Article Cite This: Environ. Sci. Technol. XXXX, XXX, XXX−XXX
pubs.acs.org/est
Potential Toxicity of Dissolved Metal Mixtures (Cd, Cu, Pb, Zn) to Early Life Stage White Sturgeon (Acipenser transmontanus) in the Upper Columbia River, Washington, United States Laurie S. Balistrieri,*,† Christopher A. Mebane,*,‡ Stephen E. Cox,§ Holly J. Puglis,∥ Robin D. Calfee,∥ and Ning Wang∥
Environ. Sci. Technol. Downloaded from pubs.acs.org by KAROLINSKA INST on 08/19/18. For personal use only.
†
United States Geological Survey, Geology, Minerals, Energy, and Geophysics Science Center, Grafton, Wisconsin 53024, United States ‡ Idaho Water Science Center, Boise, Idaho 83702, United States § Washington Water Science Center, Tacoma, Washington 98402, United States ∥ Columbia Environmental Research Center, Columbia, Missouri 65201, United States S Supporting Information *
ABSTRACT: The Upper Columbia River (UCR) received historical releases of smelter waste resulting in elevated metal concentrations in downstream sediments. Newly hatched white sturgeon hide within the rocky substrate at the sediment− water interface in the UCR for a few weeks before swim-up. Hiding behavior could expose them to metal contaminants, and metal toxicity could contribute to population declines in white sturgeon over the past 50 years. This study evaluates whether there is a link between the toxicity of dissolved metals across the sedimentwater interface in the UCR and the survival of early life stage (ELS) white sturgeon. Toxicity of dissolved metal mixtures is evaluated using a combination of previously collected laboratory and field data and recently developed metal mixture toxicity models. The laboratory data consist of individual metal (Cd, Cu, Pb, and Zn) toxicity studies with ELS white sturgeon. The field data include the chemical composition of surface and pore water samples that were collected across the sediment−water interface in the UCR. These data are used in three metal accumulation and two response models. All models predict low toxicity in surface water, whereas effects concentrations greater than 20% are predicted for 60−72% of shallow pore water samples. The flux of dissolved metals, particularly Cu, from shallow pore water to surface water likely exposes prime ELS sturgeon habitat to toxic conditions.
■
for Cu.2 Since this metal contamination poses potential risks to the health of humans, terrestrial and aquatic life, and the environment, the United States Environmental Protection Agency is currently conducting a Remedial Investigation and Feasibility Study (RI/FS) from the U.S./Canada border to Grand Coulee Dam to assess these risks (https://www.epa. gov/columbiariver/upper-columbia-river-remedialinvestigation-feasibility-study, accessed April 24, 2018). The scientific challenges of this assessment are to collect the necessary water, sediment, and biological data of this large, diverse area as well as develop approaches and models to properly interpret the data so that potential health risks are confidently evaluated. Epibenthic organisms living at the sediment−water interface and in the shallow substrates are often of interest in ecological
INTRODUCTION
The region around the Upper Columbia River Basin in northeastern Washington State, United States and southeastern Province of British Columbia, Canada is rich in mineral deposits containing copper, gold, lead, silver, and zinc with smaller amounts of antimony, arsenic, cobalt, iron, manganese, and molybdenum. The ore deposits were mined and smelted at multiple locations within the basin since the mid to late 1800s. As of 2018, the only remaining operational smelter is at Trail, British Columbia. This facility discharged metal-enriched slag and processing effluent into the free-flowing sections of the Upper Columbia River (UCR). Advances in technology through time reduced the amount of discharged dissolved metals while the release of slag from this facility ceased in 1995. However, the legacy of these activities is elevated concentrations of metals in downstream sediment.1 Laboratory studies of pore water in slag-containing sediment from the UCR indicate that dissolved concentrations of Cd, Cu, Pb, and Zn are greater than in interstitial water from reference sediments and also greater than chronic water quality criteria This article not subject to U.S. Copyright. Published XXXX by the American Chemical Society
Received: April 27, 2018 Revised: July 27, 2018 Accepted: July 30, 2018
A
DOI: 10.1021/acs.est.8b02261 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
the toxicity of metal mixtures in surface and pore water in the UCR.12 The objective of the present study is to determine whether the sensitivity of ELS sturgeon to metal-enriched water within their habitat could be a contributing factor to their poor recruitment in the UCR. Our approach is to (1) combine laboratory data from individual metal toxicity studies of ELS sturgeon8−10 with the composition of field-collected, multiple metal water samples across the sediment-water interface in the UCR11 and (2) model the data using three different competitive biotic ligand models that determine multiple metal accumulation on fish gills13−15 and two response models (i.e., independent joint action and concentration addition models) that link metal accumulation to biological response for dissolved metal mixtures.14,16,17
