A twenty-one year temporal trend of persistent organic pollutants in St. Lawrence Estuary beluga, Canada

Science of the Total Environment 485–486 (2014) 377–386

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

A twenty-one year temporal trend of persistent organic pollutants in St. Lawrence Estuary beluga, Canada Michel Lebeuf ⁎, Lena Measures, Michelle Noël, Meriem Raach, Steve Trottier Department of Fisheries and Oceans, Maurice Lamontagne Institute, 850 Route de la Mer, Mont-Joli, Québec, Canada

H I G H L I G H T S • • • • •

Most legacy POPs in beluga exhibited weak but significant decreasing linear trends. Trends of most legacy POPs were equivalently described by more than one model. Temporal trends of PBDEs were best described by a two-segment piecewise model. Biological variables did not significantly affect trends of POPs in beluga. Trends of POPs observed in beluga are in agreement with regulations.

a r t i c l e

i n f o

Article history: Received 5 December 2013 Received in revised form 26 February 2014 Accepted 20 March 2014 Available online 16 April 2014 Editor: F. Riget Keywords: Persistent organic pollutants Beluga whale Temporal trends Models Stable isotope St. Lawrence Estuary

a b s t r a c t Persistent organic pollutants (POPs) were measured in blubber from 144 stranded adult belugas (Delphinapterus leucas) found on the shores of the St. Lawrence Estuary (SLE) between 1987 and 2007. Temporal trends of POP concentrations (ln transformed) in beluga were described by three models, zero slope (ZS), linear (L) and twosegment piecewise (PW). Often two but sometimes all three models were equivalent in describing temporal trends based on Akaike's Information Criterion for small sample sizes. Over this 21-year time period, concentrations of most legacy POPs, including PCBs, DDTs and HCHs, exhibited relatively weak (≤11% per year) but significant decreasing trends in beluga. For PBDEs, temporal trends were best described by a PW model, characterizing a rapid increase until 1997–1998 followed by a slower increase for males and a steady-state for females. Potential cofactors such as blubber lipid content and carcass state of preservation did not show any significant temporal trends over the time period considered. Nitrogen stable isotope ratios (δ15N) in beluga liver, a proxy of trophic level, could not be associated to any effect on temporal trends of POP concentrations in beluga. Several POPs exhibited significant relationships with age of beluga and data were age-adjusted. Temporal trends of POP concentrations adjusted for age of beluga were reassessed but results were essentially identical as those obtained with the original POP data. Overall, POP temporal trends observed in SLE beluga are consistent with changes expected from regulations and restrictions in the use of these compounds in developed countries. Crown Copyright © 2014 Published by Elsevier B.V. All rights reserved.

1. Introduction The population of beluga (Delphinapterus leucas) from the St. Lawrence Estuary (SLE) is threatened according to the Committee on the Status of Endangered Wildlife in Canada (COSEWIC, 2004). The population is estimated at only about 10% of what it was at the beginning of the 20th century, due to intensive exploitation until 1979. Despite various actions to protect the SLE beluga population, abundance indices

⁎ Corresponding author at: Maurice Lamontagne Institute, Fisheries and Oceans Canada, 850 Route de la Mer, Mont-Joli, (Québec) G5H 3Z4, Canada. Tel.: + 1 418 775 0690; fax: + 1 418 775 0718. E-mail address: [email protected] (M. Lebeuf).

http://dx.doi.org/10.1016/j.scitotenv.2014.03.097 0048-9697/Crown Copyright © 2014 Published by Elsevier B.V. All rights reserved.

suggest that it is not recovering (Hammill et al., 2007). The SLE beluga population is concentrated at the mouth of the Saguenay River, where it occupies a relatively small area of 2790 km2 (Lemieux Lefebvre et al., 2012). The current summer home range has changed very little in the last 20 years, although the beluga range outside of summer is not well known. The SLE beluga live downstream of the Great Lakes and the St. Lawrence fluvial section, a densely populated, highly industrialized region of Canada and the United States. Although no single factor has been directly linked to the lack of recovery, this population inhabits a polluted ecosystem. The most recent analysis in the Species at Risk Recovery Strategy indicates that chemical pollution is considered a serious threat to the SLE beluga population (DFO, 2012). Since Sergeant (1980), several studies have reported high concentrations of persistent organic pollutants (POPs as defined by the Stockholm Convention, 2001) in blubber from the SLE beluga (Lebeuf, 2009). Only

