Effects of preparation on nutrient and environmental contaminant levels in Arctic beluga whale (Delphinapterus leucas) traditional foods

Matthew J. Binnington a, Ying D. Lei a, Lucky Pokiak b, James Pokiak b, Sonja K. Ostertag c, Lisa L. Loseto c, Hing M. Chan d, Leo W. Y. Yeung e, Haiyong Huang a and Frank Wania *a
aDepartment of Physical and Environmental Sciences, University of Toronto Scarborough, 1265 Military Trail, Toronto, Ontario, Canada M1C 1A4. E-mail: frank.wania@utoronto.ca; Tel: +1 416 287 7225
bTuktoyaktuk Hunters and Trappers Committee, P. O. Box 286, Tuktoyaktuk, Northwest Territories, Canada X0E 1C0
cFreshwater Institute, Fisheries and Oceans Canada, 501 University Crescent, Winnipeg, Manitoba, Canada R3T 2N6
dDepartment of Biology, University of Ottawa, 30 Marie-Curie Private, Ottawa, Ontario, Canada K1N 6N5
eDepartment of Chemistry, University of Toronto, 80 St. George Street, Toronto, Ontario, Canada M5S 3H6

Received 4th April 2017 , Accepted 19th July 2017

First published on 19th July 2017

For Canadian Arctic indigenous populations, marine mammal (MM) traditional foods (TFs) represent sources of both important nutrients and hazardous environmental contaminants. Food preparation is known to impact the nutrient and environmental contaminant content of processed items, yet the impacts of preparation on indigenous Arctic MM TFs remain poorly characterized. In order to determine how the various processes involved in preparing beluga blubber TFs affect their levels of nutrients and environmental contaminants, we collected blubber samples from 2 male beluga whales, aged 24 and 37 years, captured during the 2014 summer hunting season in Tuktoyaktuk, Northwest Territories, and processed them according to local TF preparation methods. We measured the levels of select nutrients [selenium (Se), polyunsaturated fatty acids (PUFAs)] and contaminants [organochlorine pesticides, perfluoroalkyl and polyfluoroalkyl substances (PFASs), polybrominated diphenyl ethers, polychlorinated biphenyls, polycyclic aromatic hydrocarbons (PAHs), mercury (Hg)] in raw and prepared (boiled, roasted, aged) beluga blubber TFs. The impacts of beluga blubber TF preparation methods on nutrient and environmental contaminant levels were inconsistent, as the majority of processes either did not appear to influence concentrations or affected the two belugas differently. However, roasting and ageing beluga blubber consistently impacted certain compounds: roasting blubber increased concentrations of hydrophilic substances (Se and certain PFASs) through solvent depletion and deposited PAHs from cookfire smoke. The solid–liquid phase separation involved in ageing blubber depleted hydrophilic elements (Se, Hg) and some ionogenic PFASs from the lipid-rich liquid oil phase, while PUFA levels appeared to increase, and hydrophobic persistent organic pollutants were retained. Ageing blubber adjacent to in-use smokehouses also resulted in considerable PAH deposition to processed samples. Our findings demonstrated that contaminant concentration differences were greater between the two sets of whale samples, based on age differences, than they were within each set of whale samples, due to variable preparation methods. When considering means to minimize human contaminant exposure while maximizing nutrient intake, consumption of aged liquid from younger male whales would be preferred, based on possible PUFA enhancement and selective depletion of hydrophilic environmental contaminants in this food item.

Environmental significance

This is the “Arctic dilemma”: indigenous Arctic populations gain tremendous nutritional and cultural benefits from eating traditional foods derived from marine mammals, while, at the same time, these food items are a significant source of exposure to persistent organic pollutants. We explored how traditional preparation techniques for beluga whale blubber impact its nutritional value and contaminant load. Because of their largely trivial effects on nutrient contents and contaminant burdens, preparation methods are limited as a means to increase nutrient intakes and mitigate contaminant exposures. Instead, it may be advisable to have vulnerable subpopulation eat beluga blubber food items from younger male animals.

1. Introduction

Traditional food (TF) consumption by indigenous populations throughout the Canadian Arctic remains a critical practice in ensuring relationally, culturally, spiritually, physically, and financially healthy communities.1,2 Particularly, marine mammal (MM) TFs are an important source of several nutrients such as selenium (Se) and omega-3 polyunsaturated fatty acids (ω-3 PUFAs),3,4 especially when compared to many non-nutrient dense, imported alternatives. These food items also represent a primary exposure source for persistent organic pollutants (POPs) based on the lipid content, age, and trophic level of many of the species consumed (e.g., beluga whale, bowhead whale, narwhal, polar bear, ringed seal).3–7 To better inform indigenous Arctic individuals on the respective benefits and costs of continued TF consumption, numerous literature sources have attempted to quantify the levels of key nutrients (Se, ω-3 PUFAs) and environmental contaminants in popular TFs. However, with few exceptions3,4,8–10 these studies typically report nutrient and pollutant levels in raw tissues only, and do not account for the impacts of food preparation processes on ultimate human intakes.

This remains an important oversight, as food preparation techniques regularly impact the nutrient and contaminant content of cooked items when compared to raw tissues.11–14 For environmental contaminants, these analyses have been most frequently performed in fish items,15–18 yet the effects of food preparation are often inconsistent. In most studies cooking caused POP concentrations in fish tissue to decrease.19–21 For example, Bayen et al.15 observed 13–48% decreases in organochlorine pesticide (OCP), polychlorinated biphenyl (PCB), and polybrominated diphenyl ether (PBDE) levels from pan-frying, microwave cooking, boiling, or baking salmon steaks, while Schecter et al.18 measured 5–54% reductions in total PBDE concentrations following broiling of catfish, rainbow trout, and salmon. However, several other studies have described increases in POP concentrations, no changes to food item levels, or variable effects that were dependent on food item and cooking method.12,13,22,23 For example, Rawn et al.17 observed varying effects of baking, boiling, and frying finfish on PCB and polychlorinated dibenzo-p-dioxin/furan (PCDD/F) mass changes, ranging from decreases of up to 50% for PCDD/Fs in boiled grouper to increases of up to 154% for PCDD/Fs in baked mackerel.

In addition to POPs, mercury (Hg) also represents an important environmental contaminant readily found in MM TF items,24,25 and the compound chiefly responsible for TF consumption advisories.26–28 Most studies examining the influence of food preparation on Hg levels also analyzed fish items. Generally, these investigations have noted appreciable increases to wet-weight Hg concentrations (+45–75%) in fried, baked, boiled, and smoked fillets when compared to raw samples.29–31 Typically, absolute Hg contents in fish muscle do not change via the cooking process; rather water and/or fat loss increases Hg concentrations through a solvent depletion phenomenon. This same process was also recently documented in indigenous Arctic TFs, wherein Lemire et al.9 determined that air-drying of beluga meat was found to increase its Hg concentrations by 2–3 fold due to near-complete water loss. This method to produce beluga nikku (dried meat) was also noteworthy as beluga meat represented the most important contributor to Hg intake for indigenous Arctic women of childbearing age (WCBA) from Nunavik, Québec who participated in the Lemire et al.9 study.

With respect to important nutrients derived from MM TFs, Se is a key compound sourced mainly from marine food items including fish32,33 and MMs.4 The influence of preparation on Se bioavailability in fish items has varied in the literature, with certain species- and cooking method-specific reports of decreasing Se levels34,35 increasing trends,36 as well as no statistically significant changes.33,37 Like Hg, increasing wet-weight Se concentrations are hypothesized to occur via water depletion during cooking,36 while decreasing levels are expected to be due to loss of protein-bound or aqueous Se fractions in muscle tissue.34

Finally, ω-3 PUFAs represent additional nutrients found at high levels in MM TFs, with a similarly varied range of published accounts of cooking impacts. Previous studies describing a wide range of fish species and variable preparation methods have identified significantly decreased levels of docosahexaenoic acid (DHA, 22:6n − 3) and eicosapentaenoic acid (EPA, 20:5n − 3) due to frying,38–43 while others attribute ω-3 PUFA level increases to the incorporation of cooking oil that contains these FAs into food items.14,44 However, other investigators have reported no changes to ω-3 PUFA content following cooking in fish.45–47

Ultimately, both the critical importance of MM TFs as sources of key nutrients (Se, ω-3 PUFAs) and deleterious environmental contaminants (POPs, Hg), as well as the lack of consistency in preparation impacts on the levels of these compounds in food items, suggest that greater in-depth study is required to ascertain MM TF preparation impacts. To address this data gap, we collected and prepared multiple beluga whale blubber samples during a summer 2014 field campaign in Tuktoyaktuk, Northwest Territories (NT), Canada to track any changes to Se, ω-3 PUFA, POP, and Hg concentrations over the course of preparation. Beluga whale represents a key TF species for the local Inuvialuit indigenous population, and its blubber forms the basis for two distinct TF types: (i) muktuk designates food items composed of the outer layer of blubber and its attached skin and connective tissue, while (ii) uqsuq designates food items derived from the inner layers of blubber [Tuktoyaktuk Hunters and Trappers Committee (THTC), personal communication]. Muktuk is eaten raw, boiled, roasted, or aged via immersion in uqsuq, whereas uqsuq is generally limited to its role as the fermentation medium for ageing muktuk; however, the liquid uqsuq oil formed during ageing is also used as a dipping sauce for other Arctic TFs (THTC, personal communication).

