Toxicity of oil sands acid-extractable organic fractions to freshwater fish: Pimephales promelas (fathead minnow) and Oryzias latipes (Japanese medaka)

Chemosphere 171 (2017) 168e176

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

Toxicity of oil sands acid-extractable organic fractions to freshwater fish: Pimephales promelas (fathead minnow) and Oryzias latipes (Japanese medaka) Anthony E. Bauer a, *, Richard A. Frank b, John V. Headley c, Kerry M. Peru c, Andrea J. Farwell a, D. George Dixon a a

Biology Department, University of Waterloo, 200 University Avenue West, Waterloo, Ontario N2L 3G1, Canada Aquatic Contaminants Research Division, Water Science and Technology Directorate, Environment Canada, 867 Lakeshore Road, Burlington, Ontario L7S 1A1, Canada c Aquatic Contaminants Research Division, Water Science and Technology Directorate, Environment Canada, 11 Innovation Boulevard, Saskatoon, Saskatchewan S7N 3H5, Canada b

h i g h l i g h t s
Acid extractable organic (AEO) fraction toxicity was not driven by molecular weight.
Lowest molecular weight Fraction 1 displayed lowest toxicity to both species.
Fathead minnow were more sensitive to AEO fractions than Japanese medaka.
Toxicity appears to be driven by other modes of action than solely narcosis.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 27 September 2016 Received in revised form 8 December 2016 Accepted 12 December 2016 Available online 14 December 2016

The Alberta oil sands are one of the largest global petroleum deposits and, due to non-release practices for oil sands process-affected waters, produced tailings are stored in large ponds. The acid extractable organic (AEO) compounds in oil sands process-affected water are of greatest concern due to their persistence and toxicity to a variety of aquatic biota. The present study evaluated the toxicity of the five AEO fractions to two fish species: Oryzias latipes (Japanese medaka) and Pimephales promelas (fathead minnow). The fractions (F1-F5) were comprised of AEO with increasing mean molecular weight and subsequent increases in cyclicity, aromaticity, degree of oxygenation, and heteroatom content. The lowest molecular weight fraction, F1, displayed the lowest acute toxicity to both fish species. For fathead minnow, F5 displayed the greatest toxic potency, while F2 to F4 displayed intermediate toxicities. For Japanese medaka, F2 and F3 displayed the greatest acute toxicities and F1, F4 and F5 were significantly less potent. Overall, fathead minnow were more acutely sensitive to AEO than Japanese medaka. The present study indicates that AEO toxicity may not be solely driven by a narcotic mode of action, but chemical composition such as aromaticity and heteroatom content and their relation to toxicity suggest other drivers indicative of additional modes of toxic action. © 2016 Elsevier Ltd. All rights reserved.

Handling Editor: David Volz Keywords: Oil sands Acid-extractable organics (AEO) Naphthenic acids (NA) Fractional distillation Toxicity identification evaluation (TIE) Mode of action (MOA) Narcosis

1. Introduction The Canadian oil sands, primarily located in north-eastern Alberta, are one of the largest deposits of petroleum worldwide.

Abbreviations: AEO, Acid-extractable organics; MOA, Mode of action; NA, Naphthenic acids; OSPW, Oil sands process affected water. * Corresponding author. E-mail address: [email protected] (A.E. Bauer). http://dx.doi.org/10.1016/j.chemosphere.2016.12.059 0045-6535/© 2016 Elsevier Ltd. All rights reserved.

The Alberta oil sands region consist of three large bitumen deposits; the Cold Lake, Peace River and Athabasca River deposits, which encompass a total area of 140,200 km2 (Alberta Energy, 2008). Bitumen, a naturally occurring, viscous, hydrocarbon mixture is extracted via surface mining and in-situ methods and requires further upgrading to a synthetic crude oil. In 2015, the oil sands industry produced over 2.5 million barrels of bitumen a day, with 165 billion barrels of remaining established reserves (Energy Resources Conservation Board, 2014). As surface mining practices

A.E. Bauer et al. / Chemosphere 171 (2017) 168e176

require a caustic hot water wash of the mined oil sand to extract the marketable bitumen, vast quantities of tailings are created (Alberta Energy and Utilities Board, 2014). While originally stored to maximize water recycling, under the Alberta Environmental Protection and Enhancement Act, release of oil sands process-affected water (OSPW) and other tailings into the natural environment is prohibited, and disturbed sites must be reclaimed to a natural state (Madill et al., 2001). The constituents of greatest concern in OSPW include polar organic acids (Allen, 2008a; The Royal Society of Canada (2010)), of which a subclass containing only two oxygen atoms and no other heteroatoms (i.e., RCO2H, where R is a hydrocarbon group) are referred to as classical naphthenic acids (NA). These polar organic acids occur naturally in bitumen and are detected in rivers and lakes of the oil sands region (The Royal Society of Canada (2010)). Once released from bitumen under the alkaline conditions present in the caustic wash, recycling of water for bitumen extraction results in high concentrations of these polar organics in tailings (Allen, 2008b). Historically, organic acids present in OSPW, as well as in natural waters within the deposit area, were commonly referred to in the literature as NA. However, recent literature has clearly demonstrated that classical NA and products of mild oxidation (oxy-naphthenic acids) account for less than 50% of the total abundance of organic acids in the acid-extractable fraction (Barrow et al., 2010; Grewer et al., 2010; Headley et al., 2009). Other compounds present in oil sands acid extractable fractions include organic acids that have been found to contain multiple carboxyl and dihydroxy groups, heteroatoms, diamondoid acids, and aromatic rings (Barrow et al., 2010; Grewer et al., 2010; Headley et al., 2009; Bauer et al., 2015; Frank et al., 2009; Headley et al., 2013; Rowland et al., 2011). As such, the shared solubility parameters of the newly defined oil sands organic acids with NA, warrants that this highly complex mixture is referred to hereafter as acid extractable organics (AEO). A number of studies have documented the toxic potential of OSPW, containing high concentrations of AEO, to a number of aquatic organisms including fish (The Royal Society of Canada, 2010; Nero et al., 2006). However, many studies regarding the toxicity of oil sands OSPW have used commercially prepared NA mixtures as a surrogate for AEO. Unfortunately, commercial NA have been shown to display a greater toxicity than oil sands AEO at various endpoints, possibly due to differences in composition, and are not entirely representative (Marentette et al., 2015). Therefore, there is great value in conducting studies which assess toxicity using AEO extracts rather than whole OSPW, or commercial NA. Studies have suggested that narcosis is a primary mode of action (MOA) for acute toxicity of AEO (Frank et al., 2009, 2010; Tollefson et al., 2012). Narcosis is the nonspecific action of toxicants which disrupt the integrity of cellular membranes causing loss in membrane function. The surfactant nature of AEO means that their relative narcotic potencies can be considered as a function of their aqueous solubility (log Kow), where more hydrophobic compounds more likely cause membrane disruption. Although narcosis appears to be the main driver for AEO toxicity, studies have revealed the potential presence of other MOA (Tollefson et al., 2012; Scarlett et al., 2013; Swigert et al., 2015) which raises questions about the relationship between AEO structure and chemical composition to toxicity. Much of our understanding of the structure-toxicity relationship for AEO is derived from the following studies. Indigenous microbial communities have been shown to degrade AEO and reduce the toxicity of OSPW over time (Lai et al., 1996; Han et al., 2008). Greater proportions of higher molecular weight acids are present in aged tailings as a result of the slower degradation of more