378

M. Lebeuf et al. / Science of the Total Environment 485–486 (2014) 377–386

a few studies reported long term (N 10 years) time trends of either persistent organochlorine or organobromine compounds (Muir et al., 1996a, 1996b; Gouteux et al., 2003; Lebeuf et al., 2004, 2007). Results showed that legacy POPs such as polychlorinated biphenyls (PCBs) or various organochlorine pesticides (OCPs) are decreasing or do not exhibit significant temporal trends in SLE beluga. However, there is a need to provide an update of the current trends of these compounds in SLE beluga, especially for emerging polybrominated diphenyl ethers (PBDEs) that showed a strong increase during the 1980–90s. Temporal trend studies of POPs are useful, among other things, to report current contamination trends and to document the effects of government regulations or changes in industrial applications (Lebeuf and Nunes, 2005; Hickey et al., 2006). Several biological variables such as sex and age have been recognized to influence the contamination of beluga and other marine mammals with POPs. Females are generally less contaminated with POPs than males because of an efficient transfer of most of these chemicals to their offspring during gestation and lactation (e.g. Desforges et al., 2012). A relationship between age and concentration of POPs is reported in marine mammals, including beluga, and the relationship is often more frequently demonstrated in males than females (Stern et al., 2005). Martineau et al. (1987) and Muir et al. (1996b) reported age vs POP concentration relationships in stranded SLE beluga while Hobbs et al. (2003) observed the same relationship in biopsied SLE beluga. Consequently, age of beluga, or standard length sometimes used as a proxy, could bias temporal trends of POPs if not accounted for (Hoguet et al., 2013). Changes in the ecosystem could also affect the contamination of beluga with POPs without changing the overall contamination of the ecosystem. For instance, it has been suggested that changes in environmental factors such as climate variability or changes in trophic dynamics could drastically affect POP concentrations in fish (French et al., 2006). Several fish populations, including Atlantic cod (Gadus morhua), Rainbow smelt (Osmerus mordax), Atlantic tomcod (Microgadus tomcod), and American eel (Anguilla rostrata) have drastically declined in the SLE (Castonguay et al., 1994; Myers et al., 1997). These fish species are believed to be part of the beluga diet (Vladykov, 1946; Hickie et al., 2000). Changes in fish populations may result in a change in the diet of beluga which could have an important effect on the level of contaminants acquired by beluga. Consequently, changes in the ecosystem, in particular in the structure of the food web, influence the interpretation of temporal trends of contaminants (Hebert and Weseloh, 2006). One way to assess the degree of change in the diet of beluga is by measuring nitrogen stable isotopes (15N and 14N) in beluga tissues over time. It is recognized that nitrogen stable isotopes are transferred differently from prey to predators, and the ratio (15N/ 14 N) increases up the food chain (Minagawa and Wada, 1984; Kelly, 2000). To the best of our knowledge, no attempts have been made to link temporal trends of POPs in beluga with temporal trends in trophic position. The main objective of this study was to provide an update on temporal trends of POPs in SLE beluga over a 21-year time period, between 1987 and 2007, including PBDEs for which recent regulations came into force in North America (Ward et al., 2008). Different models including zero slope (ZS), linear (L) and two-segment piecewise (PW) were used to characterize POP temporal trends in SLE beluga. This study also examined the influence of biological variables, including sex, age and trophic position of beluga, on temporal trends of POPs in beluga. 2. Materials and methods 2.1. Previously published and new data PBDE data for 28 male and 26 female belugas were previously reported for the 1987–1999 time period by Lebeuf et al. (2004). This study reports new PBDE data for 16 males and 14 females for the

1987–1999 time period and extends this time period to 2007 with new PBDE data for 34 males and 24 females. PCB and OCP data were reported previously by Lebeuf et al. (2007) for 44 males and 42 females for the 1987–2002 time period. This study reports new PCB and OCP data for an additional 14 males and 4 females for the 1987–2002 time period and an extension of this time period to 2007 with new data for 20 males and 20 females.

2.2. Sampling Samples were obtained from 66 female and 78 male stranded belugas found on the shores of the SLE between 1987 and 2007. During that time period, about 300 beluga carcasses, mostly adults, were found on the shores of the SLE as part of the SLE beluga carcass monitoring program initiated in 1983. This study reports data on approximately half of the carcasses found on the shores of the SLE during the 21-year time period examined. Stranded belugas were found between March and December and the day of stranding expressed using the Julian calendar was calculated from the date of stranding recorded for each animal. Four animals were collected outside of the SLE. Based on their similar levels of POPs and nitrogen stable isotope ratios compared to other SLE belugas, these four individuals were considered to be from the SLE population and not from populations in the Canadian Arctic. Standard length of each animal was measured from the rostrum to the notch of the tail fluke. Carcass state of preservation was classified as good, fair or poor (codes 2 to 4) according to the classification of Geraci and Lounsbury (2005), although intermediate coding was also used (e.g. 2.5). Beluga carcasses collected prior to 1997 were not systematically coded except for those that were subject to a necropsy at the veterinary laboratory of the University of Montreal (Saint-Hyacinthe, Quebec). However, the large majority of beluga carcasses examined in this study were subject to a necropsy, namely 74% and 77% for males and females, respectively. The age of beluga was determined by counting growth layer groups (GLGs) on longitudinal tooth sections for each animal. The GLGs of some belugas may have been underestimated due to difficulty in reading worn growth layers. According to Stewart et al. (2006), the age of beluga in years corresponds to the number of GLGs and age in years is used in this study. In addition, teeth (available for 84% of beluga) were systematically read again in order to standardize the reading method, validate or correct previous readings. Only adult animals of 10 years or older were included in this study. A block of skin–blubber–muscle was collected at 60–70% of the body length from the rostrum, approximately midway between the spinal column and the mid-lateral region of each individual. The thickness of the blubber layer was recorded except for some belugas collected prior to 1997 that did not undergo a necropsy. Most blubber samples collected prior to 1997 were initially separated into three layers identified as distal (adjacent to the skin), middle (mid-blubber) and proximal (adjacent to the muscle), placed in individual solvent-rinsed glass jars and stored at −20 °C. In order to characterize the full depth of the blubber, layers were combined in equal quantities and homogenized before POP chemical analysis. For belugas sampled in 1997 and after, a block of blubber extending from the skin to the muscle was collected, wrapped in solvent-rinsed aluminum foil and placed in a sealed plastic bag, and stored at −20 °C until analysis. A subsample of the full depth of blubber (i.e. from the skin to the muscle) was taken from the block of blubber, homogenized and analyzed for POPs. Among belugas from which blubber was sampled for POP analysis, 31 females and 31 males also had their liver sampled for nitrogen stable isotope analysis. Liver samples were collected from carcasses, immediately wrapped in solvent-rinsed foil, placed in a sealed plastic bag and then stored at − 20 °C (Raach et al., 2011). Subsamples of liver were taken, homogenized and analyzed for nitrogen stable isotopes. Liver samples from beluga carcasses stranded before 1993 were not available.