risk assessments of mining impacted sites. One species of interest in the UCR is white sturgeon (Acipenser transmontanus). These fish are large and slow-growing (up to 6 m in length and 816 kg in weight as adults), attain an average age of 100 years, reach reproductive maturity at ∼26 years for females, and have a spawning periodicity of 4 years.3 White sturgeon spawn in freshwater habitats with gravel, cobbles, and boulders during peak river flow. Newly hatched sturgeon spend a few weeks hiding in these hard substrates and associated interstitial fine grain sediment before reaching their swim-up stage. Hence, this behavior likely exposes ELS sturgeon in the UCR to metal-enriched sediment and pore water. Globally, nearly all sturgeon are considered threatened or endangered because of overfishing and physical or chemical degradation of habitat.3 The history of recruitment for three spatially distinct white sturgeon groups in the UCR indicates population declines began around the time of dam construction and flow regulation from about 1967−1977 and populations have remained low since.4 Using a weight-ofevidence approach with emphasis on the spatial and temporal patterns of recruitment failure, McAdam4 evaluated 12 hypotheses, including overfishing, flow regulation, and changes in geomorphology, water quality (i.e., temperature, total gas pressure, turbidity, nutrients, contaminants in the water column), and fish communities, as possible causes for population declines and concluded that substrate alteration of spawning sites by fine sediment was the most plausible cause. The present study reconsiders the role of metal contaminants in white sturgeon population declines in the UCR. The reasons are 3-fold. First, recent studies of dissolved metal (Cd, Cu, Pb, Zn) toxicity provide dose−response curves for early life stage (ELS) white sturgeon (exposure age 1−30 days post hatch (dph)). Unlike previous acute studies in which death was the only end point evaluated,5−7 these recent studies evaluated more sensitive sublethal, short-term effects through the “effective mortality” end point (i.e., lack of hiding, loss of equilibrium, and death).8−10 For fish that survived short-term exposures, biomass end points have been tracked as a longerterm, sublethal response. Biomass is a combined end point calculated by multiplying growth and mortality responses together.9 ELS sturgeon are ∼10x more sensitive to Cu than older juvenile sturgeon (61−100 dph) when using 50% effects concentrations to compare sensitivity.5,6,8,9 Thus, the impact of metal-enriched pore water in the UCR on sturgeon during their vulnerable early life stage needs evaluation. Second, the challenge of collecting water samples across the sediment− water interface (i.e., the habitat of ELS sturgeon) in the swiftly flowing UCR has been overcome. The development and use of a profile sampler that simultaneously collects water samples above (7.5 cm), at (0 cm), and below (4.5 and 14.5 cm) the sediment-interface has provided in situ dissolved metal data to assess potential metal toxicity to ELS sturgeon during their hiding phase.11 This field sampling technique collects water samples that are representative of in situ chemistry and hydrology and allows a more realistic evaluation of dissolved metal exposure to benthic fish than either water column sampling or centrifuged pore water from conventional sediment sampling. And third, recently developed models that consider the simultaneous interactions of multiple dissolved metals with biotic ligands and the biological effects of metal accumulation on biotic ligands can be used to predict
■
MATERIALS AND METHODS Methods. 2.1.1. Laboratory Toxicity Studies. The toxicity studies of individual dissolved metals (Cd, Cu, Pb, and Zn) to ELS sturgeon were conducted at the United States Geological Survey Columbia Environmental Research Center in Columbia, Missouri. The details of the experimental setup and test conditions are in previous publications,8−10,18 and the dissolved metal concentrations and biological results used in the present study are summarized in the Supporting Information (SI) (Table S1). Data from the laboratory toxicity studies were selected and characterized as shorter-term based on an end point of effective mortality and longer-term based on an end point of biomass. Longer term (24−53 days) toxicity tests started with 1−27 dph white sturgeon in exposure concentrations of dissolved Cd ( Cd, whereas the 2site accumulation model indicates that the sensitivity is Cd ∼ Cu > Pb > Zn. The sensitivity sequence for the 4-site model is Pb > Cd ∼ Cu > Zn. From the perspective of metal load on the biotic ligand, consistent identification of the most toxic metal to ELS sturgeon is not possible because of differences in binding constants for biotic ligand−metal reactions among the three biotic ligand−metal accumulation models. 