169

recalcitrant acids, likely due to their greater structural complexity (Han et al., 2008; Del Rio et al., 2006). A study that utilized the Microtox® assay (Vibrio fischeri bioluminescence) to assess toxic effects of AEO fractions exhibiting a proportional shift to higher molecular weight acids, found a reduction in toxicity with increasing median AEO fraction molecular weight (Frank et al., 2009). The aforementioned studies have collectively provided supportive evidence, at least for the endpoints assessed, that lower molecular weight acids are more toxic than higher molecular weight acids. This finding is contrary to what one would expect for baseline narcosis of O2 AEO (classical NA) in which it has been demonstrated that higher molecular weight compounds exhibit lower aqueous solubilities (greater log Kow) and therefore greater toxicity (Frank et al., 2010; Jones et al., 2011). The discrepancy is likely because an increase in molecular weight for NA is solely due to an increase in carbon number, whereas an increase molecular weight for AEO is a result of contributions by other compounds (including oxygenated groups) (Bauer et al., 2015), which can increase solubility (lower log Kow) and therefore toxicity. Quantitative structureactivity relationship models have been used to provide further insight to this observation, noting that while increasing carbon number would likely increase toxicity, the addition of polar sub-groups would likely decrease toxicity (Frank et al., 2010). It is, therefore, expected that more recalcitrant, higher molecular weight acids contain additional polar sub-groups which may be responsible for reducing toxicity. Chemical characterization of five oil sands organic acid fractions using high-resolution electrospray ionization mass spectrometry (ESIMS) and synchronous fluorescence spectroscopy (Bauer et al., 2015) demonstrated this to be true. The fractions were isolated using a Kugelrohr distillation apparatus at stepped temperatures (130  C, 160  C, 190  C, 220  C, and >220  C), resulting in increasing mean molecular weight expressed in Daltons (Da) for each fraction (F1: 237 Da; F2: 240 Da; F3: 257 Da; F4: 308 Da; and F5: 355 Da, respectively). Fraction 5 (F5) represented residual AEO components that did not distill at 220  C, therefore, encompassing a range of AEO present in the original extract. A subsequent study that further chemically characterized these same organic acid fractions found that with increasing fraction mean molecular weight there were concurrent increases in double-bond equivalents, degree of aromaticity, relative heteroatom abundance, and oxygenated functional groups (Table 1) (Bauer et al., 2015). This shows that there was an overall increase in functional groups that could likely confer a reduction in toxicity with increasing fraction mean molecular weight, justifying assessments of their respective toxicities. Therefore, using the same five fractions characterized previously, the primary objective of the current study was to assess their toxicity to two small-bodied, freshwater fish species and compare the relative sensitivities and tolerances of the two fish species. 2. Materials and methods 2.1. Fish husbandry Breeding culture procedures for fathead minnow (Pimephales promelas) and Japanese medaka (Oryzias latipes) were approved by the institutional animal review board and Animal Care Committee at the University of Waterloo (UW-ACC). Culturing conditions maintained for fathead minnow followed Environment Canada fathead toxicity test protocols (Environment Canada, 2011), while culturing conditions for Japanese medaka were based on USEPA protocols (USEPA, 1991). Culture water consisted of 25% well water

170

A.E. Bauer et al. / Chemosphere 171 (2017) 168e176

Table 1 Chemical characteristics of acid-extractable organics fractions.a Fraction

Distillation Temp ( C)

Mean Molecular Weight (Da)

Molecular Weight Range (Da)

Proportion of AEOb

Total Ion Classes

O2 DBEc

F1 F2 F3 F4 F5

130 160 190 220 >220

237 240 257 308 355

162e335 195e274 209e310 255e322 122e547

8.3% 23.8% 26.7% 18.9% 10.0%

2 4 11 21 60

3,4 3,4 4,7 7,8 3,4

AEO ¼ acid extractable organics, DBE ¼ double-bond equivalents. a Information derived from Bauer et al. (2015). b Relative proportion to the combined AEO determined by residue weight. c DBE of O2 contributing up to 50% of each fraction.

(v/v) and 75% milli-Q water (v/v) on a re-circulating system, with water renewed weekly. Well water was diluted in order to maintain water hardness below 150 mg/L as CaCO3, pH ~8, and conductivity below 1000 mS/cm. Aeration of water was performed by air stones which maintained dissolved oxygen above 75% saturation. All fish were fed at least once a day with frozen brine shrimp and/or flaked food at a rate not exceeding what could be consumed in 10 min. Following the acclimation period, fish were transitioned into breeding conditions. Fathead minnow and Japanese medaka were separated into breeding sets consisting of a male:female ratio of 1:2 and placed into breeding tanks. To encourage breeding, water temperature was raised to 26 ± 1  C for fathead minnow and 28 ± 1  C for Japanese medaka, with a 16:8 (light:dark) photoperiod. After adult fathead minnow females laid their eggs on the underside of breeding tiles, embryos were removed by scraping tiles with blunt forceps. Fertilized Japanese medaka egg clutches could be found attached to the female oviduct pore. Embryo clutches were collected 1e3 h after the beginning of the light cycle, by placing females on a damp paper towel and removing eggs from their oviduct pore with blunt forceps. Embryos from both species were examined under a dissecting microscope and those near late blastula stage and free of fungal growth were selected and set aside for testing. 2.2. Source of oil sands AEO fractions An oil sands AEO extract collected and extracted in 2005 was fractionated as described previously (Frank et al., 2008). Briefly, 3000 L of OSPW was collected from the inflow pipe of West In-Pit settling basin of Syncrude Canada Ltd. Organic acids were isolated and concentrated from OSPW following an acid extraction and centrifugation procedure (Frank et al., 2008). The AEO extract was purified using a dichloromethane wash to remove neutral organics and filtered with diethylaminoethyl-cellulose to remove humic-like substances. The resulting extract was dissolved in 0.05 M sodium hydroxide and stored in 2-L amber glass bottles at 4  C. A 500-mL sample was taken from the original AEO extract to provide material for fractionation. Fractionation was conducted using a Kugelrohr distillation apparatus at increased stepped temperatures (130  C, 160  C, 190  C, and 220  C ±5  C) and held constant for 20 min, which yielded four fractions (F1-F4) and a fifth fraction with the residual non-distilled residue (F5), with respectively increasing mean molecular weight (237 Da, 240 Da, 257 Da, 308 Da, and 355 Da) (Bauer et al., 2015; Frank et al., 2008). An additional OSPW sample was obtained in 2009 at the same location, from the same inflow pipe for West In-Pit settling basin of Syncrude Canada Ltd., and therefore, represents a similar water source. The methods utilized for extraction of AEO were also identical to those described above. This extract was acquired for toxicity bioassays of pre-fractionated AEO and is hereafter also referred to as the “AEO extract”; it is not the original extract used in the fractionation. Consequently, we acknowledge that there may be