M. Lebeuf et al. / Science of the Total Environment 485–486 (2014) 377–386

2.3. Chemical analysis Blubber samples were analyzed for several POPs, including PCBs, OCPs and PBDEs according to analytical methods used at the Maurice Lamontagne Institute, Department of Fisheries and Oceans Canada (Lebeuf et al., 2007; Raach et al., 2011). In brief, each blubber subsample was chemically dried with sodium sulfate before being transferred to a glass column. A single 13C12 PCB (IUPAC number 170) was added to the column before the extraction procedure. Lipids and lipophilic compounds were extracted from the sample with dichloromethane (DCM). The extraction solution was split into three parts. The first part (c.a. 40%) was saved as a backup, and the second part (c.a. 10%) was used to gravimetrically determine the lipid content of the sample. The third part of the extraction solution received a mixture of up to eight 13 C12 PCBs and eleven 13C or 2H OCPs and four 13C12 PBDEs as surrogate compounds, and was prepared for purification. Lipids were removed from the remaining extract by gel permeation chromatography using Bio-beads SX-3. The extract was further cleaned by elution through a two-layer column packed with neutral hydrated (5%) silica and alumina. The final extract was reduced in volume and spiked with an instrument performance solution containing two 13C12 PCBs (IUPAC numbers 111 and 189). Quantification of POPs was performed using a Varian 3400CX series gas chromatograph (Varian) equipped with a Varian Saturn IV ion trap, a Varian 1078 split/splitless programmable injector (5 μl injection volume) operated in splitless mode, and a Varian 8200CX autosampler or using a ThermoQuest Trace GC gas chromatograph (Thermo Fisher Scientific) equipped with a Finnigan PolarisQ ion trap, a ThermoQuest PTV split/splitless programmable injector (2 μl injection volume) operated in splitless mode, and a ThermoQuest AS2000 autosampler. For each sample, quantification of PCBs, OCPs and PBDEs was performed in separate runs. The chromatographic separation of the POPs was achieved using a 30 meter DB-5MS column (0.25 mm ID, 0.25 μm film thickness; J&W Scientific) with helium as the carrier gas. The ion source was operated in electron impact ionization mode and the ion trap in MS/MS mode. Concentrations of PCBs, OCPs and PBDEs were calculated using relative response factors determined from a five-point calibration curve. All reported concentrations were corrected for procedural losses of the surrogate compounds. Forty-one singly and coeluting PCB congeners were measured (IUPAC numbers 8, 15, 18, 28/31, 33, 37, 40, 44, 49, 52, 66/70, 74, 87, 95, 99, 101, 105, 110, 118, 128, 138, 149, 151, 153, 156, 170, 171, 177, 180, 183, 187, 191, 194,

379

195, 199, 205, 206, 208, and 209) and reported as PCBs. The OCPs analyzed were p,p′-DDT, p,p′-DDD, p,p′-DDE, o,p′-DDT, o,p′-DDD and o,p′DDE and reported as DDTs, α and γ-chlordane, cis and trans-nonachlor were reported as CHLs, α, γ-hexachlorocyclohexanes were reported as HCHs, tris(4-chlorophenyl)methane and tris(4-chlorophenyl)methanol were reported as TCPs and finally hexachlorobenzene (HCB) and Mirex were reported as single compounds. Eleven PBDE congeners were quantified (IUPAC numbers 17, 28, 47, 49, 66, 99, 100, 153, 154, 155, and 183) and reported as PBDEs. Attempts to measure highly brominated octa to deca-PBDEs were unsuccessful, as most samples were below detection limits (b 0.1 ng/g wet weight). The limits of quantification of individual PCB, PBDE, and OCP varied between 2 and 5 ng/g wet weight. 2.4. Quality assurance and quality control Samples were analyzed in batches of 10 samples, with one spiked blank and one certified reference material (SRM-1945), a blubber sample from a pilot whale (Globicephala sp.) obtained from the National Institute of Standards and Technology (NIST, 2007). Reference material replicates (n = 19) resulted in an average percent coefficient of variation of 9% for PCBs, 16% for OCPs, and 11% for ΣPBDEs, which indicates a fairly high degree of reproducibility. Several POPs with certified values in SRM-1945, including 29 PCBs, 14 OCPs and 5 PBDEs were systematically measured in all samples. The accuracy of the analyses was −1% for Σ26PCBs (IUPAC numbers 18, 28, 31, 44, 49, 52, 66, 70, 87, 95, 99, 101, 105, 110, 118, 128, 138, 149, 151, 153, 156, 170, 180, 183, 187, 194, 195, 206, and 209), −1% for DDTs, −4% for CHLs, −11% for HCHs, 4% for HCB, −5% for Mirex and −7% for Σ5PBDEs (IUPAC numbers 47, 99, 100, 153, and 154). 2.5. Nitrogen stable isotope analysis Samples of liver homogenates for stable isotope analysis were freeze dried and pulverized to a fine powder using a ball mill grinder. Samples were washed with methanol, dichloromethane and hexane in succession, and then centrifuged, discarding the supernatant. The lipid-free tissue was oven dried, ground using a mortar and pestle, and stored in desiccation vials until analyzed. Approximately 0.3 mg of dried tissue was used in the simultaneous analysis of stable N isotopes (15N and 14 N) on a Delta Plus Continuous Flow Stable Isotope Ratio Mass