In this study, we sought to address the following questions: (i) how do the various processes involved in producing beluga muktuk and uqsuq products affect the levels of Se, ω-3 PUFAs, POPs, and Hg? (ii) More specifically, how does the muktuk and uqsuq ageing process, and its characteristic uqsuq solid/liquid phase separation, affect ω-3 PUFA levels and the partitioning of POPs? (iii) How does the degree of food preparation change to Se, Hg, and POP levels in beluga TFs compare to their measured extent of hydrophobicity? (iv) Can food preparation impacts on beluga TFs serve as the basis for potential dietary advisories to ensure populations most sensitive to the consumption of these nutrients and contaminants (ex. Arctic indigenous WCBA) experience appropriate intakes?

2. Methods

2.1. Beluga blubber, muktuk, and uqsuq collection

Beluga TF products were collected from 2 animals caught offshore of Tuktoyaktuk, NT during the 2014 summer hunting season by local Inuvialuit subsistence hunters and manuscript co-authors L. Pokiak and J. Pokiak. The whales will be referred to by their respective biomonitoring study IDs: HI-14-06 and HI-14-11. These belugas were members of the Eastern Beaufort Sea population, with a home range including respective summering and wintering grounds in the Beaufort and Bering Seas. The population typically congregates proximate to the Mackenzie River Estuary, and selected individuals are harvested on the shore of nearby Hendrickson Island (69° 30′ N, 133° 35′ W). Immediately upon butchering of the 2 sampled animals, their muscle and blubber tissues were returned to Tuktoyaktuk for further preparation; at this point sampling began.

2.2. Beluga TF preparation

From each whale, a large chunk of blubber (∼500 cm × 250 cm × 20 cm) was designated for sampling, from which smaller pieces were derived following each step in the muktuk and uqsuq preparation processes. The steps involved in preparation of each subsequent muktuk and uqsuq TF are outlined in Fig. 1, and briefly described below.
image file: c7em00167c-f1.tif
Fig. 1 Sampling protocol for evaluating the impact of beluga blubber food preparation. Sampling began following overnight air-drying (A), after which muktuk and uqsuq were separated for further preparation. Muktuk was hang-dried (B), then subjected to one of the following processes: boiling in a large drum (C), boiling in a small pot (D), roasting (E), ageing in uqsuq for either 2 (F) or 5 days (G). Uqsuq was sampled (H) then stored as an ageing medium for 2 (I) or 5 days (J), at which time separation into oil (K) was first observed.

The first blubber samples from both animals were collected following 1–2 days of outdoor air-drying on the hunters' shoreline properties (Fig. 1A). Following air-drying, blubber was separated into the components ultimately forming muktuk and uqsuq TF products. The outer layer of blubber (∼4–5 cm) and skin was cut away to produce muktuk TFs, while the remaining inner blubber was cut into long thin strips to produce uqsuq TFs; connective tissue was discarded. Chunks of outer blubber and skin (muktuk precursors) were cut into smaller pieces connected by thin uncut cords of tissue to hang-dry atop wooden racks (Fig. 1B). Following this second drying step, an additional muktuk precursor sample was collected. Subsequently, dried outer blubber and skin pieces were designated for further preparation by 1 of 4 methods: (1) samples were boiled in a large drum of water with numerous other dried blubber chunks (Fig. 1C), (2) samples were boiled in a small pot to avoid potential cross-contamination with muktuk pieces from other animal body regions48 (Fig. 1D), (3) samples were roasted over an open flame (Fig. 1E), or (4) samples were set aside for ageing in raw uqsuq (Fig. 1F and G).

Uqsuq was sampled via the following scheme. Baseline samples were collected following separation of the inner blubber from the outer blubber prior to hang-drying. Then the sheets of inner blubber were cut into thin ∼20 cm strips and collected in a small plastic bucket for several days of fermentation (Fig. 1H). Second and third uqsuq samples were collected following 2 and 5 days of ageing (Fig. 1I and J). At the 5 day mark, characteristic uqsuq phase separation was first observed, and samples of liquid uqsuq oil were also collected (Fig. 1K). We are aware that uqsuq may be regularly fermented for far longer than the maximum of 5 days performed in this study (THTC, personal communication), however limits on our fieldwork period prevented uqsuq fermentation for longer periods. Additionally, raw muktuk samples were also aged via immersion in uqsuq, and were similarly sampled after 2 or 5 days (Fig. 1F and G). Muktuk and uqsuq samples were stored at −20 °C in a field freezer, and upon campaign completion were transported frozen via cooler to the Advanced Laboratory for Fluorinated and Other New Substances in the Environment (ALFONSE) at the University of Toronto Scarborough (UTSC) where they were again stored at −20 °C.

2.3. Measurement of nutrients

To ascertain blubber Se concentrations, samples were first microwave digested with 2.5–6 mL of concentrated nitric acid (Trace Metal grade) and 1.0 mL hydrogen peroxide (30% TraceSelect) in Teflon vessels. Digestates were transferred to pre-cleaned Teflon vials, mixed with 5.0 mL concentrated HCl (Trace Metal grade), and heated on a hot plate to 85 °C for 45 minutes. After cooling to room temperature, digestates were then transferred to clean 15 mL polypropylene Falcon tubes (Starstedt), diluted to 15 mL, and stored at 4 °C until instrumental analysis.49 Concentrations of Se were then quantified using an Agilent 8800 ICP-MS.

To measure PUFAs, blubber lipids were first extracted according to protocols previously defined.50 Blubber ω-3 PUFA concentrations were then quantified using previously outlined gas–liquid chromatography (GLC) methods51 by Lipid Analytical Laboratories at the University of Guelph Research Park.

2.4. Measurement of environmental contaminants

The panel of quantified environmental contaminants were divided into three groups based on analytical method requirements: (i) OCPs, polycyclic aromatic hydrocarbons (PAHs), PBDEs, and PCBs, (ii) perfluoroalkyl and polyfluoroalkyl substances (PFASs), and (iii) Hg.
2.4.1. Measurement of OCPs, PAHs, PBDEs, PCBs. Standards and chemicals. All environmental contaminant standards for OCPs, PAHs, PBDEs, and PCBs (both native and radiolabeled/deuterated) were purchased from Cambridge Isotope Laboratories (Tewksbury, MA). Details on these standards, as well as other reagents, are given in the ESI. Sample extraction and cleanup. Each blubber, muktuk, uqsuq, or oil sample was extracted in duplicate using a modified solvent extraction and cleanup method based on published directives from the Environment Canada National Laboratory for Environmental Testing.52 Details of this methodology, as well as relevant modifications, are provided in the ESI. Instrumental analysis. Quantification of PCBs, PAHs, and OCPs in blubber, muktuk, and uqsuq was achieved using an Agilent 7890 gas chromatograph (GC) coupled to a 7000 triple quadrupole MS/MS operated in electron ionization mode. A DB-5 column (J&W Scientific: 30 m, 0.25 mm i.d., 0.25 μm film thickness) was used for PCB, PAH, and OCP separation. PBDE quantification was performed with a Thermo Scientific Trace 1310 GC coupled to an TSQ8000 Evo triple quadrupole MS/MS equipped with a TriPlus RSH autosampler. An Rxi-5Sil MS column (RESTEK: 15 m, 0.25 mm i.d., 0.1 μm d.f.) was employed to achieve PBDE separation. Operating parameters for the two GC-MS/MS systems are described in the ESI, based on adaptations of methods by Agilent Technologies53–55 and Thermo Fisher Scientific.56
2.4.2. Measurement of PFASs. Standards and chemicals. PFAS analytical standards (both native and mass-labeled) were purchased from Wellington Laboratories (Guelph, ON). Details on these standards, as well as other reagents, are given in the ESI. Sample extraction. Each blubber, muktuk, uqsuq, or oil sample was extracted in duplicate using a modified solid phase extraction (SPE) method after alkali digestion. Details of the extraction method are provided in the ESI. Instrumental analysis. Separation and quantification of PFASs in blubber, muktuk, and uqsuq samples were performed using an Acquity ultra performance liquid chromatograph (UPLC) and a Xevo TQ-S mass spectrometer/mass spectrometer (MS/MS – Waters Corporation) operated in negative ionization mode with an atmospheric electrospray interface. An Acquity BEH C18 column (2.1 × 75 mm, 1.7 μm, 100 Å), maintained at 40 °C was used to achieve chromatographic separation. A 10 μL extract aliquot was injected onto the column, with 2 mM ammonium acetate in Milli-Q water and MeOH used as mobile phases. Detailed LC-MS/MS conditions have been reported elsewhere.57
2.4.3. Measurement of Hg. For mercury quantification, blubber samples were digested on a hot plate in a combination of nitric acid and sulfuric acid (7[thin space (1/6-em)]:[thin space (1/6-em)]3), diluted with deionized water, oxidized by the addition of bromine monochloride (BrCl) solution, left to sit overnight, and dehalogenated through addition of hydroxylamine hydrochloride (NH2OH·HCl) prior to instrumental analysis. Total mercury (THg) concentrations were then determined using a Tekran Model 2600 Cold Vapor Atomic Fluorescence Spectrophotometer (CVAFS) Mercury Analysis System.