minor differences in chemical composition between these two AEO samples depending on the ore being processed at the time of collection, as has been recently reported (Frank et al., 2016). Therefore, slight differences in toxicity between extracts from the same pond collected at different times may be expected, however we submit that this was the best comparison available. 2.3. Analytical chemistry All chemical analysis of AEO extract and fractions including derivation of concentration, ion class distribution, and molecular weight was conducted using electrospray ionization highresolution mass spectrometry and described previously (Bauer et al., 2015). In brief, an LTQ Orbitrap Velos mass spectrometer (Thermo Fisher Scientific) was operated in full-scan negative ion mode with an m/z scan range of 100e600. Infusion solvent used was 50:50 acetonitrile:water containing 0.1% ammonium hydroxide at a 200 mL/min flow rate. All ions were present in the m/z 100 to m/z 300 range which corresponded to a resolution range of 240 000 to 150 000 and the mass accuracy was less than 2 ppm error. Molecular analysis was conducted using Xcalibur Ver 2.1 (Thermo Fisher Scientific) and Composer Ver 1.0.2 (Sierra Analytics) software. By multiplying each m/z value by its relative abundance and averaging the sum of m/z values (with incorporated abundance) for each fraction, mean molecular weight (Da) for each fraction was derived. For derivation of AEO extract and fraction concentrations, a predefined 5-point regression of OSPW-derived organic acids was used to determine resulting AEO concentrations. The concentration of the extracted AEO solution was 2500 mg/L, determined by ESI-MS (Headley et al., 2002) with comparison to an aliquot of NAs previously extracted from OSPW (Rogers et al., 2002) and analyzed by Fourier transform infrared spectroscopy (Jivraj et al., 1995). Stock solutions for bioassays were prepared from this original sample. AEO fraction aromaticity was derived using synchronous fluorescence spectroscopy and described previously (Kavanagh et al., 2009). Briefly, a Perkin-Elmer Luminescence spectrometer LS50B was operated at an excitation wavelength of 250e500 nm and a Dl of 18 nm. Monochromator slit width were 5 nm, scan speed was set to 50 nm/min with a resolution of 0.5 nm, and data were interpreted using FL Winlab 3® software (PerkinElmer). 2.4. Exposure conditions Embryos of fathead minnow (Pimephales promelas) and Japanese medaka (Oryzias latipes) were exposed separately to the five AEO fractions, as well as an AEO extract. Stock solutions were prepared by diluting the oil sands AEO extract and fractions to 100 mg/L (nominal) with 75% Milli-Q water (v/v) and 25% well water (v/v; referred to as dilution water) to emulate breeding culture conditions. Each stock solution was adjusted to pH 8.0 ± 0.2 to mimic conditions present in OSPW and natural lakes in the area

A.E. Bauer et al. / Chemosphere 171 (2017) 168e176

(Allen, 2008a), and stored in amber glass bottles at 4  C. Exposure treatments were prepared by mixing stock solutions with dilution water, and included 7 nominal concentrations (3, 5, 10, 15, 30, 50, 100 mg/L for fathead minnow; 1.5, 5, 10, 15, 30, 50, 100 mg/L for Japanese medaka), a water control, and a sodium hydroxide (0.05 M) solvent control (conversion to mM concentrations; Supplemental Data, Fig. S1). Nominal concentrations were used throughout the toxicity assessments as ESI-MS analyses of samples yielded much lower measured concentrations. This observed reduction in yield was likely due to matrix effects of well water used for dilutions, resulting in ion suppression (Taylor, 2005). Lower measured treatment concentrations may also have been due to loss of compounds via adherence to embryo. In previous analysis of AEO stock concentrations nominal concentrations deviated from measured concentrations by 6.0 ± 0.3% (data not shown), and therefore we believe that the use of nominal concentrations herein was acceptable. Exposures were prepared as mg/L concentrations and then converted to mM concentrations using each fraction's mean molar mass in order to provide comparison with characterization data (Bauer et al., 2015). Water quality measurements at 3 mg/L (fathead minnow) and 1.5 mg/L (Japanese medaka), 15 mg/L, and 100 mg/L were taken daily for temperature, dissolved oxygen, and conductivity, while pH was taken at all concentrations daily using an Orion model 1230 multi-meter (Orion Research Inc., Beverly, MA). Water quality was measured from 20-mL vials, prepared the previous day, before subdividing solutions into Falcon™ 48-well tissue culture plates for toxicity testing. All embryo toxicity testing parameters were conducted in accordance with OECD Report 212 (Fish, Shortterm Toxicity Test) (OECD, 1998), although test procedures varied slightly. Embryos were selected in the morning and tests were run when embryos entered late blastula stage and terminated at two days post-hatch (~10e11 days for Japanese medaka, ~ 6e8 days for fathead minnow). Ten embryos per concentration were placed in separate wells of tissue culture plates and exposed to a treatment volume of 0.5 mL/embryo. Treatments were run in triplicate and solutions were renewed daily. Tissue culture plates containing embryos were held in a growth chamber (Conviron®, Winnipeg, MB) with photoperiod maintained at 16 h light: 8 h dark, and temperature held at 25 ± 1  C. Previous work has found that gentle agitation of Japanese medaka eggs ensures earlier, more synchronized hatch times (8 days instead of 12 days) without affecting test results (Farwell et al., 2006). Therefore, tissue culture plates holding Japanese medaka embryos were additionally placed on an orbital shaker (Lab-Line Instruments Inc., Melrose Park, IL) set to 100 rpm while in the growth chamber. 2.5. Endpoint measurements Acute toxicity was evaluated for each treatment by recording mortality at two days post-hatch (>8 days for Japanese medaka and >6 days for fathead minnow), and sublethal toxicity was assessed by recording time-to-hatch, hatch length, and abnormalities. Embryos were assessed twice daily. For larvae that successfully hatched with no major spinal deformities, hatch length was recorded to 0.1 mm using a dissecting microscope and stage micrometer. Larval abnormalities investigated were based on a blue sac disease score developed previously (Rhodes et al., 2005) and included three main abnormality types: heart abnormalities (malformation, poor circulation, hemorrhaging, and clotting), cranial and skeletal deformities (jaw, cranial size/structure deformities, and spine curvature), and yolk sac abnormalities (yolk sac and pericardial edemas). Incidence of abnormalities was recorded to determine percent normal larvae for each treatment. Treatments were compared to both water and solvent controls,

171

with those that showed greater percentage abnormalities (p  0.05) relative to controls considered significant. Exposures were terminated at two days post-hatch in order to assess hatch length growth endpoints and morphological abnormalities. 2.6. Statistical analysis Quantal data (mortality) for each fraction was tested for normality using a Shapiro-Wilk's test while median lethal concentrations (LC50) and their confidence intervals (a ¼ 0.05) were determined using Probit analysis in IBM® SPSS® (version 20). Data that did not conform to Probit assumptions of normality or homogeneity were analyzed with a trimmed Spearman-Karber test in Microsoft® Excel (Microsoft® Office, 2010) at a 20% trim level (Hamilton et al., 1977). Significant differences between fraction LC50 values were determined by the Litchfield-Wilcoxon method using Microsoft® Excel (Microsoft® Office, 2010). Quantitative data (hatch length and time-to-hatch) for each fraction were analyzed using IBM® SPSS® version 20. Analysis for quantitative data was subject to hypothesis testing using one-way ANOVA followed by Tukey's post hoc tests to determine concentrations that differed significantly (p < 0.05) from controls. Tests that did not conform to assumptions of normality and homoscedasticity required for ANOVA, were subject to Kruskal-Wallis followed by pairwise Mann-Whitney-U tests. Both parametric and non-parametric analyses were conducted on concentrations within each treatment in order to derive no-observed-effects concentrations (NOEC) and lowest-observed-effects concentrations (LOEC). For comparison, LOEC values were reported. Incidences of abnormality for the derivation of percent normal larvae were analyzed in SPSS® using the same methods of hypothesis testing described above, with the exception that the Student's t-test was used in place of a one-way ANOVA. This was because of the small sample sizes present due to low overall incidences of abnormalities. 3. Results 3.1. Mortality Fathead minnow displayed LC50 values in the range of 72 mMe134 mM when exposed to the five AEO fractions (Table 2). The F5 extract was most toxic overall (60 mM; (confidence limits 49e73 mM)) but was not significantly different from F2. F1 displayed the lowest toxicity overall (134 mM; (104e170 mM)) and was significantly different from F2, F4 and F5. The AEO extract (LC50 ¼ 72 mM; (58e96 mM)) displayed toxicity similar to F2, F4, and F5. The LC50 values of Japanese medaka embryo exposed to the five AEO fractions ranged from 149 mMe291 mM (Table 2). The most toxic fractions to medaka were F2 (157 mM; (151e162 mM)) and F3 (149 mM; (147e152 mM)), which were significantly more toxic than the other fractions and significantly different from each other. Fraction F1, which had the lowest mean molecular weight, displayed an LC50 (291 mM; (280e302 mM)) that was the least toxic relative to the other fractions, and was nearly two-fold less toxic than the most toxic fraction (F3). When the AEO extract was compared to the fraction toxicities, the AEO extract displayed a greater toxicity (155 mM; (145e161 mM)) than any individual fraction (Table 2), and was statistically similar to only F2 and F3. 3.2. Sublethal endpoints Fathead minnow mean hatch lengths ranged from 3.8 ± 0.1 to 5.2 ± 0.04 mm (Supplemental Data, Fig. S2). F3 e F5 conferred