Table 1 Biological variables for adult SLE beluga, comparison analyses between males and females and linear regressiona analyses between biological variables and year of stranding (1987–2007). Biological variable

Males Mean ± Std Dev or Median (n)

Age (year) (M = F)c Standard length (m) (M N F) Blubber lipid (%) (M = F) Blubber thickness (cm) (M = F) Carcass state of preservationd (M N F) Day of strandinge (M = F) Liver δ15N (‰)f (M N F) a

Females Range

pb (n)

r2

Mean ± Std Dev or Median (n) 43.4 ± 12.5 (65)

p (n)

r2

12–63

0.903 (65)

b0.01

3.1–4.0

0.389 (66)

0.01

64.2–97.9

0.086 (58)

0.05

Range

42.8 ± 12.8 (75)

12–72

0.988 (75)

b0.01

4.1 ± 0.2 (78)

3.0–4.6

0.325 (76)

0.01

3.6 ± 0.2 (66)

63.9–97.5

0.211 (76)

0.02

93.3 (66)

7.2 ± 1.9 (66)

3.0–13

0.304 (65)

0.02

6.5 ± 2.0 (48)

1.0–12

0.062 (46)

0.08

3.5 (78)

2.0–4.0

0.299 (78)

0.01

3.0 (66)

2.0–4.0

0.101 (66)

0.04

211 (78)

87–342

0.261 (78)

0.02

216 (66)

96–363

0.010 (66) ↓

0.10

15.5–18.9

0.381 (28)

0.03

16.4 ± 0.7 (31)

14.9–17.6

0.001 (30) ↑

0.33

91.3 (78)

16.9 ± 0.9 (31)

Biological variable = b ∗ year of stranding + a. Probability of a slope that significantly differs from zero (number of data points included in the regression). c Results of comparison between males (M) and females (F) in parentheses for each above variable (i.e. age, standard length, blubber thickness, liver δ15N (t-test) and other variables (Mann–Whitney)). d Coded according to the classification of Geraci and Lounsbury (2005). e Day of stranding expressed using Julian calendar. f Nitrogen stable isotope ratio in liver for the 1993–2007 time period. b

380

M. Lebeuf et al. / Science of the Total Environment 485–486 (2014) 377–386

Fig. 1. Temporal trends of age (year), standard length (m), blubber thickness (cm), blubber lipid content (%), carcass state of preservation, Julian day of stranding and liver δ15N (‰). Values considered as outliers (see text) were excluded from graphs (see Table 1). The statistically significant time trend is represented by a regression line (solid for males and dotted for females).

Spectrometer (Thermo Finnigan) coupled to a Carlo Erba Elemental Analyzer (CHNS-O EA1108). All results were expressed as 15N/14N ratios in conventional delta notation (δ15N) relative to atmospheric nitrogen. The analytical precision for δ15N, assessed by the repeated analysis of 7 samples, was 3‰ or less. 2.6. Data analysis Contaminant concentrations were expressed on a lipid weight basis and were natural logarithm (ln)-transformed to normalize the data

distribution prior to statistical analysis. Comparison of POP concentrations between males and females was assessed using a two-sample t test. Temporal trends of POP concentrations were independently assessed for female and male belugas. Temporal trend analyses of lntransformed POP concentrations in blubber from beluga were assessed using three distinct models of increasing complexity, a zero-slope (ZS) regression, a simple linear (L) regression and a non-linear two-segment piecewise (PW) regression model. Only models with normally distributed residuals were retained. Regression models were assessed after removing outliers identified on the basis of studentized residuals

M. Lebeuf et al. / Science of the Total Environment 485–486 (2014) 377–386

381

Table 2 Basic statistics for POP concentrations (μg/g lipid for PCBs and DDTs and ng/g lipid for the other POPs) in blubber from adult SLE beluga and comparisona analyses between males and females (1987–2007). POPs

Males

PCBs DDTs CHLs HCHs TCPs HCB Mirex PBDEs a b c

Mean ± Std Dev

Range

Geometric mean (nc)

Mean ± Std Dev

Range

Geometric mean (n)