2.5. Quality assurance and quality control

Several quality assurance and quality control strategies were employed to minimize external contamination during sample extraction. Materials used to handle samples and solvents, such as transfer pipettes, pipette tips, test tubes, falcon tubes, and septa, were checked to ensure non-detectable presence of PFAS, PCB, PBDE, and OCP compounds. Additionally, all consumables were contaminant-free. Samples were extracted in ALFONSE, a Class 100 clean laboratory constructed using materials thoroughly checked for the presence of fluorinated, chlorinated, and brominated organic chemicals, while contamination via trace gases and particles entering the space is limited by an advanced carbon and high-efficiency particulate arrestance (HEPA) air filtering scheme. Laboratory and field solvent blanks were processed using the same experimental methods as blubber samples to determine the contaminant levels introduced via shipping, handling, storage, extraction and cleanup.

2.6. Data analysis

Statistical comparisons of preparation step impact on POP concentrations were performed using R Ver 3.4.0 (R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org). Although whale sample size was limited (n = 2), combining data for both whales and all preparation steps (n = 11) produced POP-specific datasets of 22 observations, with each data point measured in duplicate. Note that data analysis was not performed for Hg, Se, or ω-3 PUFAs based on singlet measures for each preparation step in both whales.

POPs were grouped where possible, producing summed concentrations for chlordanes, dichlorodephenyltrichloroethanes (DDTs), endosulfans, hexachlorocyclohexanes (HCHs), PAHs, PBDEs, and PCBs, while dieldrin, hexachlorobenzene (HCB), and mirex were compared individually. As POP concentration datasets were not normally distributed, according to the Shapiro–Wilk test of normality, a non-parametric approach was employed. Blubber concentration replicates for each whale-preparation step combination (n = 44) were ranked, and Kruskal–Wallis one-way analyses of variance (ANOVAs) were performed on these ranks. The influence of sample whale of origin was controlled for, and thus Tukey's post hoc test could be used to identify significant differences between preparation steps. Statistical hypotheses tests were all 2-tailed where possible, using an assumed type 1 error rate probability of α = 0.05.

3. Results and discussion

3.1. Determinants of beluga TF nutrient levels

All nutrients were measured once in each blubber sample, and individual whale data were used to explore differences between preparation steps. Detectable concentrations of Se and ω-3 PUFAs were observed in all blubber samples, and the results for these compounds are depicted in Fig. 2. Total fatty acid content (g per g ww) is depicted in Fig. 2A, while further quantification of ω-3 PUFAs is shown in Fig. 2B. Se results are depicted in Fig. 2C. The appreciable error estimates for these analyses reflect the general differences in nutrient content observed between whales, though these discrepancies were less than those measured for POPs, where inter-animal differences far exceeded replicate variability. For ω-3 PUFA analyses we focused on the most nutritionally relevant congeners to humans, DHA and EPA, while also comparing other ω-3 compounds, as well as total ω-6 PUFAs (Fig. 2B).
image file: c7em00167c-f2.tif
Fig. 2 The impact of beluga blubber preparation on nutrient levels. Wet-weight contents of saturated, monounsaturated, and polyunsaturated FAs are depicted in (A), while PUFAs are further differentiated in (B). Se concentrations are depicted in (C). Nutrients were measured once for each whale, and depictions of these individual concentrations are interleaved for whales HI-14-06 and HI-14-11.
3.1.1. ω-3 PUFAs. To assess the impact of preparation methods on blubber nutrients we assessed the sum of all PUFAs (including both ω-3 and ω-6 variants). Though we could not interpret our results statistically, they do suggest minimal differences in total PUFA content between most preparation steps. However, roasted and 5 day aged muktuk may exhibit reduced PUFA levels, and 5 day aged uqsuq and uqsuq oil may possess increased amounts. The loss of PUFAs from roasted samples mirrors that of total lipid concentrations (Fig. 2A), where lipid depletion likely resulted from melting while suspended above an outdoor cookfire. The data suggest a possible decreasing PUFA content with time during muktuk ageing, while an opposite effect of increasing PUFA content in uqsuq during ageing was suggested. The ageing process typically takes place outdoors during the warm summer months, with ambient temperatures causing the solid uqsuq to liquefy into oil, potentially resulting in preferential liberation of lipid and PUFA constituents from the solid blubber. It is possible that muktuk and uqsuq subjected to longer ageing periods could exhibit further diminished and elevated PUFA levels, respectively. With sufficient time, an extended sampling protocol could detect these differences, especially given the knowledge that uqsuq ageing can be performed for several weeks or months (THTC, personal communication).
3.1.2. Se. The levels of Se in beluga blubber products were generally similar, as all preparation steps possessed levels ≤2 μg per g ww. With the exception of certain values (i.e. roast muktuk – Se: 2.91 μg per g ww), these levels are consistent with previous beluga blubber measurements.4 For detailed literature comparisons, please refer to ESI Table S23.

Our measurements suggest that Se levels in uqsuq may be lower than those in muktuk (Fig. 2C). The potential for greater Se levels in muktuk was unsurprising, as this element is found at highest levels in beluga skin, meaning muktuk products are typically an excellent dietary Se source.4 Though all larger blubber and muktuk pieces included skin, skin was not contained in the actual ∼1 g subsamples digested for Se analysis. Thus, the greater Se amounts in muktuk samples were likely due to their proximity to the skin. Conversely, Se was found at lowest levels in uqsuq oil, suggesting an aversion to this near-total lipid phase. Roasted muktuk may possess the greatest Se concentrations, likely due to the lipid depletion phenomenon discussed above. As lipid melted away from roasted samples, the blubber phase containing Se decreased in volume causing concentrations to rise. This mechanism has frequently been identified in fish preparation analyses as responsible for increased Se levels following cooking.36 As Se is nearly absent from uqsuq oil its presence in roasted muktuk lipid drippings is likely minimal, suggesting the bulk of Se is retained in the muktuk, effectively increasing its concentrations.

3.2. Determinants of beluga TF neutral contaminant levels

3.2.1. OCPs, PBDEs, PCBs, FOSA. We focus the discussion here on neutral POPs found in ≥50% of our blubber samples, which included all target OCP, PBDE, and PCB compounds except o,p′-dichlorodiphenyldichloroethylene (DDE), cis-chlordane, and PBDE-183. Depictions of OCP, PBDE, and PCB concentration replicates are interleaved for each whale in Fig. 3 and 4. Summed concentrations of chlordanes, DDTs, endosulfans, and HCHs are displayed in Fig. 3A–D, respectively, while summed concentrations of PCBs and PBDEs are displayed in Fig. 4A and B, respectively. Fig. 4C also illustrates the levels of neutral PFAS perfluorooctacesulfonamide (FOSA), based on its similarity to OCPs, PBDEs, and PCBs in preparation influence. The concentrations of OCPs, PCBs, PBDEs, and FOSA are compared to literature reports for beluga blubber in ESI Tables S24–S26, and S28, respectively. ESI Fig. S1 depicts preparation step impacts on additional OCPs: dieldrin, HCB, and mirex.
image file: c7em00167c-f3.tif
Fig. 3 The impact of beluga blubber preparation on OCP levels. Summed, lipid-normalized concentrations of chlordanes (A), DDTs (B), endosulfans (C), and HCHs (D) were measured in duplicate for each preparation step, and depictions of these replicate pairs are interleaved for belugas HI-14-06 and HI-14-11.

image file: c7em00167c-f4.tif
Fig. 4 The impact of beluga blubber preparation on PCB, PBDE, FOSA, and PAH levels. Lipid-normalized concentrations of summed PCBs (A), summed PBDEs (B), FOSA (C), and summed PAHs (D) were measured in duplicate for each preparation step, and depictions of these replicate pairs are interleaved for belugas HI-14-06 and HI-14-11. Significant differences were detected between preparation steps for PCBs, PBDEs, and PAHs while controlling for whale identity; those that do not share any letter designators were distinct according to Kruskal–Wallis ANOVA with Tukey multiple comparisons (p < 0.05).