3.1.3. Tox-Response Models for the Laboratory Toxicity Studies. First developed by Stockdale, Tipping, Lofts, Ormerod, Clements, and Blust,17 the toxicity function, Tox, is designed to account for site-specific chemical speciation of dissolved metals, metal accumulation at the biotic ligand, and the relative potency of each metal. This toxicity function is combined with associated biological responses to provide a single dose (Tox)-response curve that incorporates the chemical speciation and biological impacts of multiple metals. For our calculations, the potency of each metal (Cd, Cu, Pb) is evaluated relative to Zn, that is, the potency coefficient for Zn, αZn, is defined to be 1. Thus, if ELS sturgeon are more sensitive to a given metal accumulation than to Zn accumulation, then the potency coefficient for that metal is greater than 1. If sturgeon are less sensitive to a given metal than Zn, then the potency coefficient for the metal is less than 1. Because the three biotic ligand−metal accumulation models predict different metal loads and metal sensitivities, the potency coefficients, summarized in SI Table S4, also vary among the three models. For example, potency coefficients for Cd vary from 0.10 ± 0.02 (1-site accumulation model) to 3.2 ± 0.5 (2-site accumulation model) to 53 ± 10 (4-site accumulation model). These differences in potency coefficients are reciprocal to the metal-biotic ligand binding constants used in the models, which vary by 2 orders of magnitude. The 1-site model with its Cd-BL log K value of 7.59 had the lowest Cd potency factor and the 4-ste model with its Cd-BL log K value of 5.4 had the highest Cd potency factor (SI Table S4).
interactions among the dissolved metals, inorganic ligands (e.g., OH−, HCO3−, Cl−), and organic ligands (i.e., DOC and biotic ligands). The Tox versus response curve developed from the laboratory studies is used to predict sturgeon response at the Tox values of the dissolved metal mixtures.14,16
■
RESULTS AND DISCUSSION Development of Predictive ELS Sturgeon-Response Models. Predictive response models for ELS sturgeon during exposure to field-collected dissolved metal mixtures in the UCR are developed using data from laboratory studies of individual metal toxicity to ELS sturgeon. Model development entails three steps. The first is to collate data from previously published toxicity studies that examined the response of ELS sturgeon to variable dissolved concentrations of individual metals (Cd, Cu, Pb, Zn). The second step is to use biotic ligand models to predict the accumulation of metal on the biotic ligand for the variable solution compositions in the toxicity studies. And the third step is to relate predicted biotic ligand−metal accumulation to observed biological responses. 3.1.1. Laboratory Dose−Response. Although this study selected shorter-term toxicity data based on an end point of effective mortality and longer-term data on an end point of biomass, the biological responses to changes in either end point are the same. That is, similar dose (free metal ion concentration)-response curves are observed for the two different end points with effective mortality and reduction in biomass increasing as free ion concentrations of the metals increase (Figure 1). The use of free ion concentrations, which
Figure 1. Log free metal ion concentrations for Cd, Cu, Pb, and Zn versus shorter-term effective mortality or longer-term reduction in biomass from laboratory toxicity testing of early life stage white sturgeon. Data were fit with a logistic equation. Solution speciation was determined with WHAM 7.
are determined from equilibrium speciation calculations, accounts for different compositions among solutions (e.g., DOC and major ion concentrations). Fitting the combined data sets yields a free ion concentration at 20% biological response (EC20) of 41 ± 3 nM Cd2+, 0.29 ± 0.02 nM Cu2+, 0.9 ± 0.1 nM Pb2+, and 1750 ± 72 nM Zn2+. From a free ion concentration perspective, the relative toxicity of the free metal ions to ELS sturgeon is Cu2+ > Pb2+ > Cd2+ > Zn2+. On a molar basis, Cu2+ is about 3 times more toxic than Pb2+, 137 times more toxic than Cd2+, and 5830 times more toxic than Zn2+ to ELS sturgeon. D
DOI: 10.1021/acs.est.8b02261 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
Figure 2. Calculated log biotic ligand−metal (BL-Me) fractional accumulation (upper panel) or log Tox (toxicity function) (lower panel) using three different sets of equilibrium constants for biotic ligand−cation reactions versus shorter-term effective mortality or longer-term reduction in biomass from laboratory toxicity testing of early life stage sturgeon. Data were fit with logistic equations.