81.1 105 4791 58.2 107 613 955 421

3.6–238 0.7–297 292–9455 18.6–270 2.7–260 35.0–2137 37.4–4003 32.9–1308

68.5 (78) 74.8 (78) 4303 (78) 49.5 (78) 85.6 (78) 518 (78) 739 (78) 297 (78)

20.0 16.8 1436 65.0 21.9 205 739 419

0.4–115 0.3–246 222–7364 13.8–226 1.3–160 24.9–1881 5.8–3993 18.4–1875

13.1 (66) 6.2 (66) 1059 (66) 53.6 (66) 13.6 (64) 152 (66) 481 (66) 224 (66)

± ± ± ± ± ± ± ±

40.8 62.7 1789 38.9 51.9 351 679 311

± ± ± ± ± ± ± ±

20.7 40.0 1337 42.6 28.3 311 699 392

b0.001; M b0.001; M b0.001; M 0.43 b0.001; M b0.001; M 0.010; M 0.15

Comparison between males and females was based on ln-transformed data (t-test). Probability of a difference of POP concentrations between males (M) and females (F); sex with the highest values indicated. Number of samples with detectable concentration.

exceeding the value of 3. The ZS regression model was used to test the scenario where POP concentrations did not change over time whereas the L regression model indicated a constant rate of change of POP (because POP concentrations were ln-transformed) during the 1987–2007 time period. The PW regression model assumed that the temporal trend of POP concentrations was composed of two time periods, the first one starting at the earliest year in the time-series until the breakpoint year and the second one starting from the breakpoint year until the latest year in the series. The breakpoint year was determined with the quasiNewton method using an iterative convergence procedure by varying the breakpoint annually within the time-series but with a minimum of three years in each time period. Akaike's Information Criterion for small sample sizes (AICc) was calculated for each regression model and used to compare models by determining ΔAICc (i.e. AICc value of the model − AICc value of best fit model). Models with ΔAICc b 2 were considered indistinguishable from each other (Burnham and Anderson, 2002). Linear models, or time periods in the case of the PW model, were further characterized from the estimated slope (b) and intercept (a) [ln[POP] = b ∗ year of stranding + a] by calculating the half-life time (t ½ ) or the doubling time (t2), using the expression [ln(2) ∕ b]. Alternatively the rate of decrease or increase of contaminant concentrations in percent per year was calculated using the expression [(1−e−b) ∗ 100]. Biological variables characterized by semi-quantitative values, including carcass state of preservation or quantitative values not normally distributed, namely blubber lipid content and date of stranding converted to day of year (Julian calendar), were compared between males and females using a Mann–Whitney (M–W) analysis. Values for normally

Table 3 Results of regression modelsa for temporal trend analyses of POP concentrations in blubber from adult male and female SLE belugas (1987–2007). POPs

Males

Females

N

Modelsb

n

Models

PCBs

74

65

L (↓); PW (2004; n.s., n.s.)

DDTs CHLs

73 72

ZS; L (n.s.); PW (1990; n.s., n.s.) L (↓); PW (2005; ↓, n.s.) ZS

65 66

HCHs TCPs HCB Mirex

78 72 76 76

L (↓); PW (1990; n.s., ↓) ZS; L (↓); PW (1990; n.s, n.s.) L (↓); PW (1992; n.s., ↓) ZS; PW (1990; n.s., n.s.)

66 63 66 64

PBDEs

78

PW (1997; ↑, ↑)

65

L (↓); PW (2004; ↓, n.s.) ZS; L (↓); PW (1999; n.s., n.s.) L (↓); PW (1993; ↓, ↓) L (↓); PW (1999; n.s., n.s.) ZS; L (↓); PW (1990; n.s., n.s.) ZS; L (↓); PW (1990; n.s., n.s.) PW (1998; ↑, n.s.)

a

p; M/Fb

Females

Zero-slope (ZS); linear (L); piecewise (PW; breakpoint year); linear segment trend: n.s. (not significantly different from zero); increasing (↑); decreasing (↓). b Model in bold letters is the best fit, i.e. lowest AICc (Akaike's Information Criterion for small sample sizes), models reported are equivalent based on ΔAICc b 2 relative to the best fit.