172

A.E. Bauer et al. / Chemosphere 171 (2017) 168e176

Table 2 Acute and sublethal toxic effects of AEO fractions. Fraction

Acute LC50a (mM)

Hatch Length NOEC (mM)

fathead minnow

Japanese medaka

AEO 1 2 3 4 5 AEO 1 2 3 4 5

72 (58e96) 134 (107e170) 71 (55e90) 104 (88e126) 92 (77e111) 59 (49e73) 155 (149e161) 291 (280e302) 157 (151e162) 149 (147e152) 209 (173e251) 192 (183e201)

c

nm >211 >208 58 32 42 nm >211 125 117 97 141

LOEC (mM)

Size reduction (mm)b

nm –d e 117 49 85 nm e 208 195 162 282

nm e e 0.4 1.4 0.6 nm e 0.5 0.3 0.4 0.6

NOEC ¼ no observed effect concentration, LOEC ¼ lowest observed effects concentration, AEO ¼ acid extractable organics. a This column displays LC50 mean ± confidence limits for fraction acute toxicities. b Column values describe reduction in size at the highest concentration compared to controls: (mean hatch length for controls) - (mean hatch length at highest concentration). c “nm” refers to data where no measurement was obtained. d “–” indicates no significant effect on hatch length, and therefore, no NOEC or LOEC was calculated.

significant reductions in hatch length (p  0.05) compared to controls, while F1 and F2 had no effect on hatch length at concentrations with <100% mortality. Because no LOEC could be derived for F1 and F2, effect concentrations could not be calculated. The LOEC for F3, F4, and F5 were 116.7 mM, 48.7 mM, and 84.5 mM, respectively (Table 2). At the highest concentrations tested (50 mg/L) F4 displayed the greatest inhibition of growth (average 1.4 mm reduction), followed by F5 (0.6 mm reduction) and F3 (0.4 mm reduction) when compared to controls (Table 2). Although lower molecular weight fractions (F1 and F2) displayed lower toxic potencies than higher molecular weight fractions (F3, F4, and F5), fathead minnow did not show a strong molecular weight to toxicity relationship for hatch length. Japanese medaka exposed to AEO fractions displayed mean hatch lengths ranging from 4.2 ± 0.05 to 4.9 ± 0.04 mm (Supplemental Data, Fig. S3). Only fractions F2 to F5 caused a significant reduction in larval hatch length (p  0.05), and only at the highest concentrations. Because of the low number of individuals at the highest concentration of each fraction, we were not confident in concluding that a trend was present. Fraction F5 had the greatest impact on hatch length (Table 2), reducing average length by 0.6 mm at the highest concentration (100 mg/L), followed by F2 (0.4 mm reduction at 50 mg/L). The significant effect (p < 0.05) observed in F5 was at the 100 mg/L concentration (282 mM) while F2-F4 displayed effects at the 50 mg/L concentration (208 mM, 195 mM, 162 mM, respectively). LOEC for medaka hatch length ranged from 162 to 282 mM for F2 e F5 (Table 2). The range in mean hatch time for fathead minnow was 4.2 ± 0.1 to 5.9 ± 0.3 days, while the range in mean hatch time for Japanese medaka was 7.6 ± 0.1 to 9.3 ± 0.2 days (Supplemental Data, Figs. S4 and S5). When fractions were compared for Japanese medaka, F5 appeared to have the greatest impact on hatch time by delaying hatch an average 1.3 days compared to controls, followed by F1 (1 day delay) and F4 (0.5 day delay). Although F5 displayed the greatest impact on hatch time this effect was only observed at 100 mg/L (282 mM) concentration whereas F1 and F4 displayed delays in hatch at the 50 mg/L concentration (211 mM and 162 mM, respectively). Similar to fathead minnow, there was no substantial data to allow for comparison between fractions. Overall, fractions only displayed significant delays in hatch time (p  0.05) relative to controls at the highest concentrations where number of individuals was lowest (due to higher mortalities), thus reducing confidence that dose-response trends were present.

3.3. Abnormalities Control fish displayed total incidence of abnormalities <7% for fathead minnow and Japanese medaka across all fractions. Fathead minnow only displayed abnormalities that were significantly different from controls for F4 at 30 mg/L and 50 mg/L concentrations (p ¼ 0.001), but all fractions appeared to display an increase in abnormalities at higher concentrations. Japanese medaka displayed abnormalities which were significantly higher than controls only at the highest concentration (100 mg/L, p ¼ 0.001) assessed for F5. Of the abnormality endpoints assessed, the most common in both species were yolk sac/pericardial edema (Fig. 1). 4. Discussion 4.1. Narcosis Investigation of the relationship between AEO molecular weight and toxicity, using the same fractions assessed herein, has reported that lower molecular weight AEO fractions were more toxic than high molecular weight fractions to Vibrio fischeri (Microtox®) (Frank et al., 2008). By comparing molecular weight and surfactant nature of AEO to Vibrio fischeri toxicity, polar narcosis was proposed as the predominant MOA for AEO (Frank et al., 2009, 2010). In a companion study to this current work (Bauer et al., 2015), the same five oil sand AEO fractions investigated by Frank et al. (2008) were further characterized to determine organic acid chemical composition associated with increasing molecular weight. This analysis revealed a relative increase in heteroatoms, carboxyl and hydroxyl groups, and overall increase in oxygenated functional groups associated with an increase in fraction mean molecular weight (Table 1, Supplemental Data, Figs. S6 and S7), which can lead to increased solubility of AEO compounds (Stanford et al., 2007). We, therefore, would not have been surprised to observe a similar decrease in toxicity with increasing molecular weight fraction, supporting a narcotic MOA as observed in the Frank et al. (2008) study (Frank et al., 2008). When fraction toxicities were compared, the Japanese medaka bioassay exhibited a response supportive of a narcosis MOA as the main toxic driver. For Japanese medaka F2 and F3 displayed the highest acute toxicity compared to other fractions. Conversely, F4 and F5 displayed significantly lower acute toxicity than F2 and F3. Much of the increase in molecular weight for F4 and F5 can be attributed to the relative increase in