The combination of metal accumulation and potency coefficients results in a single Tox versus biological response (i.e., effective mortality or reduction in biomass) curve for each of the three biotic ligand−metal accumulation models (Figure 2 lower panel). Values of Tox at 50% response (Toxc) of the ELS sturgeon determined from the logistic fits presented in SI Table S3 are 0.033 ± 0.003 (1-site accumulation model), 0.100 ± 0.007 (2-site accumulation model), and 0.086 ± 0.009 (4site accumulation model), when Zn is defined as the reference metal. Tox values at 50% response would be different if a different metal (e.g., Cd) is defined as the reference metal. Because the Tox-response curves incorporate the effects of dissolved Cd, Cu, Pb, and Zn concentrations on effective mortality and reduction in biomass of ELS sturgeon, these curves are used in the Tox addition approach for assessing potential impacts to sturgeon health using multiple dissolved metal concentrations in field-collected water from the UCR (see below). Application of Predictive ELS Sturgeon-Response Models to Field-Collected Dissolved Metal Mixtures. 3.2.1. Free Metal Ion Concentrations Across the Sediment− Water Interface in the UCR. A comparison of free metal ion concentrations of Cd, Cu, Pb, and Zn in water just above and below the sediment−water interface in the UCR with EC20 concentrations of the free metal ions from the laboratory toxicity studies indicates that Cd 2+ , Pb 2+ , and Zn 2+
concentrations are below EC20 values at all depths, except 1 deep pore water sample for Pb2+ (Figure 3). The same cannot be said for Cu2+ concentrations, which, as previously noted, are the most toxic to ELS sturgeon. Although no samples at 7.5 cm above the interface had Cu 2+ concentrations greater than the EC20 concentrations, a few at the sediment-water interface (0 cm) and ∼50% of the pore water samples below the interface had Cu2+ concentrations greater than the laboratory derived EC20. In addition, there are gradients in the concentrations of Cu2+, Pb2+, and Zn2+ across the sediment-water interface with fluxes of these free cations from pore water at 4.5 cm below the interface to the interface. The Sums of Toxic Units (TU) for the free metal ions are dominated by Cu2+ and, as illustrated in SI Figure S1, are greater than 1 for a few samples at the interface and for 50% of pore water samples at 4.5 cm below the interface. These results suggest that ELS sturgeon in their hiding phase at or below the sediment−water interface are likely exposed to toxic concentrations and fluxes of Cu2+ at some sites in the UCR. 3.2.2. Predicted Metal Accumulation on ELS Sturgeon Gills in the UCR. Because the 3 biotic ligand−metal accumulation models have different equilibrium constants for biotic ligand−metal reactions, the accumulation of metal among the models is different for a given water composition. These differences are illustrated in Figure 4 for predicted metal E
DOI: 10.1021/acs.est.8b02261 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
across the sediment-water interface in the UCR could adversely impact sturgeon health and survival. Predicted shorter-term effective mortality and longer-term reduction in biomass in all surface water (i.e., 7.5 cm above the interface) samples and most samples at the sediment−water interface (0 cm) are low relative to health impacts in shallow (4.5 cm below interface) and deeper (14.5 cm below the interface) pore water (Figure 5). Independent joint action using results from the 3 metal accumulation models predicts a median adverse response (i.e., effective mortality or reduction in biomass) of 11−17% for surface water (SW), 15−19% for water at the sediment-water interface (SWI), 36−47% for shallow pore water (SPW), and 58−65% for deep pore water (DPW). The corresponding predicted values using Tox addition are 5−6% for SW, 8−11% for SWI, 32−51% for SPW, and 68−83% for DPW. From another perspective and illustrated in theSI Figure S2, the percentages of field samples predicted to have greater than EC20 are 11−17% for SW, 24−43% for SWI, and 57−72% for SPW and DPW using the independent joint action approach and 0−11% for SW, 14−24% for SWI, and 60−71% for SPW and DPW using the Tox addition approach. The results indicate that predicted toxicity of dissolved metal mixtures across the sediment−water interface to ELS sturgeon in the UCR is remarkably similar among the different metal accumulation and response models. Even though the models varied greatly in their structure and in their intermediate