Notably, whale HI-14-06 consistently exhibited lipid-adjusted OCP, PBDE, and PCB concentrations that exceeded those of whale HI-14-11 by 2–3 fold (Fig. 3 and 4). Because the degradation half-lives of these POPs in the environment and biota are on the order of years58–61 this POP exposure pattern suggests that beluga HI-14-06 is older than HI-14-11, and/or that HI-14-06 is male and HI-14-11 is female.62 At the time of their capture, the ages and sexes of belugas HI-14-06 and HI-14-11 were determined, the former by tooth analysis.63 Both individuals were male and aged 37 and 24 years, respectively. Notably, the older animal (HI-14-06) lived through the emissions peaks of several legacy POPs and therefore experienced higher environmental legacy POP levels during its development; it is more likely to exhibit elevated OCP and PCB levels.62 Thus, the age difference between belugas appears sufficient for the impact of birth-year proximity to POP emissions peaks to differentiate their POP exposures. Otherwise, differences in environmental contaminant exposure between similarly aged and sexed wildlife could also arise from interindividual differences in diet,64 metabolic transformation capability,65 migratory behaviour,66 or blubber dynamics.67

Alternatively, the ageing process may simply exacerbate POP concentration differences already present due to blubber stratification. Krahn et al.48 previously described the variability of OCP and PCB levels in beluga by blubber depth, and found that distribution was inconsistent between whales. Maximum POP levels are often measured in whale outer blubber layers,68 but the findings by Krahn et al.48 indicate that greatest POP concentrations may rather be found in inner blubber. Additionally, Krahn et al.48 determined that blubber stratification patterns were consistent between chemical classes for each examined whale, while in our 2 belugas this was not always the case; beluga HI-14-06 exhibited higher PBDE and PCB levels in inner blubber, but greater OCP levels in outer. However, note that both the current and Krahn et al.48 studies were characterized by small population size (n = 2 and n = 6, respectively).

In contrast with the influence of age, the distinct preparation processes examined did not exert a reliable effect on neutral POPs across the range of contaminants examined (Fig. 3 and 4). Thus, the limited data from our study suggest that baseline whale contaminant concentrations, influenced by age, sex, diet, migration etc., are more relevant to variability in human legacy POP exposure than the methods used to prepare beluga blubber TFs. However, there did appear to be some consistent impacts of preparation on neutral POP levels. For certain chemical groups greater concentrations in uqsuq versus muktuk were suggested, presumably driven by the phase separation involved in ageing uqsuq. Specifically, the elevated lipid content of the liquid oil served as a more appropriate residence phase for these hydrophobic chemicals (∼4 < log[thin space (1/6-em)]KOW < ∼10),69,70 causing greater partitioning into liquid oil versus certain muktuk phases, significantly so for PCBs and PBDEs (p < 0.05 − Fig. 4A and B).

Unlike for Se, roasting muktuk did not appreciably affect OCP, PBDE, or PCB levels in either whale, despite the appreciable removal of lipid via the cooking process. This suggests that these contaminants are likely well-mixed within the blubber compartment, and that a corresponding amount of chemical mass is removed alongside lipid mass melted off roasted muktuk. This explains why the amplification that was observed for hydrophilic compounds was not present for OCPs, PBDEs, and PCBs.

3.2.2. PAHs. We again focused on PAHs found in ≥50% of our blubber samples, which included all target PAH compounds except benzo(b)fluoranthene, benzo(k)fluoranthene, and benzo(g,h,i)perylene. Depictions of summed PAH concentration replicates are interleaved in Fig. 4D. The concentrations of PAHs are also compared to literature reports for beluga blubber in ESI Table S27. Our discussion of concentration determinants for PAHs is separated from our findings for other neutral contaminants (OCPs, PBDEs, and PCBs) based on appreciable differences in these factors.

Primarily, the relationship between PAH concentrations and whale age was the reverse of that observed for OCPs, PBDEs, and PCBs, where levels in beluga HI-14-11 exceeded those in HI-14-06 by over an order of magnitude (Fig. 4D). This reverse exposure trend was surprising, and likely due to the much shorter environmental and biotic degradation half-lives for PAHs compared to legacy POPs.71 Thus, it is likely that the bulk of measured PAH levels were introduced during preparation. This hypothesis is also supported by previous reports that PAHs are poorly biomagnified in aquatic food webs.72,73 Under direction from co-authors L. and J. Pokiak, both belugas were prepared over the course of 2–3 days adjacent to smokehouses (i.e. sources of smoke from organic material combustion), yet the HI-14-11 smokehouse was more regularly and recently used. Also, whale HI-14-11 was prepared within the bounds of Tuktoyaktuk proper, whereas whale HI-14-06 was prepared in a more remote location. Further, under direction from J. Pokiak uqsuq from beluga HI-14-11 was aged within his smokehouse, as compared to whale HI-14-06 uqsuq which was aged under a shelter in L. Pokiak's backyard, several kilometers away from his smokehouse. These differences, which reflect local variability in preparation customs, likely contributed to the greater PAH levels in muktuk and uqsuq from beluga HI-14-11.

This set of findings is perhaps most noteworthy due to our primary interest in the modulatory effect of TF preparation on beluga blubber nutrient and environmental contaminant concentrations,11,17 rather than introduction of these compounds.74 However, PAH results indicate that muktuk and uqsuq preparation may also deposit environmental contaminants into blubber, in addition to altering levels present at baseline. The close ties between beluga preparation and hunting camp smokehouses (THTC, personal communication) suggest that consistent elevation of PAH levels in prepared beluga blubber products may be a common occurrence across the Canadian Arctic. For certain processes this may be unavoidable, as the equipment necessary to hang dry muktuk, or beluga meat (nikku), is frequently stored and used beside or inside a smokehouse (THTC, personal communication). However, for certain preparation processes, such as uqsuq ageing, the introduction of PAHs during processing could be mitigated by the choice of storage location. The storage sites for fermenting uqsuq pails are varied, and the use of space beside or inside a smokehouse can potentially be avoided. The influence of uqsuq ageing locale can be plainly seen in our data, where baseline uqsuq samples for both whales contained similar levels of PAHs (∼5 ng per g lipid), but ageing performed inside a smokehouse (HI-14-11) resulted in aged uqsuq oil exhibiting nearly 5-fold greater total PAH amounts.

PAHs also appeared to be more affected by certain preparation processes when compared to other neutral environmental contaminants (Fig. 4D). Specifically, roasting muktuk over a cookfire appeared to elevate PAH levels above all other preparation processes, excluding air-dried muktuk and aged uqsuq oil for whale HI-14-06 (Fig. 4D). In fact, significantly greater PAH levels were found in roasted muktuk versus baseline and 2 d aged uqsuq when controlling for whale identity (p < 0.05). Since PAHs are readily derived from combusted organic matter,75 the proximity of these cooked muktuk pieces to the woodfire smoke introduced appreciable PAH amounts.

PAHs also appear to be sensitive to the phase separation occurring during uqsuq ageing, as both whales exhibited uqsuq oil PAH concentrations that appeared to exceed those from all preparation steps except HI-14-11 air-drying, hang-drying, and roasting (plus HI-14-06 air-dried muktuk, with the same caveat stated above). Note that significantly greater sum PAH levels were detected in uqsuq oil versus baseline uqsuq (p < 0.05). The sensitivity of PAHs to uqsuq ageing was surprising given a lack of impact on OCPs, PBDEs, and PCBs. As PAHs generally exhibit lower hydrophobicities than most other measured chemicals according to octanol–water partition coefficient values,76 we hypothesized they would be less sensitive to uqsuq phase separation. However, the values for heavier PAH congeners (BaP, BghiP, IP – log[thin space (1/6-em)]KOW > 6) are comparable to those for many OCPs, PBDEs, and PCBs. We anticipated that the tendency of neutral organic contaminants to partition into aged uqsuq oil would correlate with hydrophobicity, but this relationship was not apparent, as some of the least hydrophobic chemicals we measured were most amplified in uqsuq oil (i.e. NAP, ACN log[thin space (1/6-em)]KOW < 4). Thus, we hypothesize that PAH concentrations were increased during uqsuq ageing primarily due to deposition and not because of phase-separation-based enrichment. Especially because greater deposition was noted in whale HI-14-11 due to its greater proximity to, and duration nearby, local PAH sources (smokehouses, cookfires).

3.3. Determinants of beluga TF ionogenic contaminant levels

3.3.1. PFASs. Due to low detection for several of the 13 targeted ionogenic PFASs in beluga blubber, results were limited to the five compounds found in ≥50% of samples: perfluorononanoate (PFNA), perfluorodecanoate (PFDA), perfluoroundecanoate (PFUnDA), perfluorododecanoate (PFDoDA), and perfluorooctane sulfonate (PFOS). Depictions of summed PFCA and PFOS concentration replicates are interleaved in Fig. 5A and B, respectively; note that data for FOSA is described above alongside other neutral contaminants; As PFASs accumulate primarily in the blood and liver of beluga whales,77,78 there is limited literature data quantifying blubber levels. However, both Kelly et al.77 and Ostertag et al.79 provided tissue distribution data, including blubber, for several PFAS compounds; these concentrations are compared to values for blubber, muktuk, and uqsuq from the current study in ESI Table S28. Our findings align well with these 2 published datasets, and suggest that PFOS and FOSA are found more readily in beluga blubber than other PFASs. For example, measured levels of PFOS only (in ng per g ww) were similar to total perfluorinated carboxylic acid concentrations (ΣPFCAs – range of 4 to 6 compounds) in all 3 studies.
image file: c7em00167c-f5.tif
Fig. 5 The impact of beluga blubber preparation on PFAS and Hg levels. Wet-weight concentrations of summed PFCAs (A) and PFOS (B) were measured in duplicate for each preparation step, and depictions of these replicate pairs are interleaved for belugas HI-14-06 and HI-14-11. Hg (C) was measured once for each whale, and depictions of these individual concentrations are interleaved for whales HI-14-06 and HI-14-11. Significant differences were detected between preparation steps for PFCAs and PFOS while controlling for whale identity; those that do not share any letter designators were distinct according to Kruskal–Wallis ANOVA with Tukey multiple comparisons (p < 0.05).

