humic substances. Environ Chem
8:225–235.
Trenfield M, Harford A, Mooney T, Ellis M, Humphrey C, van Dam R. 2017.
Toxicity of contaminated waters from Gulungul Creek Tributary 2 in
2015 and 2016. Internal Report 652. Supervising Scientist Branch,
Darwin, Northern Territory, Australia.
Trenfield MA, Ng JC, Noller BN, Markich SJ, van Dam RA. 2011. Dissolved
organic carbon reduces uranium bioavailability and toxicity. 2. Uranium
[VI] speciation and toxicity to three tropical freshwater organisms.
Environ Sci Technol 45:3082–3089.
Trenfield MA, Ng JC, Noller BN, Markich SJ, van Dam RA. 2012a. Dissolved
organic carbon reduces uranium toxicity to the unicellular eukaryote
Euglena gracilis.Ecotoxicology 21:1013–1023.
Trenfield MA, Ng JC, Noller BN, Markich SJ, van Dam RA. 2012b. Dissolved
organic carbon reduces the toxicity of aluminum to three tropical
freshwater organisms. Environ Toxicol Chem 31:427–436.
Trenfield MA, Pease CJ, Walker SL, Mooney T, Tybell L, Humphrey C, van
Dam RA, Harford AJ. 2020. Standardised chronic toxicity test protocols
and culturing methods for a suite of tropical freshwater species. Aus-
tralas Bull Ecotox Environ Cont 6:1–80.
Turner K, Tayler K, Tyrrell JWR. 2019. Revised Ranger mine water
quality objectives for Magela Creek and Gulungul Creek. Internal
Report 659. Supervising Scientist Branch, Darwin, Northern Territory,
Australia.
US Environmental Protection Agency. 1993. Method 350.1. Determi-
nation of ammonia nitrogen by semi‐automated colorimetry.
Cincinnati, OH.
US Environmental Protection Agency. 2017. Water Quality Standards
Handbook. Chapter 3. Water quality criteria. EPA‐823‐B‐17‐001. Office
of Water, Office of Science and Technology, Washington, DC. [cited
2020 April]. Available from: https://www.epa.gov/sites/production/files/
2014‐10/documents/handbook‐chapter3.pdf
van Dam RA, Hogan AC, Harford AJ. 2017. Derivation and implementation
of a site‐specific water quality limit for uranium in a high conservation
value ecosystem. Integr Environ Assess Manag 13:765–777.
van Dam RA, Hogan A, McCullough C, Houston M, Humphrey C, Harford A.
2010. Aquatic toxicity of magnesium sulfate and the influence of calcium
in very low ionic concentration water. Environ Toxicol Chem
29:410–421.
van Dam RA, Trenfield MA, Markich SJ, Harford AJ, Humphrey CL, Hogan
AC, Stauber JL. 2012. Reanalysis of uranium toxicity data for selected
freshwater organisms and the influence of dissolved organic carbon.
Environ Toxicol Chem 31:2606–2614.
van Genderen E, Adams W, Dwyer R, Garman E, Gorsuch J. 2015. Modeling
and interpreting biological effects of mixtures in the environment: In-
troduction to the metal mixture modeling evaluation project. Environ
Toxicol Chem 34:721–725.
van Regenmortel T, De Schamphelaere KA. 2018. Mixtures of Cu, Ni, and
Zn act mostly noninteractively on Pseudokirchneriella subcapitata
growth in natural waters. Environ Toxicol Chem 37:587–598.
Vijver MG, Elliott EG, Peijnenburg WJGM, De Snoo GR. 2011. Response
predictions for organisms water‐exposed to metal mixtures: A meta‐
analysis. Environ Toxicol Chem 30:1482−1487.
Wilde K, Stauber J, Markich S, Franklin N, Brown P. 2006. The effect
of pH on the uptake and toxicity of copper and zinc in a tropical
freshwater alga (Chlorella sp.). Arch Environ Contam Toxicol
51:174−185.
2346 Environmental Toxicology and Chemistry, 2021;40:2334–2346—M.A. Trenfield et al.
© 2021 Commonwealth of Australia. Environmental Toxicology and Chemistry © 2021 SETAC wileyonlinelibrary.com/ETC