results, that is, which metals were accumulating to greater extent or contributing more to the overall toxicity, the overall toxicity predictions were similar. While these results support the practical value of these metal−-mixture modeling approaches, as noted elsewhere,14,16 it is not obvious which if any of the various, related modeling approaches should be considered optimal. The modeling approaches are very different. The 1-site accumulation model has started with linear free energy relationships (LFERs),24 which were then adjusted to fit toxicity data.14 The 2-site accumulation model used actual accumulations on fish gills to determine binding coefficients,13 which avoided some problems of unconstrained calibrations and nonunique solutions that may arise when metals accumulation is treated as a model construct rather than being constrained by data. The 4-site accumulation model is a compendium model, with log K values partially rooted in actual accumulations, but also adjusted to fit toxicity data.15 Because of the differing model structures, we did not expect the models to provide similar toxicity predictions. The use of the three different modeling approaches gives some indication of model associated uncertainties and confidence in the outcomes. Metal sensitive and vulnerable ELS sturgeon that spend part of their life cycle hiding at or below the sediment-water interface will be exposed to toxic concentrations and fluxes of dissolved Cu2+ ions, at some, but not all, locations in the UCR. This exposure is predicted to result in increases in effective mortality and reductions in biomass of ELS sturgeon. Hence, the presence of metal contaminants in the habitat of ELS sturgeon could be a contributing factor in the loss of generations of white sturgeon and to their population declines in the UCR. This modeling work provides a bridge between laboratory derived toxicity data and field exposure data. The integration of laboratory studies using advanced end points and in situ porewater measurements through recently developed multiple
Figure 3. Free metal ion concentrations for Cd, Cu, Pb, and Zn versus depth across the sediment-water interface in the upper reaches of the Columbia River. The whisker plots indicate ranges, 25th percentiles, medians, 75th percentiles, and outliers. The EC20 values for early life stage sturgeon determined from laboratory toxicity testing of individual metals are shown as dashed red lines.
accumulation on ELS sturgeon in shallow pore water from the UCR.
Figure 4. Calculated fractional metal loads (BL-metal) on early life stage sturgeon gills using three different sets of equilibrium constants for biotic ligand-cation reactions (1-site, 2-site, 4-site accumulation models) for the compositions of shallow pore water (−4.5 cm below the sediment-water interface) in the upper reaches of the Columbia River. The whisker plots indicate ranges, 25th percentiles, medians, 75th percentiles, and outliers.
For these pore water compositions, metals occupy less than 10% and, in many cases, less than 1% of the total biotic ligand sites of ELS sturgeon. The metal loads are Cd ∼ Cu ∼ Zn ≫ Pb for the 1-site accumulation model, Cu > Pb ∼ Zn ≫ Cd for the 2-site accumulation model, and Cu ∼ Zn ≫ Pb > Cd for the 4-site accumulation model. 3.2.3. Predictions of Potential Health Impacts to ELS Sturgeon in the UCR. Both independent joint action and Tox addition are used to predict whether the accumulation of metal on ELS sturgeon gills exposed to dissolved metal mixtures F
DOI: 10.1021/acs.est.8b02261 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
Figure 5. Predicted shorter-term effective mortality or longer-term reduction in biomass for early life stage sturgeon across the sediment−water interface in the upper reaches of the Columbia River. Three different metal accumulation models (1-site, 2-site, 4-site) and two different response models [independent joint action (upper panel) and Tox addition (lower panel)] are used in the calculations. The whisker plots indicate ranges, 25th percentiles, medians, 75th percentiles, and outliers. The dashed lines connect the median values.
metal−mixture BLMs13−15,24 advances the state of science of lab to field toxicity extrapolations.
■
Investigation/Feasibility Study (RI/FS). The funds for the Upper Columbia River Project are provided by Teck American Incorporated via the Department of Interior Central HazMat fund and are subsequently distributed to and managed by the United States Geological Survey Washington Water Science Center.