Compared to legacy POPs, the greater similarity in PFAS concentrations between whales was unsurprising given the recency of PFAS emissions as compared to legacy POPs. Also, the influences of most preparation methods on PFAS levels were minor, with the exceptions of roasting muktuk and ageing uqsuq. Roasting led to increased PFOS concentrations compared to certain other muktuk treatments (i.e. air-dry, hang-dry, boil pot, age 2 d: p < 0.05 – Fig. 5B), likely based on the same phase depletion effect discussed above for Se. Additionally, PFASs were reduced in uqsuq oil, exhibiting significantly lower concentrations for all PFAS comparisons except to pot-boiled and 2 d-aged muktuk for PFOS (p < 0.05; Fig. 5A and B). Notably, levels of PFNA, PFDA, PFUnDA, and PFOS concentrations all fell below detection limits in uqsuq oil.

The effects of these two preparation steps suggest that the phase composition of different beluga blubber products contributes strongly to their varied PFAS concentrations. As maximal PFAS concentrations are frequently detected in beluga blood,77,78 the blood content of samples likely contributed strongly to sample PFAS variability, especially since there was no way to control the blood contents of either field-sampled blubber pieces or extracted 1 g subsamples. Selective lipid phase depletion during muktuk roasting may elevate PFOS concentrations, due to a greater phase proportion of blood/water in the remaining blubber. Notably, this solvent depletion effect mirrors that observed in previous studies of preparation impacts on environmental contaminant levels.9,29–31

The 4 PFCAs and PFOS are estimated to possess pKa values <1,80,81 and thus at physiological pH all PFCAs and PFOS would be >99% ionized. Thus, the uqsuq ageing depletion effect observed for highly ionized PFASs appears to be directly related to the hydrophilicity of these chemicals. As uqsuq oil possessed the highest sampled lipid content (>80% for both whale samples – Fig. 2A), PFCAs and PFOS were either retained in solid uqsuq, or preferentially partitioned into a non-lipid phase in the aged uqsuq mixture (i.e. blood proportion of immersed muktuk).

Interestingly, significant decreases to uqsuq oil PFCA and PFOS concentrations during ageing were not accompanied by equivalent corresponding increases to their levels in aged muktuk and uqsuq. It is important to note that uqsuq samples were taken from a large mixture of blubber immersed in the same ageing pail, as per co-author instruction (see Fig. 1H–J). As the uqsuq removed from this mixture after ageing represented a fraction of the total prepared, it is possible that pail PFAS contents were not well-mixed. Thus, PFAS levels may have varied substantially within the uqsuq mixture. In other words, our aged uqsuq and uqsuq oil samples may not have been representative of the whole uqsuq pail based on an uneven distribution of these contaminants during the ageing process. A follow-up study could clarify this issue via a more thorough uqsuq sampling program.

3.3.2. Hg. The levels of Hg in beluga blubber are displayed in Fig. 5C, and for all preparation steps concentrations were ≤1 μg per g ww. With the exception of certain values (i.e. 5 day aged uqsuq – 1.04 μg per g ww), these levels are consistent with previous beluga blubber measurements.4 For detailed literature comparisons, please refer to ESI Table S23.

Generally it did not seem as if proximity to the skin differentiated Hg levels in muktuk versus uqsuq, unlike Se. Though Hg levels are also found to be higher in beluga skin than blubber,4 these discrepancies are much lower than those for Se, where beluga skin Se concentrations may exceed those in blubber by as much as two orders of magnitude. This suggests that Hg may be more evenly distributed throughout blubber depth, and thus consumption of muktuk with skin as a means to maximize Se intake would not also result in a reciprocal increase to Hg consumption.

Further, no major differences in Hg concentrations were suggested between distinct preparation processes. Particularly, roasting muktuk did not cause an increase to Hg levels through the same lipid depletion mechanism observed for Se and PFASs. This finding mirrors that of the larger food preparation literature, where even equivalent cooking methods for the same compound do not necessarily cause equal solute concentration changes.11 However, Hg levels may increase during uqsuq ageing; though this effect was driven largely by only one of the two belugas. A mechanistic explanation for this phenomenon, if not simply due to experimental variability, could be that solvent depletion is causing Hg to be retained in solid uqsuq as lipid oil liquefies away. Correspondingly, the lowest levels of Hg were found in liquid uqsuq oil (Fig. 5C) suggesting an aversion to this predominantly lipid phase similar to Se and PFASs. Further evidence for this process is suggested by the potentially greater Hg levels in solid muktuk aged for 5 days rather than 2, yet this same effect was not observed for Se. This again may arise from the differences between Se and Hg in blubber distribution by depth.

3.4. Comparing TF processing impact to the food preparation literature

Notably, our TF processing results mirror those from the food preparation literature, wherein methods were found to either increase, decrease, or exert no significant effect on nutrient and environmental contaminant concentrations depending on the specific processes and compounds examined.9,11,14,33,36,45 TF preparation did exert meaningful effects on nutrient and contaminant concentrations for select processes expected to cause a significant change in the nutrient/pollutant storage phase.11,14,17 In particular, two processes were found to regularly impact nutrient and contaminant levels in muktuk and uqsuq products: (i) ageing uqsuq appeared to preferentially concentrate certain hydrophobic compounds while selectively depleting water-soluble substances, and (ii) roasting muktuk may inflate the levels of certain compounds based on solvent depletion. Specifically, ω-3 PUFA (Fig. 2B), PCB (Fig. 4A), and PBDE (Fig. 4B) levels may be enhanced in aged uqsuq oil, while hydrophilic Se (Fig. 2), ionogenic PFAS (Fig. 5A and B), and Hg (Fig. 5C) concentrations were often below method detection limits in this phase. Presumably the selective enrichment effect observed for uqsuq oil would be even further enhanced if measurements were performed for longer than the 5 day period in the present study. Additionally, roasting muktuk appeared to increase concentrations for certain compounds through either depletion of the solvent phase (Se, PFOS), or through deposition from organic material combustion (PAHs). The lack of roasting impact on hydrophobic neutral POPs was presumably because the lipid that is melted away from muktuk samples was well-mixed, resulting in minimal neutral POP concentration changes.

These results present challenges to broad recommendations on the utility of TF preparation in maximizing nutrient intake while minimizing contaminant consumption. However, based on the enhancement of ω-3 PUFAs, depletion of hydrophilic contaminants (Hg, ionogenic PFASs), and a lack of hydrophobic POP enrichment (OCPs, PCBs), uqsuq oil appears to be a preferable beluga blubber TF. Even more so, since risk from the only group of pollutants that was significantly enhanced in uqsuq oil (PAHs) can be mitigated by avoiding uqsuq ageing adjacent to local PAH sources. However, uqsuq oil also lacks Se, due to exclusion of beluga skin. Thus, uqsuq oil could be supplemented with beluga products including skin, such as boiled muktuk, to account for the lack of Se in uqsuq oil.