distributed variables, namely age, standard length, blubber thickness and liver δ15N, were compared using a two-sample t test (t). Variables including age, standard length, thickness of blubber, carcass state of preservation, day of stranding, lipid content of blubber and δ15N in liver were assessed for male and female belugas as a function of year of stranding using a linear regression model after removing outliers. Significant relationships between POP concentration and age were observed and an adjustment (correction) was made according to the method described by Hebert and Weseloh (2006). Briefly, the linear regression of the natural logarithm of measured POP concentrations against age was obtained after removing outliers. The residuals of the regression analysis were determined and added to the grand mean POP concentration to obtain the final age-adjusted POP concentrations. Temporal trends of adjusted contaminant concentrations were reassessed according to the three models described above. All statistical analyses were performed using Systat 10 software (SPSS, 2000) with statistical significance set at α = 0.05. 3. Results 3.1. Biological variables and nitrogen stable isotope ratio There was no statistical difference between males and females in mean or median values for most biological variables examined including age, blubber lipid content, blubber thickness and the day of stranding (Table 1). About 10% of males (n = 8) and females (n = 7) exhibit blubber lipid content less than 80%. A possible link between low blubber lipid content and other biological parameters was tested. Results indicate that no biological variable, including blubber thickness and the state of preservation of the carcass, was significantly different (t; p N 0.05) between belugas of low (b 80%) and high (N 80%) blubber lipid content. However, statistical analysis showed that males were on average longer (t; p b 0.001) and their carcass was in a poorer state of preservation than females (M–W; p = 0.009). Males had significantly higher δ15N ratios in liver than females (Table 1). Temporal trends of biological variables were examined by linear regression analysis for both sexes. There were no statistically significant temporal trends in any biological variables examined except day of stranding and δ15N for females (Table 1, Fig. 1). Hence, no significant temporal trends were found in age, standard length, lipid content, blubber thickness and carcass state of preservation for both sexes, and day of stranding for males over the 21-year time period examined (1987–2007). 3.2. POP concentrations and temporal trends Mean concentrations and standard deviation of individual POP or group of POPs are calculated for male and female belugas for the entire time period examined (Table 2). Geometric means were much lower

382

M. Lebeuf et al. / Science of the Total Environment 485–486 (2014) 377–386

than (arithmetic) means, up to 63%, indicating the influence of some extreme values in mean concentrations of POPs. For most POPs, males exhibited significantly higher concentrations in their blubber than females. However, mean concentrations of HCHs and PBDEs in males and females were not significantly different during the 1987–2007 time period.

Temporal trends of POP data were assessed by examining three distinct regression models (Table 3). The L regression model was generally the best fitted model to describe POP data, especially for females, but frequently ZS and/or PW regression models were equivalent according to the ΔAICc. Temporal trends (slope) of legacy POPs were characterized by rates of decreasing concentrations equal or slower than 11% per year.

Fig. 2. Temporal trends of POP concentrations in blubber from adult male and female belugas between 1987 and 2007. Statistically significant time trends are represented by regression lines (solid for males and dotted for females) with depuration/accumulation rates (% per year) and corresponding half-life/doubling time in years in parentheses.

M. Lebeuf et al. / Science of the Total Environment 485–486 (2014) 377–386

383

Table 4 Basic statistics for age adjusted POP concentrations (μg/g lipid for PCBs and DDTs and ng/g lipid for the other POPs) in blubber from adult SLE beluga and comparisona analyses between males and females (1987–2007). POPs

Males

PCBs DDTs CHLs TCPs HCB Mirex a b c

Mean ± Std Dev

Range

Geometric mean (nc)

84.0 111 4995 109 619 937

24.2–220 23.0–328 2029–9352 37.0–282 131–2193 195–3113

77.0 (71) 96.8 (68) 4770 (70) 100 (71) 535 (73) 776 (73)

± ± ± ± ± ±

37.3 58.4 1507 48.9 328 610

Mean ± Std Dev 17.0 ± 44.9 21.8 ± 30.2 735 ± 632

Range See Table 2 0.4–266 See Table 2 2.6–190 See Table 2 55.0–4113

Geometric mean (n) 6.2 (65) 13.5 (62) 548 (63)

b0.001; M b0.001; M b0.001; M b0.001; M b0.001; M 0.005; M

Comparison between males (M) and females (F) was based on ln-transformed data (t-test). Probability of a difference in POP concentrations between males and females with the sex with the highest values indicated. Number of samples with detectable concentration.

Temporal trends for some POP concentrations in males, namely PCBs, CHLs and Mirex, were not significantly different than zero. Three POP temporal trends were best described by a single model, namely ZS for CHLs in males and PW for PBDEs in males and females. Results of the best fitted model for each POP along with their rate of decrease or increase, when relevant, are presented in Fig. 2. Emerging PBDEs exhibited a rapid rate of increasing concentrations, exceeding 20% per year, in the first half of the time period examined but the rate of increasing concentrations in males diminished during the last 10 years whereas the PBDE temporal trend in females leveled off and was not significantly different than zero during that time period.

3.3. Age-adjusted POP concentrations and temporal trends The effect of age on POP concentrations in beluga was examined for both sexes by linear regression analyses. Statistically significant linear relationships were observed for 6 POPs in males and 3 POPs in females. POP concentrations adjusted for age of beluga were calculated according to the method described by Hebert and Weseloh (2006). The statistical significance of the difference in mean concentrations of POPs between males and females, using age-adjusted POP concentrations, was unchanged (Table 4). Temporal trends were also reassessed using age-adjusted POP concentrations (Table 5). For ΣDDTs in males, the best fit model was changed from PW to L, although both models were considered equivalent based on ΔAICc. For the other POPs in males, the same models were retained and the best fitted models were unchanged except that model L was added to ZS and PW equivalent models for Mirex. Results of regression models were unchanged for females. Best fitted models are presented in Fig. 3 for relationships of age-adjusted POP concentrations over time along with rates of decrease when significant.