ASSOCIATED CONTENT
S Supporting Information *
■
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.8b02261. Four tables containing summary data from laboratory toxicity studies (Table S1, 5 pages), equilibrium constants for biotic ligand−cation reactions (Table S2) and logistic parameters for fitting curves of biotic ligand−metal accumulations (Table S3) or toxicity functions (Table S4) versus biological response, and two figures containing Sum of Toxic Units (Figure S1) or fraction of field samples with >20% adverse response (Figure S2) versus depth across the sediment−water interface (PDF)
■
REFERENCES
(1) Johnson, A.; Norton, D.; Yake, B.; Twiss, S. Transboundary metal pollution of the Columbia River (Franklin D. Roosevelt Lake). Bull. Environ. Contam. Toxicol. 1990, 45 (5), 703−710. (2) Paulson, A. J.; Cox, S. E. Release of elements to natural water from sediments of Lake Roosevelt, Washington, USA. Environ. Toxicol. Chem. 2007, 26 (12), 2550−2559. (3) Pikitch, E. K.; Doukakis, P.; Lauck, L.; Chakrabarty, P.; Erickson, D. L. Status, trends and management of sturgeon and paddlefish fisheries. Fish Fish. 2005, 6 (3), 233−265. (4) McAdam, D. S. O. Retrospective weight-of-evidence analysis identifies substrate change as the apparent cause of recruitment failure in the upper Columbia River white sturgeon (Acipenser transmontanus). Can. J. Fish. Aquat. Sci. 2015, 72 (8), 1208−1220. (5) Little, E. E.; Calfee, R. D.; Linder, G., Toxicity of copper to earlylife stage Kootenai River white sturgeon, Columbia River white sturgeon, and rainbowtrout. Arch. Environ. Contam. Toxicol. 2012.63400 (6) Vardy, D. W.; Oellers, J.; Doering, J. A.; Hollert, H.; Giesy, J. P.; Hecker, M. Sensitivity of early life stages of white sturgeon, rainbow trout, and fathead minnow to copper. Ecotoxicology 2013, 22 (1), 139−147. (7) Vardy, D. W.; Santore, R. C.; Ryan, A. C.; Giesy, J. P.; Hecker, M. Acute toxicity of copper, lead, cadmium, and zinc to early life stages of white sturgeon (Acipenser transmontanus) in laboratory and Columbia River water. Environ. Sci. Pollut. Res. 2014, 21, 1−12. (8) Calfee, R. D.; Little, E. E.; Puglis, H. J.; Scott, E.; Brumbaugh, W. G.; Mebane, C. A. Acute sensitivity of white sturgeon (Acipenser
AUTHOR INFORMATION
Corresponding Authors
*(L.S.B.) E-mail:
[email protected]. *(C.A.M.) E-mail:
[email protected]. ORCID
Laurie S. Balistrieri: 0000-0002-6359-3849 Christopher A. Mebane: 0000-0002-9089-0267 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was done as part of the United States Department of Interior’s efforts on the Upper Columbia River Remedial G
DOI: 10.1021/acs.est.8b02261 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
transmontanus) and rainbow trout (Oncorhynchus mykiss) to copper, cadmium, or zinc in water-only laboratory exposures. Environ. Toxicol. Chem. 2014, 33 (10), 2259−2272. (9) Wang, N.; Ingersoll, C. G.; Dorman, R. A.; Brumbaugh, W. G.; Mebane, C. A.; Kunz, J. L.; Hardesty, D. K. Chronic sensitivity of white sturgeon (Acipenser transmontanus) and rainbow trout (Oncorhynchus mykiss) to cadmium, copper, lead, or zinc in laboratory water-only exposures. Environ. Toxicol. Chem. 2014, 33 (10), 2246− 2258. (10) Puglis, H. J.; Calfee, R. D.; Little, E. E., Behavioral effects of copper on larval White Sturgeon. Environ. Toxicol. Chem. accepted, pending revisions. (11) Cox, S. E.; Brumbaugh, W. G.; Balistrieri, L. S.; Wolf, R. E.; Adams, M.; Spanjer, A. J.; Olsen, T. D., Trace elements concentrations in pore water and surface