3.5. Comparing muktuk and uqsuq contaminant concentrations to regulatory exposure thresholds

Our findings are particularly pertinent when considering which chemicals are most likely to convey deleterious dietary exposure risks. We compared blubber contaminant concentrations to the latest version of the Agency for Toxic Substances and Disease Registry (ATSDR) Minimal Risk Levels (MRLs) to ascertain which chemicals posed greatest risk;82,83 these data are listed in Table 1. MRLs represent estimates of daily human exposure to hazardous substances that are unlikely to cause appreciable risk for non-cancer endpoints over specific durations.82 They are also based on no observed adverse effect levels in in vitro toxicity bioassays, modulated by uncertainty factors for conversion to human intake rates.82 As MRLs are formulated as intake rates (mg contaminant per kg bw and day), we estimated the necessary blubber intake rates (g blubber per day) at which a generic indigenous Arctic woman (70 kg) would surpass MRL thresholds based on mean blubber TF contaminant concentrations (ng per g ww) from each beluga. Note that our measured blubber POP concentrations are listed on a wet-weight basis in Table 1, to allow for calculation of wet-weight based blubber intakes. As the goal of the uncertainty factors are to be protective of individuals most sensitive to exposures, we applied them conservatively to calculate only the minimum blubber intake rates at which MRLs would be exceeded.
Table 1 Comparisons of measured beluga blubber POP concentrations to ASTDR Minimum Risk Levels (MRL) for human exposure
Pollutant [HI-14-06] ng per g ww [HI-14-11] ng per g ww MRL chemical intake ratea ng per day Durationb Uncertainty factorc HI-14-06 HI-14-11
Min MRL intake rate g ww per day Risk character. ratioe Min MRL intake rate g ww per day Risk character. ratioe
a Formulated in units ng per day; conversion from published units mg per kg bw and day, assuming blubber consumption by a generic 70 kg woman. b ASTDR exposure durations: intermediate – exposure for more than 14 days, but less than 1 year; chronic – exposure for 1 year or longer. c Mathematical adjustment of MRL bounds based on uncertainty, and applied to the estimated blubber intake rates to generate minimal risk levels. These factors are applied to account for variability in interindividual human sensitivities, human-animal differences, etc. d ΣPCB minimal risk level was derived from Health Canada's Toxicological Reference Value,82 as the MRL from ASTDR was based only on Aroclor 1254; an uncertainty factor of 300 was assumed based on the MRL for Aroclor 1254. e Risk characterization ratios were quantified by comparing the calculated minimal MRL blubber intake rate to the observed consumption rate of beluga blubber by participants in the Inuit Health Survey (0.6 g d−1).
OCP p,p′-DDT 143.9 28.9 3.5 × 104 Intermediate 100 2.4 × 100 2.5 × 10−1 1.2 × 101 5.0 × 10−2
Dieldrin 30.4 7.8 7.0 × 103 Intermediate 100 2.3 × 100 2.6 × 10−1 9.0 × 100 6.7 × 10−2
3.5 × 103 Chronic 100 1.2 × 100 5.2 × 10−1 4.5 × 100 1.3 × 10−1
Endosulfan 15.9 5.7 3.5 × 105 Intermediate 100 2.2 × 102 2.7 × 10−3 6.1 × 102 9.8 × 10−4
3.5 × 105 Chronic 100 2.2 × 102 2.7 × 10−3 6.1 × 102 9.8 × 10−4
HCB 154.0 146.7 7.0 × 103 Intermediate 90 5.1 × 10−1 1.2 × 100 5.3 × 10−1 1.1 × 100
4.9 × 103 Chronic 300 1.1 × 10−1 5.7 × 100 1.1 × 10−1 5.4 × 100
α-HCH 9.3 9.7 5.6 × 105 Chronic 100 6.0 × 102 1.0 × 10−3 5.8 × 102 1.0 × 10−3
β-HCH 7.6 6.4 4.2 × 104 Intermediate 300 1.8 × 101 3.3 × 10−2 2.2 × 101 2.7 × 10−2
γ-HCH 80.2 61.2 7.0 × 102 Intermediate 1000 8.7 × 10−3 6.9 × 101 1.1 × 10−2 5.2 × 101
Mirex 33.0 6.9 5.6 × 104 Chronic 100 1.7 × 101 3.5 × 10−2 8.1 × 101 7.4 × 10−3
trans-Chlordane 11.5 5.1 4.2 × 104 Intermediate 100 3.7 × 101 1.6 × 10−2 8.2 × 101 7.3 × 10−3
4.2 × 104 Chronic 100 3.7 × 101 1.6 × 10−2 8.2 × 101 7.3 × 10−3
PAH ACN 0.5 2.1 4.2 × 107 Intermediate 300 2.8 × 105 2.1 × 10−6 6.7 × 104 9.0 × 10−6
ANT 1.9 16.9 7.0 × 108 Intermediate 100 3.7 × 106 1.6 × 10−7 4.1 × 105 1.4 × 10−6
FLA 0.5 3.4 2.8 × 107 Intermediate 300 1.9 × 105 3.2 × 10−6 2.7 × 104 2.2 × 10−5
FLU 0.5 6.7 2.8 × 107 Intermediate 300 1.9 × 105 3.2 × 10−6 1.4 × 104 4.3 × 10−5
NAP 2.3 29.4 4.2 × 107 Intermediate 90 2.0 × 105 3.0 × 10−6 1.6 × 104 3.8 × 10−5
ΣPBDE 17.1 6.9 2.1 × 102 Intermediate 300 4.1 × 10−2 1.5 × 101 1.0 × 10−1 5.9 × 100
ΣPCBd 1248.5 377.6 9.1 × 103 300 2.4 × 10−2 2.5 × 101 8.0 × 10−2 7.5 × 100
PFOS + FOSA 2.4 2.3 2.1 × 103 Intermediate 90 9.7 × 100 6.2 × 10−2 1.0 × 101 5.9 × 10−2

Ultimately, 4 groups of contaminants stood out as conveying the greatest human exposure risk from beluga blubber TFs: HCB, γ-HCH, ΣPBDEs, and ΣPCBs. Depending on the beluga analyzed, a generic 70 kg indigenous Arctic woman might possibly exceed MRL thresholds through consumption of as little as 0.02 g per day of blubber from HI-14-06 for ΣPCBs, or 0.008 g per day for γ-HCH. These values are significant as human biomonitoring work performed throughout the Canadian Arctic via the Inuit Health Survey previously estimated mean beluga blubber consumption among the sample population (n = 2074) to be approximately 0.6 g per day for strictly beluga blubber and 11.0 g per day for muktuk products containing skin and fat.4 Thus we also calculated simple Risk Characterization Ratios (RCRs) by comparing our calculated MRL blubber intake rates to this observed figure (0.6 g per day), where any value exceeding 1 represents a potential adverse effect risk.84 The only compounds to exhibit RCRs >1 were the same 4 chemical groups identified above (HCB, γ-HCH, ΣPBDEs, ΣPCBs).

However, certain caveats bear mentioning with regard to these observations. Firstly, the Inuit Health Survey intake data represent the mean of over 2000 participants of varying ages, where older, male, high TF consumers heavily skewed consumption. Plus intake rates were based on food frequency questionnaire data,4,85 which are often highly unreliable.86–88 Additionally, exposure guidelines are often targeted specifically to WCBA83 based on the sensitivity of the developing fetus and newborn to toxicological effects; thus inferences about the suitability of blubber TFs should also account for varied age-and-stage sensitivity. Finally, certain chemicals with the lowest MRL intakes also possessed MRLs with the highest uncertainty factors: HCB, ΣPBDEs, and especially γ-HCH. This suggests that MRLs for γ-HCH may be overly conservative by a factor of as much as 1000, which would place them much closer to contaminants of lesser concern.

4. Conclusions

We conducted a series of experiments to investigate the impact of beluga blubber TF preparation on the levels of nutrients and environmental contaminants in these food items. Our results suggest that beluga blubber preparation can have appreciable effects on the concentrations of certain compounds, depending mainly on the process utilized and the chemical properties of the examined substance. Ageing uqsuq was the method primarily responsible for many blubber preparation process impacts we detected, due to the distinctive phase separation between solid fat and liquid oil observed following sufficient fermentation time. Uqsuq oil contained a higher total lipid proportion and ω-3 PUFA content, and its phase separation selectively depleted hydrophilic compounds that do not preferentially partition into lipid-rich phases. Thus uqsuq oil was found to contain near-detection limit levels of Se, Hg, and ionogenic PFASs such as PFCAs and PFOS. However, uqsuq ageing did not consistently enrich hydrophobic POP concentrations in uqsuq oil. Though uqsuq oil occasionally appeared to possess greater levels of OCPs, PBDEs, and PCBs, this effect was minor. Ultimately, our findings support previous food preparation investigations that have identified variable impacts of preparation depending on the types of food and substances examined, as well as measurements often characterized by inconsistency in preparation impacts.

Interestingly, our analysis of PAH variability following beluga blubber preparation indicated that preparation methods may also introduce environmental contaminant concentrations to TFs, rather than simply modulate levels already contained therein. As beluga blubber preparation is regularly conducted within or adjacent to indigenous Arctic smokehouses, and liberally uses woodfires to boil water used in the cooking process, PAHs derived from organic material combustion were readily introduced into blubber products. We found that certain methods, such as directly roasting muktuk over an open flame or ageing muktuk were sufficient to elevate PAH concentrations above levels measured in alternatively prepared blubber TFs. Notably, the impact of ageing on PAH deposition was also influenced by proximity to, and duration nearby a functional smokehouse. These findings suggest that adjustments to beluga blubber TF preparation methods can not only potentially reduce environmental contaminant levels but also prevent their incorporation.

Collectively these results provide some guidance on our goal to provide beluga blubber TF consumption advice. Based on the inconsistency of uqsuq ageing in enriching hydrophobic POP concentrations within uqsuq oil, and the corresponding depletion of certain environmental contaminants from this phase through the same process, it would seem that uqsuq oil is the preferred product for consumption, particularly since ω-3 PUFA levels are also maximized in this product. Though this food item was also correspondingly low in Se, this intake loss can be mitigated by consumption of beluga skin included in alternative muktuk products, with boiled food items likely representing the best choice. Additionally, preparation of beluga TFs physically removed from the environment surrounding an in-use smokehouse can mitigate PAH exposure. All told, we determined that variable beluga blubber TF preparation techniques can differentiate nutrient and pollutant concentrations in these food items, and that aged uqsuq oil is likely the best option for individuals wishing to maximize nutrient content while minimizing pollutant exposure.