A.E. Bauer et al. / Chemosphere 171 (2017) 168e176

Fig. 1. Percentage incidence of abnormality type for Japanese medaka (n ¼ 60) and fathead minnow (n ¼ 104). Data represents individuals pooled across all fractions and concentrations. Bars represent mean percentage incidence ± standard error.

heteroatom content (specifically oxygen-containing species) and greater degree of aromaticity (Supplemental Data, Figs. S6 and S7) than F2 and F3 (Bauer et al., 2015). Additionally, F4 and F5 contained doubly charged ions, which was likely due to the presence of dicarboxyl and dihydroxy groups (Bauer et al., 2015). There is evidence to suggest that additional oxygenated functional groups serve to increase the water solubility (hydrophilicity) of compounds present in crude oil (Stanford et al., 2007) and NA surrogates, and are associated with a reduction in toxicity (Frank et al., 2009, 2010). This evidence is also supported by a recent study which found that AEO fractions containing mono- and dioxygenated ion classes were more toxic than those fractions containing poly-oxygenated classes (Morandi et al., 2015). Similar to research relating increased carboxylic acid content to reductions in toxicity to the same fractions tested in the present study (Supplemental Data, Figs. S8 and S9) (Frank et al., 2009), an increase in oxygenated groups and aromaticity likely increase aqueous solubility and resulted in the observed decrease in toxicity for F4 and F5, when compared with F2 and F3. The relatively low toxicity observed in F1 is plausible if we consider that the relative solubility of an acidic compound imparted by its hydrophilic carboxyl functional group is much greater for a smaller compound than a larger one. Thus, an overall relative reduction in the compound's hydrophobic surface area among similar compounds results in relatively higher solubilities and lower toxicities. For fathead minnow, only the toxicities of F1 and F2 were similar to those seen in Japanese medaka, and appeared to be driven by narcosis. The respective toxic potencies of F1 and F2 were likely due to the same chemical properties described for Japanese medaka.

4.2. Additional MOAs Only some of the toxicity observed for both fish species could be explained by a narcotic MOA, and overall toxicity trends were not comparable to a similar study exposing the same fractions to V. fischeri (Frank et al., 2008). Because previous studies have proposed that NA toxicity to fish may be directed by MOA in addition to narcosis (Tollefson et al., 2012; Scarlett et al., 2013; Swigert et al., 2015), and due to the fact that all fractions were themselves complex mixtures of varying complexity, examination of additional MOAs was merited. When fraction exposure to fathead minnow was compared, acute toxicity exhibited a trend which was dissimilar to that observed for Japanese medaka. Specifically, F4 and F5 were more toxic to fathead minnow than Japanese medaka. It is possible that fathead minnow are more sensitive than Japanese medaka because of a different, additional MOA. The proposal of an additional MOA is

173

in agreement with recent literature regarding fathead minnow exposed to commercial NA (Swigert et al., 2015) and organic fractions of OSPW (Morandi et al., 2015). An electrophilic reactivity MOA may explain some of the high toxicity observed for fathead minnow in F4 and F5. Electrophilic toxicants are electron-poor compounds that bond covalently with electron-rich nucleophilic sites on biological macromolecules such as proteins, DNA, peptides, €bel et al., 2011). The result is the direct addition, and lipids (Schwo conjugation, or substitution of the toxicant with the heteroatomic moiety of the endogenous molecule in an untargeted manner. This has been shown to cause mutagenicity and carcinogenicity, hepatotoxicity, and elevated acute toxicity above narcosis through the covalent reaction with fish gill membranes (Von der Ohe et al., 2005). Recent research on OSPW and NA has found that increased presence of aromatic rings and DBE coincides with an increase in toxicity to exposed fish (Scarlett et al., 2013; Morandi et al., 2015). The increased toxicity associated with aromatic NAs has been linked to their electrophilic reactivity (using values for energy of the lowest unoccupied molecular orbital; ELUMO) (Scarlett et al., 2013), wherein compounds with lower ELUMO values (i.e., increased aromatics) exhibit increased electrophilic reactivity and toxicity (Schultz, 1989). Therefore, it is likely that the greater degree of aromatic compounds present in F4 and F5 (Supplemental Data, Fig. S6) contribute to their higher toxicities in fathead minnow, due to an increase in electrophilic reactivity. However, this relationship is simply speculative because quantitative validation of ELUMO values was not calculated herein. It is also probable that additional MOA include oxidative stress, which may account for some of the high toxicity observed in fathead minnow exposures. This theory is supported by research reporting that fathead minnow embryo exposed to OSPW displayed greater reactive oxygen species (ROS) (He et al., 2012), as well as higher abundances of transcripts responsible for the removal of ROS (Wiseman et al., 2013), when compared to controls. Oxidative stress is the result of an imbalance between exposure to ROS and an organism's ability to detoxify them with endogenous antioxidants. ROS are highly reactive, oxygen-centered radicals that contain an unpaired electron, allowing it to remove electrons from other molecules and generate similar chain reactions. Unmitigated oxidative stress can result in damage to DNA, carbohydrates, proteins, lipids, and ultimately cell and tissue damage. Although the presence of ROS in each fraction is unknown, we propose that they would be most likely present in F5 because it contains the highest abundance of oxygen-containing species (Table 1, Supplemental Data, Fig. S7). With respect to Japanese medaka exposures, it is possible that the relatively high toxicities observed for F2 and F3 were due to contribution from an additional MOA elicited by compounds therein. Although F2 and F3 exhibited an expected higher toxicity compared to F1, F4 and F5, the similarity in toxic potency between the two was not expected. These findings were surprising because F2 and F3 contained compounds which had quite dissimilar DBE and aromatic profiles indicating potentially different degrees of ringed structures, aromaticity and functional groups (Bauer et al., 2015). Scarlett et al. (2013) observed that alicyclic NA displayed lesser toxicities to zebrafish than polycyclic monoaromatic NA. This insight is supported in the present study, where we observed a parallel between the lowest aromatic content and the lowest toxicity in F1. Moreover, monoaromatic content markedly increased successively in F2 and F3 (Bauer et al., 2015), as did toxicity. As such, specific aromatic compounds may be important in assessing toxic mode of action, and is perhaps explained by recent discoveries identifying structural similarities between NA compounds and the drug Ibuprofen (Tollefson et al., 2012). Ibuprofen, a monoaromatic, non-steroidal anti-inflammatory drug (DBE ¼ 5), has been shown to cause sublethal toxicity in aquatic invertebrates and fish (Gravel