Table 5 Results of regression modelsa for temporal trend analyses of age-adjusted POP concentrations in blubber from adult male and female SLE belugas (1987–2007). POPs

PCBs DDTs CHLs TCPs HCB Mirex a

p; M/Fb

Females

Males

Females b

N

Models

71 68 70 71 72 73

ZS; L (↓); PW (1990; n.s., n.s.) L (↓); PW (2005; ↓ n.s.) ZS ZS; L (n.s.); PW (1990; n.s, n.s.) L (↓); PW (1998; ↓, ↓) ZS; L (n.s.); PW (1990; n.s., n.s.)

n 64 62 63

Models See Table 3 L (↓); PW (2004; ↓ n.s.) See Table 3 L (↓); PW (1999; n.s., n.s.) See Table 3 ZS; L (↓); PW (1990; n.s., n.s.)

Zero-slope (ZS); linear (L); piecewise (PW; breakpoint year); linear segment trend: n.s. (not significantly different from zero); increasing (↑); decreasing (↓). b Model in bold letters is the best fit, i.e. lowest AICc (Akaike's Information Criterion for small sample sizes), models reported are equivalent based on ΔICc b 2 relative to the best fit.

4. Discussion 4.1. Temporal trends of POPs in beluga Updated temporal trends of legacy POPs in SLE beluga reported from this study were very similar to those reported previously (Lebeuf et al., 2007). Most temporal trends of legacy POPs exhibited decreasing concentrations at slow rates best described by the linear (L) model (Fig. 2). These results are also in agreement with decreasing trends of POPs reported in sediment cores collected in the distribution area of the SLE beluga (Lebeuf and Nunes, 2005). This study also compares the fitting of POP data among three models namely, ZS, L and PW. For legacy POPs, mostly regulated in the 1970s, current changes in concentrations are more difficult to detect compared to the marked decline in concentrations of POPs generally observed soon after a change in regulation. This situation likely applies to TCPs for which the origin has not been clearly identified but it is presumably related to the production of DDT (Buser, 1995). This might explain that for some POPs, particularly in males, concentration trends were best described by the ZS model (Table 3). The statistical power of a time-series depends on several variables including not only its length and the number of samples available in each year but also the between-year variation which could hide temporal trends of POP concentrations (Rigét et al., 2010). On several occasions, the significance of the temporal trend was different among models considered equivalent according to the ΔAICc (Tables 3 and 5). This indicates that the temporal trend of POP concentrations in SLE beluga is relatively uncertain for many legacy POPs. The PW model is of a particular interest because it provides the possibility of identifying a change in temporal trend over the time-series (French et al., 2006, 2011). On several occasions the PW model was the best fitted model or as an equivalent best fitted model with a breakpoint located at the beginning or at the end of the time-series (Table 3). This may indicate an artifact since the PW model selected a short segment (three-four years) which permits reduction of the sum of squares of the residuals without leading to significant temporal trends. This seems to be the case for DDTs in males where the last time period (segment) covers only three years (the minimum required) and shows high variability in reported concentrations (Fig. 2). The overall rate of decreasing DDT concentrations in males that resulted from the L model is 5.6% per year, which is higher than that obtained from the PW model. The most striking change in POP temporal trends in SLE beluga was observed for PBDEs. PBDE data are best described by a two-segment piecewise (PW) model which seems particularly appropriate for more recently regulated POPs. The PW model indicated that a change in temporal trends occurred at the end of the 1990s for both males and females. Unexpectedly, the breaking point in PBDE temporal trends occurred several years before the adoption of Canadian and US legislative regulations for this group of chemicals (Ward et al., 2008). Possible explanations are earlier regulations in some European countries

384

M. Lebeuf et al. / Science of the Total Environment 485–486 (2014) 377–386

Fig. 3. Temporal trends of age-adjusted POP concentrations in blubber from adult male and female belugas between 1987 and 2007. Statistically significant time trends are represented by regression lines with (solid for males and dotted for females) depuration/accumulation rates (% per year) and corresponding half-life/doubling time in years in parentheses.

(e.g. Germany) leading to a reduction of PBDEs through imported products, and/or the switch to alternative flame retardants by the industry prior to regulation such as voluntary phase-out of PBDE production in North America (Ward et al., 2008). 4.2. Effects of biological variables on temporal trends of POPs in beluga The elimination of POPs by maternal transfer is selective and depends on their Kow (n-octanol/water partitioning coefficient) (Ikonomou and Addison, 2008; Frouin et al., 2012; Desforges et al., 2012). Most POPs are characterized by intermediate Kow, ranging from 105 to 107, and exhibit significant differences in concentrations between males and females as reported in this study (Table 1). For low Kow POPs such as HCHs, however, elimination rates are fast, preventing long term accumulation in beluga and resulting in similar concentrations in males and females. For high Kow POPs such as Mirex, concentrations are generally different between males and females but the difference is small (Table 1) because these POPs have slow elimination rates, including during maternal transfer (Hoguet et al., 2013). Lower or similar concentrations of POPs in females compared to males are not expected to affect temporal trends. However, temporal trends of POPs in females are also affected by the intensity of POP elimination (maternal transfer) relative