We thank the Tuktoyaktuk HTC, Aurora Research Institute, and Fisheries and Oceans Canada for their support in the field. Also, we are grateful to Dr Sohee Kang (University of Toronto Scarborough) for her assistance with data analysis strategies. Finally, we acknowledge funding support from the Natural Sciences and Engineering Research Council of Canada, the Northern Scientific Training Program, the Arctic Institutes of North America, and the Northern Contaminants Program of Indigenous and Northern Affairs Canada (project H-07).


  1. AMAP Assessment, Human Health in the Arctic. Arctic Monitoring and Assessment Programme (AMAP), Oslo, Norway, 2015, vol. vii, p. 165 Search PubMed.
  2. AMAP Assessment, Human Health in the Arctic. Arctic Monitoring and Assessment Programme (AMAP), Oslo, Norway, 2009, vol. vii, p. 256 Search PubMed.
  3. H. V. Kuhnlein, H. M. Chan, D. Leggee and V. Barthet, J. Food Compos. Anal., 2002, 15, 545–566 CrossRef CAS.
  4. B. D. Laird, A. B. Goncharov, G. M. Egeland and H. M. Chan, J. Nutr., 2013, 143, 923–930 CrossRef CAS PubMed.
  5. K. Borgå, A. T. Fisk, P. F. Hoekstra and D. C. G. Muir, Environ. Toxicol. Chem., 2004, 23, 2367–2385 CrossRef.
  6. S. G. Donaldson, J. Van Oostdam, C. Tikhonov, M. Feeley, B. Armstrong and P. Ayotte, et al. , Sci. Total Environ., 2010, 408, 5165–5234 CrossRef CAS PubMed.
  7. J. Van Oostdam, S. G. Donaldson, M. Feeley, D. Arnold, P. Ayotte and G. Bondy, et al. , Sci. Total Environ., 2005, 351–352, 165–246 CrossRef CAS PubMed.
  8. P. F. Hoekstra, T. M. O'Hara, S. M. Backus, C. Hanns and D. C. G. Muir, Environ. Res., 2005, 98, 329–340 CrossRef CAS PubMed.
  9. M. Lemire, M. Kwan, A. E. Laouan-Sidi, G. Muckle, C. Pirkle, P. Ayotte and É. Dewailly, Sci. Total Environ., 2015, 509–510, 248–259 CrossRef CAS PubMed.
  10. S. K. Moses, A. V. Whiting, D. C. G. Muir, X. Wang and T. M. O'Hara, Int. J. Circumpolar Health., 2009, 68, 354–371 CrossRef PubMed.
  11. J. L. Domingo, Crit. Rev. Food Sci. Nutr., 2011, 51, 29–37 CrossRef CAS PubMed.
  12. G. Perelló, R. Martí-Cid, V. Castell, J. M. Llobet and J. L. Domingo, Food Chem. Toxicol., 2009, 47, 709–715 CrossRef PubMed.
  13. G. Perelló, R. Martí-Cid, V. Castell, J. M. Llobet and J. L. Domingo, Food Control, 2010, 21, 178–185 CrossRef.
  14. J. Weber, V. C. Bochi, C. P. Ribeiro, A. d. M. Victório and T. Emanuelli, Food Chem., 2008, 106, 140–146 CrossRef CAS.
  15. S. Bayen, P. Barlow, H. K. Lee and J. P. Obbard, J. Toxicol. Environ. Health, Part A, 2005, 68, 253–265 CrossRef CAS PubMed.
  16. S. P. Bhavsar, X. Zhang, R. Guo, E. Braekevelt, S. Petro, N. Gandhi, E. J. Reiner, H. Lee, R. Bronson and S. A. Tittlemier, Environ. Int., 2014, 66, 107–114 CrossRef CAS PubMed.
  17. D. F. Rawn, K. Breakell, V. Verigin, S. A. Tittlemier, L. Del Gobbo, M. L. Diamond, L. Vanderlinden and D. Sit, J. Agric. Food Chem., 2013, 61, 989–997 CrossRef CAS PubMed.
  18. A. Schecter, O. Päpke, K. C. Tung, T. Brown and A. Musumba, Toxicol. Environ. Chem., 2006, 88, 207–211 CrossRef CAS.
  19. J. Moya, K. G. Garrahan, T. M. Poston and G. S. Durell, Bull. Environ. Contam. Toxicol., 1998, 60, 845–851 CrossRef CAS PubMed.
  20. R. A. Sherer and P. S. Price, Qual. Assur., 1993, 2, 396–407 CAS.
  21. N. D. Wilson, N. M. Shear, D. J. Paustenbach and P. S. Price, J. Exposure Anal. Environ. Epidemiol., 1998, 8, 423–440 CAS.
  22. M. Rose, S. Thorpe, M. Kelly, N. Harrison and J. Startin, Chemosphere, 2001, 43, 861–868 CrossRef CAS PubMed.
  23. M. E. Zabik, M. J. Zabik, A. M. Booren, M. Nettles, J.-H. Song, R. Welch and H. Humphrey, J. Agric. Food Chem., 1995, 43, 993–1001 CrossRef CAS.
  24. H. M. Chan and O. Receveur, Environ. Pollut., 2000, 110, 1–2 CrossRef CAS PubMed.
  25. J. L. Kirk, I. Lehnherr, M. Andersson, B. M. Braune, L. Chan and A. P. Dastoor, et al. , Environ. Res., 2012, 119, 64–87 CrossRef CAS PubMed.
  26. Government of the Northwest Territories, Mercury Levels in Fish, Health and Social Services, Yellowknife, NT, 2012, p. 2 Search PubMed.
  27. Nunavik Regional Board of Health and Social Services, Results from the Nunavik Child Development Study (NCDS), Public Health Implications, Kuujjuaq, QC, 2011, p. 4 Search PubMed.
  28. Nunavut Tunngavik Incorporated, Release of Inuit Health Survey Report on Contaminant Assessment in Nunavut, Iqaluit, NU, 2012, p. 376 Search PubMed.
  29. J. Burger, C. Dixon, S. Boring and M. Gochfeld, J. Toxicol. Environ. Health, Part A, 2003, 66, 817–828 CrossRef CAS PubMed.
  30. V. A. Gremiachikh, I. I. Tomilina and V. T. Komov, Gig. Sanit., 2007, 64–67 CAS.
  31. J. N. Morgan, Adv. Exp. Med. Biol., 1999, 459, 195–211 CrossRef CAS PubMed.
  32. T. E. Fox, E. G. H. M. Van den Heuvel, C. A. Atherton, J. R. Dainty, D. J. Lewis, N. J. Langford and H. M. Crews, et al. , Eur. J. Clin. Nutr., 2004, 58, 343–349 CrossRef CAS PubMed.
  33. C. T. Martins, C. Almeida and P. C. Alvito, Food Anal. Method., 2011, 4, 77–83 CrossRef.
  34. M. B. Atta, L. A. El-Sebaie, M. A. Noaman and H. E. Kassab, Food Chem., 1997, 58, 1–4 CrossRef CAS.
  35. M. S. Bratakos, T. F. Zafiropoulos, P. A. Siskos and P. V. Ioannou, Int. J. Food Sci. Technol., 1988, 23, 585–590 CrossRef CAS.
  36. C. Afonso, S. Costa, C. Cardoso, R. Oliveira, H. M. Lourenço and A. Viula, et al. , Environ. Res., 2015, 143, 130–137 CrossRef CAS PubMed.
  37. S. Mierke-Klemeyer, R. Larsen, J. Oehlenschläger, H. Maehre, E. O. Elvevoll and N. M. Bandarra, et al. , Eur. Food Res. Technol., 2008, 227, 827–833 CrossRef CAS.
  38. J. Bakar, E. Z. Rahimabadi and Y. Man, Food Sci. Technol., 2008, 41, 2144–2150 CAS.
  39. M. Candela, I. Astiasaran and J. Bello, J. Agric. Food Chem., 1998, 46, 2793–2796 CrossRef CAS.
  40. M. Candela, I. Astiasaran and J. Bello, Food Chem., 1997, 58, 227–231 CrossRef CAS.
  41. M. I. Gladyshev, N. N. Sushchik, G. A. Gubanenko, S. M. Demirchieva and G. S. Kalachova, Food Chem., 2006, 96, 446–451 CrossRef CAS.
  42. N. M. Stephen, R. Jeya Shakila, G. Jeyasekaran and D. Sukumar, J. Food Sci. Technol., 2010, 47, 174–181 CrossRef CAS PubMed.
  43. A. U. Türkkan, S. Cakli and B. Kilinc, Food Bioprod. Process., 2008, 86, 163–166 CrossRef.
  44. D. Larsen, S. Y. Quek and L. Eyres, Food Chem., 2010, 119, 785–790 CrossRef CAS.
  45. F. de Castro, H. Sant'Ana, F. M. Campos, N. M. B. Costa, M. T. C. Silva, A. L. Salaro and S. D. C. C. Franceschini, Food Chem., 2007, 103, 1080–1090 CrossRef.