174

A.E. Bauer et al. / Chemosphere 171 (2017) 168e176

and Vijayan, 2006; Heckmann et al., 2007). Similar to mammals, research suggests a MOA in aquatic organisms which interferes with prostaglandin synthesis by competitive inhibition of cyclooxygenase enzymes (Heckmann et al., 2008). Because F1 has a maximum DBE of 7 in low proportion, these compounds are likely not in appreciable quantities, and are likely present in greater quantities in F2-F4 (Table 1). These data imply that a specific subset of AEO may induce greater acute toxicities when at elevated proportions within a fraction or extract, and also illustrate the possibility that components within AEO may exert a MOA similar to ibuprofen. 4.3. Abnormalities Previous studies that exposed fish to natural bitumen sources, process-affected sediments, and OSPW have reported similar types of abnormalities but at higher incidence rates than the present study (Colavecchia et al., 2004, 2007; Peters et al., 2007). However, unlike these previous studies, the AEO extraction employed in the present study was designed to isolate only polar organics and exclude contaminants such as polycyclic aromatic hydrocarbons (PAH) which have been shown to elicit high incidence rates of abnormalities in Japanese medaka (Farwell et al., 2006; Rhodes et al., 2005). Previous analysis has verified that the AEO extract contains less than 0.3 mg/L of the 16 priority PAHs (Kavanagh et al., 2012) as defined by the United States Environmental Protection Agency. The incidences of abnormalities to fathead minnow reported here more closely resemble research conducted by Marentette et al. (2015) where exposure to an AEO extract displayed a dose-dependent increase in abnormalities resulting in >60% incidence of total abnormality at 25 mg/L AEO. In the present study, at concentrations above 30 mg/L, an average of 19e62% of fathead minnow and 4e57% of medaka displayed some form of abnormality across all fractions. The present study established that edema was the most common abnormality (Fig. 1), resulting from the grouping of yolk sac and pericardial edema. Marentette et al. (2015) also observed that the majority of abnormalities caused by oil sands AEO manifest as cardiovascular abnormalities, which included edemas (Marentette et al., 2015). It is difficult to ascertain which fractions significantly contributed to increased abnormality rates because there was a great deal of variability between replicates, and at higher concentrations the number of surviving fish was very low. Further experimentation is suggested in order to assess abnormalities across fractions. 4.4. Species comparison In general, the toxicities between fractions for both fish species tested did not differ greatly and were within a two-fold separation. Overall, fathead minnow appeared to be more sensitive to AEO exposure than Japanese medaka for both acute and subacute toxicity (Table 2). Disparities observed for toxicities between fathead minnow and Japanese medaka may be related to differences in embryo chorion permeability. The Japanese medaka chorion has many filamentous protrusions at ~23 mm intervals which are 130e140 mm in diameter (Iwamatsu, 1992). In contrast, the chorion surface of the fathead minnow possesses evenly distributed pores ~0.2 mm in diameter and ~2 mm apart (Lillicrap, 2010), which could allow for greater permeability of AEO compounds. Another possibility for greater sensitivity of fathead minnow is the differences in developmental regimes between species. Japanese medaka complete organogenesis prior to hatch (Villalobos et al., 2000), whereas fathead minnow complete midfinal stages of organogenesis up to 4 days post-hatch (Scudder et al., 1988). Although both species are capable of embryonic

detoxification of select xenobiotics (Colavecchia et al., 2007; Jovanovic et al., 2011; Lindstrom-Seppa et al., 1994; Wisk and Cooper, 1992; Wu et al., 2011), it is quite possible that underdeveloped organs in fathead minnow could impact the embryonic metabolism of AEO. It is also possible that the natural variation in incubation duration between the two species accounts for some of the observed disparity in toxic responses. The range in mean hatch time for Japanese medaka was 7.6e9.3 days while the range in mean hatch time for fathead minnow was 4.2e5.9 days, reflecting natural variations in hatch times between the two species (Supplemental Data, Figs. S4 and S5). Finally, the recommended exposure temperature for Japanese medaka is 26e30  C while exposures herein were kept at 25 ± 1  C. It is possible that the lower temperatures resulted in a lower sensitivity of Japanese medaka, and that toxicities would be more similar to those observed for fathead minnow at greater temperatures. A comparison of the two species with respect to abnormalities shows that fathead minnow display greater variability for abnormality rates. Nonetheless, for both species incidences increased slightly with increasing concentration, and the percentage pooled abnormality types were very similar between species (Fig. 1). This observation indicates that the mechanisms affecting developmental abnormalities from AEO exposure are likely the same. 5. Conclusions The present research suggests that, regarding toxicity assessments of the subfractions generated from an OSPW extract, toxicity to AEO does not appear to be related solely to narcosis. Additional MOA may manifest via oxidative stress, electrophilic reactivity, and enzyme inhibition. Although there were differences in the relative toxicities of fractions between species, within individual fish species fraction toxicities were very similar, and in many cases fraction toxicities were not significantly different (p  0.05, Table 2). The toxicity trends observed in response to exposures to AEO fractions differed from those observed using the Microtox® assay in a previous study (Frank et al., 2008). Even within the present study, a common trend was not observed between fish species, demonstrating the importance of incorporating a complement of aquatic species in order to accurately assess the toxicity of highly complex organic mixtures in OSPW. Further complementary analyses are required to support future AEO characterization and understanding of associated toxicity, as this study was based solely on one mode of detection. The focus of the present investigation was analyzing the AEO component (namely the carboxylic acids) of isolated OSPW fractions, for which ESI-MS in negative ion mode is the best method. However recent work has brought attention to the fact that other soluble organics that likely contribute to OSPW toxicity are best detected in positive mode (Morandi et al., 2015; Pereira et al., 2013). The authors acknowledge that other components could be contributing to toxicity which we did not detect and which is the focus of future investigation. Moreover, this research advocates for supplementary evaluation of the mode of toxic action that the various compounds within AEO exert. Disclaimer Due to the lack of representative analytical standard methodologies for the analysis of total NAs or AEO, the results presented herein should be considered internally consistent, and may not be directly comparable to other results. For further information on the methods and analytical procedures used in this study, please contact the corresponding author.