water near the sedimentwater interface in the Upper Columbia River,(2015); U.S. Geological Survey Data Release: Washington, https://dx.doi.org/10.5066/ F73N21GJ, 2016. (12) Van Genderen, E.; Adams, W.; Dwyer, R.; Garman, E.; Gorsuch, J. Modeling and interpreting biological effects of mixtures in the environment: Introduction to the metal mixture modeling evaluation project. Environ. Toxicol. Chem. 2015, 34 (4), 721−725. (13) Balistrieri, L. S.; Mebane, C. A. Predicting the toxicity of metal mixtures. Sci. Total Environ. 2014, 466−467, 788−799. (14) Farley, K. J.; Meyer, J. S. Metal Mixture Modeling Evaluation project: 3. Lessons learned and steps forward. Environ. Toxicol. Chem. 2015, 34 (4), 821−832. (15) Santore, R. C.; Ryan, A. C. Development and application of a multimetal multibiotic ligand model for assessing aquatic toxicity of metal mixtures. Environ. Toxicol. Chem. 2015, 34 (4), 777−787. (16) Balistrieri, L. S.; Mebane, C. A.; Schmidt, T. S.; Keller, W. B. Expanding metal mixture toxicity models to natural stream and lake invertebrate communities. Environ. Toxicol. Chem. 2015, 34 (4), 761− 776. (17) Stockdale, A.; Tipping, E.; Lofts, S.; Ormerod, S. J.; Clements, W. H.; Blust, R. Toxicity of proton-metal mixtures in the field: Linking stream macroinvertebrate species diversity to chemical speciation and bioavailability. Aquat. Toxicol. 2010, 100 (1), 112− 119. (18) Wang, N.; Calfee, R. D.; Beahan, E.; Brumbaugh, W. G.; Dorman, R. A.; Hardesty, D. K.; Ingersoll, C. G.; Kunz, J. L.; Little, E. E.; Mebane, C. A.; Puglis, H. J., Acute and chronic sensitivity of white sturgeon (Acipenser transmontanus) and rainbow trout (Oncorynchus mykiss) to cadmium, copper, lead, or zinc in laboratory wateronly exposures. U.S. Geological Survey Scientific Investigations Report 2013−5204 2013. (19) Paquin, P. R.; Gorsuch, J. W.; Apte, S.; Bately, G. E.; Bowles, K. C.; Campbell, P. G. C.; Delos, C. G.; Di Toro, D. M.; Dwyer, R. L.; Galvez, F.; Gensemer, R. W.; Goss, G. G.; Hogstrand, C.; Janssen, C. R.; McGeer, J. C.; Naddy, R. B.; Playle, R. C.; Santore, R. C.; Schneider, U.; Stubblefield, W. A.; Wood, C. M.; Wu, K. B. The biotic ligand model: a historical overview. Comp. Biochem. Physiol., Part C: Toxicol. Pharmacol. 2002, 133, 3−35. (20) Lofts, S. User’s Guide to WHAM7; NERC Centre for Ecology and Hydrology, 2012; p 97. (21) Tipping, E.; Lofts, S.; Sonke, J. E. Humic Ion-Binding Model VII: a revised parameterisation of cation-binding by humic substances. Environmental Chemistry 2011, 8, 225−235. (22) Bryan, S. E.; Tipping, E.; Hamilton-Taylor, J. Comparison of measured and modelled copper binding by natural organic matter in freshwaters. Comp. Biochem. Physiol., Part C: Toxicol. Pharmacol. 2002, 133, 37−49. (23) Tipping, E.; Rey-Castro, C.; Bryan, S. E.; Hamilton-Taylor, J. Al(III) and Fe(III) binding by humic substances in freshwaters, and implications for trace metal speciation. Geochim. Cosmochim. Acta 2002, 66 (18), 3211−3224. (24) Carbonaro, R. F.; Atalay, Y. B.; Di Toro, D. M. Linear free energy relationships for metal−ligand complexation: Bidentate
binding to negatively-charged oxygen donor atoms. Geochim. Cosmochim. Acta 2011, 75 (9), 2499−2511. (25) De Schamphelaere, K. A. C.; Janssen, C. R. A biotic ligand model predicting acute copper toxicity for Daphnia magna: the effects of calcium, magnesium, sodium, potassium, and pH. Environ. Sci. Technol. 2002, 36 (1), 48−54.
H
DOI: 10.1021/acs.est.8b02261 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