  46. E. Fajmonová, J. Zelenka, T. Komprda, D. Kladroba and I. Sarmanová, Czech J. Anim. Sci., 2003, 48, 85–92 Search PubMed.
  47. M. R. Neff, S. P. Bhavsar, E. Braekevelt and M. T. Arts, Food Chem., 2014, 164, 544–550 CrossRef CAS PubMed.
  48. M. M. Krahn, D. P. Herman, G. M. Ylitalo, C. A. Sloan, D. G. Burrows and R. C. Hobbs, et al. , J. Cetacean Res. Manag., 2004, 6, 175–189 Search PubMed.
  49. R. D. Evans, N. M. Grochowina, N. Basu, E. M. O'Connor, B. E. Hickie and K. Rouvinen-Watt, et al. , Chemosphere, 2016, 144, 1582–1588 CrossRef CAS PubMed.
  50. J. Folch, M. Lees and G. H. Sloane Stanley, J. Biol. Chem., 1957, 226, 497–509 CAS.
  51. Y. E. Zhou, S. Kubow and G. M. Egeland, Int. J. Circumpolar Health., 2011, 70, 498–510 CrossRef PubMed.
  52. The National Laboratory for Environmental Testing, 2006-2007 Schedule of Services, Environment Canada, Burlington, Ontario, Canada, 2007, p. 8 Search PubMed.
  53. P. L. Wylie and C.-K. Meng, A method for the trace analysis of 175 pesticides using the Agilent triple quadropole GC/MS/MS, Agilent Technologies, Inc., Wilmington, DE, 2009, p. 14 Search PubMed.
  54. P. Fürst and C. Sandy, Determination of dioxin-like and non-dioxin-like polychlorinated biphenyl congeners in foodstuffs and animal feed using the Agilent 7000 Triple Quadrupole GC/MS/MS system, Agilent Technologies, Inc., Berkshire, United Kingdom, 2011, p. 12 Search PubMed.
  55. M. Pinchin, S. Verik and K. Brady, Analyzing wastewater effluents for PAHs and PBDEs using the Agilent 7000 Triple Quadrupole GC/MS/MS, Agilent Technologies, Inc., Cheadle, United Kingdom, 2012, p. 8 Search PubMed.
  56. I. de Dobbeleer, J. Gummersbach, H.-J. Huebschmann, A. Mayer and P. Silcock, Analyzing PBDEs in house dust samples with the Thermo Scientific TSQ Quantum XLS Ultra GC-MS/MS in EI-SRM mode, Thermo Fisher Scientific, Dreieich, Germany, 2012, p. 6 Search PubMed.
  57. L. Yeung and S. Mabury, Environ. Chem., 2016, 13, 102–110 CrossRef CAS.
  58. P. S. Howard, R. S. Boethling, W. F. Jarvis, W. Meylan and E. M. Michalenko, Handbook of Environmental Degradation Rates, Lewis Publications, Chelsea, MI, 1991, p. 776 Search PubMed.
  59. R. Ritter, M. Scheringer, M. MacLeod, C. Moeckel, K. C. Jones and K. Hungerbühler, Environ. Health Perspect., 2011, 119, 225–231 CrossRef CAS PubMed.
  60. F. Wania and D. Mackay, Environ. Pollut., 1999, 100, 223–240 CrossRef CAS PubMed.
  61. F. Wong, I. T. Cousins and M. MacLeod, Environ. Int., 2013, 59, 168–174 CrossRef CAS PubMed.
  62. M. J. Binnington and F. Wania, Environ. Toxicol. Chem., 2014, 33, 1415–1426 CrossRef CAS PubMed.
  63. R. Stewart, S. E. Campana, C. M. Jones and B. E. Stewart, Can. J. Zool., 2006, 84, 1840–1852 CrossRef.
  64. L. L. Loseto, G. A. Stern, D. Deibel, T. L. Connelly, A. Prokopowicz, D. R. S. Lean, L. Fortier and S. H. Ferguson, J. Marine Syst., 2008, 74, 1012–1024 CrossRef.
  65. J.-P. W. Desforges, P. S. Ross and L. L. Loseto, Environ. Toxicol. Chem., 2013, 32, 1132–1142 CrossRef CAS PubMed.
  66. G. M. Donaldson, B. M. Braune, A. J. Gaston and D. G. Noble, Arch. Environ. Contam. Toxicol., 1997, 33, 430–435 CrossRef CAS PubMed.
  67. J. Wolkers, I. C. Burkow, C. Lydersen, S. Dahle, M. Monshouwer and R. F. Witkamp, Sci. Total Environ., 1998, 216, 1–11 CrossRef CAS PubMed.
  68. A. Aguilar and A. Borrell, Mar. Environ. Res., 1991, 31, 275–286 CrossRef CAS.
  69. L.-J. Bao, J. You and E. Y. Zeng, Environ. Toxicol. Chem., 2011, 30, 1731–1738 CrossRef CAS PubMed.
  70. H. Xiao, N. Li and F. Wania, J. Chem. Eng. Data, 2004, 49, 173–185 CrossRef CAS.
  71. D. Mackay, W.-Y. Shiu, K. C. Ma and S. C. Lee, Handbook of Physical-Chemical Properties and Environmental Fate for Organic Chemicals, CRC Press, Boca Raton, 2nd edn, 2006, vol. I–IV, p. 919 Search PubMed.
  72. I. Takeuchi, N. Miyoshi, K. Mizukawa, H. Takada, T. Ikemoto, K. Omori and K. Tsuchiya, Mar. Pollut. Bull., 2009, 58, 663–671 CrossRef CAS PubMed.
  73. Y. Wan, X. Jin, J. Hu and F. Jin, Environ. Sci. Technol., 2007, 41, 3109–3114 CrossRef CAS PubMed.
  74. M. Roszko, K. Szymczyk and R. Jędrzejczak, J. Environ. Sci. Health, Part B, 2015, 50, 884–895 CrossRef CAS PubMed.
  75. N. D. Forsberg, D. Stone, A. Harding, B. Harper, S. Harris and M. M. Matzke, et al. , J. Agric. Food Chem., 2012, 60, 6899–6906 CrossRef CAS PubMed.
  76. Y. G. Ma, Y. D. Lei, H. Xiao and F. Wania, J. Chem. Eng. Data, 2010, 55, 819–825 CrossRef CAS.
  77. B. C. Kelly, M. G. Ikonomou, J. D. Blair, B. Surridge, D. Hoover, R. Grace and F. A. P. C. Gobas, Environ. Sci. Technol., 2009, 43, 4037–4043 CrossRef CAS PubMed.
  78. J. L. Reiner, S. G. O'Connell, A. J. Moors, J. R. Kucklick, P. R. Becker and J. M. Keller, Environ. Sci. Technol., 2011, 45, 8129–8136 CrossRef CAS PubMed.
  79. S. K. Ostertag, B. A. Tague, M. M. Humphries, S. A. Tittlemier and H. M. Chan, Chemosphere, 2009, 75, 1165–1172 CrossRef CAS PubMed.
  80. D. Brooke, A. Footitt and T. A. Nwaogu, Environmental risk evaluation report: perfluorooctanesulphonate (PFOS), Environment Agency, Chemicals Assessment Section, Wallingford, UK, 2004, p. 104 Search PubMed.
  81. K.-U. Goss, Environ. Sci. Technol., 2008, 42, 456–458 CrossRef CAS PubMed.
  82. Agency for Toxic Substances and Disease Registry, Minimal Risk Levels (MRLs), Atlanta, GA, 2016, p. 16, available from: http://www.atsdr.cdc.gov/mrls/pdfs/atsdr_mrls.pdf Search PubMed.
  83. Health Canada, Federal Contaminated Site Risk Assessment in Canada, Part II: Health Canada Toxicological Reference Values (TRVs) and Chemical-Specific Factors, Ministry of Health, Ottawa, ON, 2010, p. 69 Search PubMed.
  84. F. A. Verdonck, A. Souren, M. B. van Asselt, P. A. Van Sprang and P. A. Vanrolleghem, Integr. Environ. Assess. Manage., 2007, 3, 333 CrossRef.
  85. B. D. Laird, A. B. Goncharov and H. M. Chan, Environ. Int., 2013, 59, 33–40 CrossRef CAS PubMed.
  86. E. Archer, G. Pavela and C. J. Lavie, Mayo Clin. Proc., 2015, 90, 911–926 CrossRef PubMed.
  87. E. Archer, G. A. Hand and S. N. Blair, PLoS One, 2013, 8, e76632 CAS.
  88. N. V. Dhurandhar, D. Schoeller, A. W. Brown, S. B. Heymsfield, D. Thomas and T. I. A. Sørensen, et al. , Int. J. Obes., 2015, 39, 1109–1113 CrossRef CAS PubMed.


Electronic supplementary information (ESI) available. See DOI: 10.1039/c7em00167c
Current Address – MTM Research Centre, School of Science and Technology, Örebro University, S-701 82 Örebro, Sweden.

This journal is © The Royal Society of Chemistry 2017