A.E. Bauer et al. / Chemosphere 171 (2017) 168e176

Acknowledgments Financial support for the present research was provided by the Natural Sciences and Engineering Research Council Discovery Grant (NSERC RGPIN - 8155) to D. G. Dixon. We thank K. Oakes, M. Servos, and R. Hall from the Biology Department at the University of Waterloo for their valuable guidance. We thank M. Ryan and C. Futher for animal care and wetlab assistance during the maintenance of fish cultures at University of Waterloo. Many thanks to L. Bauer and A. Bauer from Caledon Aquatics for all of their fish culturing assistance. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.chemosphere.2016.12.059. References Alberta Energy, 2008. Alberta Oil Sands: Resourceful. Responsible. (ISBN 97807785-7348-7). Alberta Energy and Utilities Board, 2014. Energy Ministry of Energy Annual Report. Allen, E.W., 2008. Process water treatment in Canada's oil sands industry: I. Target pollutants and treatment objectives. J. Environ. Eng. Sci. 7, 123e138. Allen, E.W., 2008. Process water treatment in Canada's oil sands industry: I. Target pollutants and treatment objectives. J. Environ. Eng. Sci. 7, 123e138. Barrow, M.P., Witt, M., Headley, J.V., Peru, K.M., 2010. Athabasca oil sands process water: characterization by atmospheric pressure photoionization and electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry. Anal. Chem. 82, 3727e3735. Bauer, A.E., Frank, R.A., Headley, J.V., Peru, K.M., Hewitt, L.M., Dixon, D.G., 2015. Enhanced characterization of oil sands acid-extractable organics fractions using electrospray ionization-high-resolution mass spectrometry and synchronous fluorescence spectroscopy. Environ. Toxicol. Chem. 34, 1001e1008. Colavecchia, M.V., Backus, S.M., Hodson, P.V., Parrott, J.L., 2004. Toxicity of oil sands to early life stages of fathead minnows (Pimphales Promelas). Environ. Toxicol. Chem. 23, 1709e1718. Colavecchia, M.V., Hodson, P.V., Parrott, J.L., 2007. The relationships among CYP1A induction, toxicity, and eye pathology in early life stages of fish exposed to oil sands. J. Toxicol. Environ. Health, Part A 70, 1542e1555. Del Rio, L.F., Hadwin, A.K.M., Pinto, L.J., MacKinnon, M.D., Moore, M.M., 2006. Degradation of naphthenic acids by sediment micro-organisms. J. Appl. Microbiol. 101, 1049e1061. Energy Resources Conservation Board, 2014. ST98-2014 Alberta's Energy Reserves 2013 and Supply/Demand Outlook 2014e2023. Environment Canada, 2011. EPS 1/RM/22. Biological Test Method: Test of Larval Growth and Survival Using Fathead Minnows. EPS 1/RM/22. Farwell, A., Nero, V., Croft, M., Bal, P., Dixon, D.G., 2006. Modified japanese medaka embryo-larval bioassay for rapid determination of developmental abnormalities. Arch. Environ. Contam. Toxicol. 51, 600e607. Frank, R.A., Kavanagh, R., Burnison, B.K., Arsenault, G., Headley, J.V., Peru, K.M., Solomon, K.R., 2008. Toxicity assessment of generated fractions from an extracted naphthenic acid mixture. Chemosphere 72, 1309e1314. Frank, R.A., Fischer, K., Kavanagh, R., Burnison, B.K., Arsenault, G., Headley, J.V., Peru, K.M., Van Der Kraak, G., Solomon, K.R., 2009. Effect of carboxylic acid content on the acute toxicity of oil sands naphthenic acids. Environ. Sci. Technol. 43, 266e271. Frank, R.A., Sanderson, H., Kavanagh, R., Burnison, B.K., Headley, J.V., Solomon, K.R., 2010. Use of a (Q)SAR model to predict the toxicity of naphthenic acids. J. Toxicol. Environ. Health, Part A 73, 319e329. Frank, R.A., Milestone, C.M., Rowland, S.J., Headley, J.V., Kavanagh, R.J., Lengger, S.K., Scarlett, A.G., West, C.E., Peru, K.M., Hewitt, L.M., 2016. Assessing spatial and temporal variability of acid-extractable organics in oil sands process-affected waters. Chemosphere 160, 303e313. Gravel, A., Vijayan, M.M., 2006. Salicylate disrupts interrenal steroidogenesis and brain glucocorticoid receptor expression in rainbow trout. Toxocol. Sci. 93, 41e49. Grewer, D.M., Young, R.F., Whittal, R.M., Fedorak, P.M., 2010. Naphthenic acids and other acid-extractables in water samples from Alberta: what is being measured? Sci. Total Environ. 408, 5997e6010. Hamilton, M.A., Russo, R.C., Thurston, R.V., 1977. Trimmed Spearman-Karber method for estimating median lethal concentrations in toxicity bioassays. Environ. Sci. Technol. 11, 714e719. Han, X., Scott, A.C., Fedorak, P.M., Bataineh, M., Martin, J.W., 2008. Influence of molecular structure on the biodegradability of naphthenic acids. Environ. Sci. Technol. 42, 1290e1295. He, Y., Patterson, S., Wang, N., Hecker, M., Martin, J.W., Gamal El-Din, M., Giesy, J.P., Wiseman, S.B., 2012. Toxicity of untreated and ozone-treated oil sands processaffected water (OSPW) to early life stages of the fathead minnow (Pimephales

175

promelas). Water Res. 46, 6359e6368. Headley, J.V., Peru, K.M., McMartin, D.W., Winkler, M., 2002. Determination of dissolved naphthenic acids in natural waters by using negative-ion electrospray mass spectrometry. J. AOAC Int. 85, 182e187. Headley, J.V., Peru, K.M., Armstrong, S.A., Han, X., Martin, J.W., Mapolelo, M.M., Smith, D.F., Rogers, R.P., Marshall, A.G., 2009. Aquatic plant-derived changes in oil sands naphthenic acid signatures determined by low-, high-, and ultrahighresolution mass spectrometry. Rapid Commun. Mass Spectrom. 23, 515e522. Headley, J.V., Peru, K.M., Mohamed, M.H., Frank, R.A., Martin, J.W., Hazewinkle, R.R.O., Humphries, D., Gurprasad, N.P., Hewitt, L.M., Muir, D.C.G., Lindeman, D., Strub, R., Young, R.F., Grewer, D.M., Whittal, R.M., Fedorak, P.M., Birkholz, D.A., Hindle, R., Reisdorph, R., Wang, X., Kasperski, K.L., Hamilton, C., Woudneh, M., Wang, G., Loescher, B., Farwell, A., Dixon, D.G., Ross, M.S., Dos Santos Pereira, A., King, E., Barrow, M.P., Fahlman, B., Bailey, J., McMartin, D.W., Borchers, C.H., Ryan, C.H., Toor, N.S., Gillis, H.M., Zuin, L., Bickerton, G., McMaster, M.E., Sverko, E., Shang, D., Wilson, L.D., Wrona, F.J., 2013. Chemical fingerprinting of oil sands naphthenic acids in environmental samples - a review of analytical methods. J. Environ. Sci. Health A 48, 1145e1163. Heckmann, L.H., Callaghan, A., Hooper, H.L., Connon, R., Hutchinson, T.H., Maund, S.J., Sibly, R.M., 2007. Chronic toxicity of ibuprofen to Daphnia magna: effects on life history traits and population dynamics. Toxicol. Lett. 172, 137e145. Heckmann, L., Sibly, R.M., Connon, R., Hooper, H.L., Hutchinson, T.H., Maund, S.J., Hill, C.J., Bouetard, A., Callaghan, A., 2008. Systems biology meets stress ecology: linking molecular and organismal stress responses in Daphnia magna. Genome Biol. 9, R40. Iwamatsu, T., 1992. Morphology of filaments on the chorion of oocytes and eggs in the medaka (developmental biology). Zool. Sci. 9, 589e599. Jivraj, M.N., MacKinnon, M.D., Fung, B., 1995. Naphthenic Acids Extraction and Quantitative Analyses with FTIR Spectroscopy. Syncrude Analytical Methods Manual, fourth ed. Jones, D., Scarlett, A.G., West, C.E., Rowland, S.J., 2011. Toxicity of individual naphthenic acids to Vibrio fischeri. Environ. Sci. Technol. 45, 9776e9782. Jovanovic, B., Anastasova, L., Rowe, E.W., Zhang, Y., Clapp, A.R., Palic, D., 2011. Effects of nanosized titanium dioxide on innate immune system of fathead minnow (Pimephales promelas Rafinesque, 1820). Ecotoxicol. Environ. Saf. 74, 675e683. Kavanagh, R., Burnison, B.K., Frank, R.A., Solomon, K.R., Van Der Kraak, G., 2009. Detecting oil sands process-affected waters in the Alberta oil sands region using synchronous fluorescence spectroscopy. Chemosphere 76, 120e126. Kavanagh, R.J., Frank, R.A., Burnison, B.K., Young, R.F., Fedorak, P.M., Solomon, K.R., Van Der Kraak, G., 2012. Fathead minnow (Pimephales promelas) reproduction is impaired when exposed to a naphthenic acid extract. Aquat. Toxicol. 116e117, 34e42. Lai, J.W.S., Pinto, L.J., Kiehlmann, E., Bendell-Young, L.I., Moore, M.M., 1996. Factors that affect the degradation of naphthenic acids in oil sands wastewater by indigenous microbial communities. Environ. Toxicol. Chem. 15, 1482e1491. Lillicrap, A.D., 2010. The Use of Zebrafish Embryos as an Alternative Approach for Ecotoxicity Testing. University of Exeter, Exeter. Lindstrom-Seppa, P., Korytko, P.J., Hahn, M.E., Stegeman, J.J., 1994. Uptake of waterborne 3,3',4,4'-tetrachlorobiphenyl and organ and cell-specific induction of cytrochrome P4501A in adult and larval fathead minnow Pimephlaes promelas. Aquat. Toxicol. 28, 147e167. Madill, R.E.A., Orzechowski, M.T., Chen, G., Brownlee, B.G., Bunce, N.J., 2001. Preliminary risk assessment of the wet landscape option for reclamation of oil sands mine tailings: bioassays with mature fine tailing pore water. Environ. Toxicol. 16, 197e208. Marentette, J.R., Frank, R.A., Bartlett, A.J., Gillis, P.L., Hewitt, L.M., Peru, K.M., Headley, J.V., Brunswick, P., Shang, D., Parrott, J.L., 2015. Toxicity of naphthenic acid fraction components extracted from fresh and aged oil sands processaffected waters, and commercial naphthenic acid mixtures, to fathead minnow (Pimephales promelas) embryos. Aquat. Toxicol. 164, 108e117. Morandi, G.D., Wiseman, S.B., Pereira, A., Mankidy, R., Gault, I.G.M., Martin, J.W., Giesy, J.P., 2015. Effects-directed analysis of dissolved organic compounds in oil sands process-affected water. Environ. Sci. Technol. 49, 12395e12404. Nero, V., Farwell, A., Lee, L.E.J., Van Meer, T., MacKinnon, M.D., Dixon, D.G., 2006. The effects of salinity on naphthenic acid toxicity to yellow perch: gill and liver histopathology. Ecotoxicol. Environ. Saf. 65, 252e264. OECD, 1998. OECD Guidelines for the Testing of Chemicals [212]: Fish, Short-term Toxicity Test on Embryo and Sac-fry Stages. Pereira, A.S., Bhattacharjee, S., Martin, J.W., 2013. Characterization of oil sands process-affected waters by liquid chromatography orbitrap mass spectrometry. Environ. Sci. Technol. 47, 5504e5513. Peters, L.E., MacKinnon, M.D., Van Meer, T., van den Heuvel, M.R., Dixon, D.G., 2007. Effects of oil sands process-affected waters and naphthenic acids on yellow perch (Perca flavescens) and Japanese medaka (Orizias latipes) embryonic development. Chemosphere 67, 2177e2183. Rhodes, S., Farwell, A., Hewitt, L.M., MacKinnon, M.D., Dixon, D.G., 2005. The effects of dimethylated and alkylated polycyclic aromatic hydrocarbons on the embryonic development of the Japanese medaka. Ecotoxicol. Environ. Saf. 60, 247e258. Rogers, V.V., Liber, K., MacKinnon, M.D., 2002. Isolation and characterization of naphthenic acids from Athabasca oil sands tailings pond water. Chemosphere 48, 519e527. Rowland, S.J., Scarlett, A.G., Jones, D., West, C.E., Frank, R.A., 2011. Diamonds in the rough: identification of individual naphthenic acids in oil sands process water.

176

A.E. Bauer et al. / Chemosphere 171 (2017) 168e176

Environ. Sci. Technol. 45, 3154e3159. Scarlett, A.G., Reinardy, H.C., Henry, T.B., West, C.E., Frank, R.A., Hewitt, L.M., Rowland, S.J., 2013. Acute toxicity of aromatic and non-aromatic fractions of naphthenic acids extracted from oil sands process-affected water to larval zebrafish. Chemosphere 93, 415e420. Schultz, T.W., 1989. Nonpolar narcosis: a review of the mechanism of action for baseline aquatic toxicity. ASTM STP 1027 Aquat. Toxicol. Hazard Assess. 12, 104e109. Philadephia, PA. €bel, J.A.H., Koleva, Y.K., Enoch, S.J., Bajot, F., Hewitt, M., Madden, J.C., Schwo Roberts, D.W., Schultz, T.W., Cronin, M.T.D., 2011. Measurement and estimation of electrophilic reactivity for predictive toxicology. Chem. Rev. 111, 2562e2596. Scudder, B.C., Carter, J.L., Leland, H.V., 1988. Effects of copper on development of the fathead minnow, Pimephales promelas Rafinesque. Aquat. Toxicol. 12, 107e124. Stanford, L.A., Kim, S., Klein, G.C., Smith, D.F., Rodgers, R.P., Marshall, A.G., 2007. Identification of water-soluble heavy crude oil organic-acids, bases, and neutrals by electrospray ionization and field desorption ionization Fourier transform ion cyclotron resonance mass spectrometry. Environ. Sci. Technol. 41, 2696e2702. Swigert, J.P., Lee, C., Wong, B.C.L., White, R., Scarlett, A.G., West, C.E., Rowland, S.J., 2015. Aquatic hazard assessment of a commercial sample of naphthenic acids. Chemosphere 124, 1e9. Taylor, P.J., 2005. Matrix effects: the achilles heel of quantitative high-performance liquid chromatographyeelectrosprayetandem mass spectrometry. Clin. Biochem. 38, 328e334. The Royal Society of Canada, 2010. Environmental and Health Impacts of Canada's

Oil Sands Industry. The Royal Society of Canada, Ottawa, ON. Tollefson, K.E., Petersen, K., Rowland, S.J., 2012. Toxicity of naphthenic acids and mixtures of these to fish liver cells. Environ. Sci. Technol. 46. USEPA, 1991. Guidelines for Culturing the Japanese Medaka, Oryzias latipes. EPA 600/3-91/064. United States Environmental Protection Agency, Washington, DC. Villalobos, S.A., Hamm, J.T., Teh, S.J., Hinton, D.E., 2000. Thiobencarb-induced embryotoxicity in medaka (Oryzias latipes): stage-specific toxicity and the protective role of chorion. Aquat. Toxicol. 48, 309e326. Von der Ohe, P.C., Kühne, R., Ebert, R., Altenburger, R., Liess, M., Schüürmann, G., 2005. Structural Alerts - a new classification model to discriminate excess toxicity from narcotic effect levels of organic compounds in the acute daphnid assay. Chem. Res. Toxicol. 18, 536e555. Wiseman, S., He, Y., Gamal El-Din, M., Martin, J.W., Jones, P.D., Hecker, M., Giesy, J.P., 2013. Transcriptional responses of male fathead minnows exposed to oil sands process-affected water. Comp. Biochem. Physiol. Part C 157. Wisk, J.D., Cooper, K.R., 1992. Effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin on benzo(a)pyrene hydroxylase activity in bmeryos of the Japanese medaka (Oryzias latipes). Arch. Toxicol. 66, 245e249. Wu, M., Shariat-Madar, B., Haron, M.H., Wu, M., Khan, I.A., Dasmanhapatra, A.K., 2011. Ethanol-induced attenuation of oxidative stress is unable to alter mRNA expression pattern of catalase, glutathine reductase, glutathione-S-transferase (GST1A), and superoxide dismutase (SOD3) enzymes in Japanese rice fish (Oryzias latipes) embryogenesis. Comp. Biochem. Physiol. Part C 153, 159e167.