to their acquisition. For instance, similar PBDE concentrations in females and males for the 1988–1999 time period (Fig. 2) were explained by a much higher acquisition of PBDEs by females than their elimination through lactation (Lebeuf et al., 2004). Since 1999, however, the rate of PBDE acquisition is slower in males and not significant in females. This difference in temporal trends between males and females is consistent with the expected difference in PBDE concentrations between males and females due to maternal transfer of POPs (Table 3). Similarly, higher concentrations of POPs transferred to the first-born calf than to subsequent calves, as shown for bottlenose dolphins (Tursiops truncatus) (Wells et al., 2005) or the lack of transfer in older and less reproductive females (Hickie et al., 2000), is expected to result in distinctive temporal trends of POP concentrations between males and females. Difference of standard length between adult males and females is commonly observed in beluga populations. For instance, Muir et al. (1990) reported longer males than females in most beluga populations from Canadian waters and length data from SLE beluga collected in 1986–1987 had similar mean and range values compared to the beluga examined in this study. To the best of our knowledge, there is no demonstrated link between the length of a beluga and its contamination, assuming animals of similar ages, and growing and feeding conditions. Male carcasses were in a significantly poorer state of preservation

M. Lebeuf et al. / Science of the Total Environment 485–486 (2014) 377–386

than females (Table 3) as previously reported by Lebeuf et al. (2007) from a smaller number of beluga carcasses collected over a shorter time period (1987–2002). Information on the effects of carcass state of preservation on POP concentrations in marine mammals is scarce. Recently, Borrell et al. (2007) monitored POP concentrations in blubber from harbor porpoises (Phocoena phocoena) sampled repeatedly over a 48 hour period post-mortem and found no significant change. Therefore, the effect of the state of preservation of a carcass with respect to blubber POP concentrations is likely negligible. Despite differences in standard length and carcass state of preservation between males and females, there were no significant temporal trends suggesting that these biological variables could influence temporal trends of POPs in SLE beluga. There were, however, two biological variables that showed significant temporal trends during the examined time period (Table 1). For instance, female carcasses were collected on average about 60 days earlier in 2007 than 20 years ago. To the best of our knowledge, there is no study indicating that blubber contamination by POPs in female beluga varies seasonally. Therefore, temporal trends on day of stranding of carcasses were not considered to affect POP contamination in females. A significant increase in δ15N values was observed in liver of female beluga (Table 1). Several studies have shown that δ15N values in biota can be used as a proxy of trophic position in a food web, ultimately resulting from the diet (Kelly, 2000). In addition, positive relationships are commonly reported between POP concentrations and δ15N values (Jardine et al., 2006). The diet of beluga has likely changed over the years as changes in the trophic structure of the ecosystem have been observed (Savenkoff et al., 2007). Baseline changes of δ15N in the SLE ecosystem should also be considered to explain the temporal trend of δ15N observed in female beluga (Fry, 1999). It is still unclear, however, why a significant temporal trend of δ15N was observed in female but not in male beluga.In a recent study, Lesage (2013) reported δ15N values in another tissue (muscle) of adult SLE beluga, but no significant trend was observed for both males and females in the 1988–2007 time period. Comparing temporal trends of δ15N between liver and muscle of beluga and the lack of information on δ15N baseline trends in the SLE do not permit associating δ15N data to any specific effect on temporal trends of POP concentrations in beluga. Concentrations of POPs with slow or intermediate elimination rates are expected to increase with the age of animals, especially for longlived species such as the beluga. However, a positive relationship between POP concentrations and age can sometimes be difficult to observe because of the high variability in the load of POPs transferred from the mother. The load of POPs transferred to calves depends on the birth order and interval, resulting in a large range of POP contamination not only in calves but also in adults (Stern et al., 2005). In this study, several significant relationships in POP concentrations vs age were observed despite the inherent sources of variability in POP concentrations applicable to the analyzed data. Likely for these reasons, males exhibited more often positive and significant relationships than females, consistent with observations reported in Canadian Arctic beluga populations (Stern et al., 2005). However, the age adjustment on POP temporal trends had only a significant effect on the best fitted model for DDTs in males but changes in rates of decrease were negligible for all ageadjusted temporal trends of POPs (Fig. 3). 5. Conclusions Over the 1987–2007 time period, concentrations of legacy POPs either slowly decreased or remained constant in SLE beluga. These results are in agreement with the behavior of persistent compounds transferred to the next generation. Concentrations of legacy POPs in blubber from males were often best fitted by the zero-slope (ZS) model, although in many cases, linear (L) and piecewise (PW) models were considered equivalent. For PBDEs, a significant change in temporal trends, best described by the PW model, was observed at the end of the 1990s. This change in temporal

385

trend is in agreement with international regulations and/or the switch to alternative flame retardants by the industry prior to regulations in North America. Concentrations of POPs are expected to continue to be detected in SLE beluga into the foreseeable future. Acknowledgments Thanks to Robert Michaud, Pierre Béland, Daniel Martineau and Stéphane Lair for their contribution in collecting samples and biological data from stranded beluga. Age determination of beluga resulted from the meticulous work of Yves Morin and Véronique Lesage. Stable isotope analyses were performed at the Environmental Isotope Lab (University of Waterloo, ON, Canada). Thanks to Bruno Gouteux, Brigitte Desrosiers and Sylvie St-Pierre for their help in POP analyses. References Borrell A, Aguilar A, Zeljkovic S, Brouwer A, Besselink HT, Koopman H, et al. Post-mortem stability of blubber DLCs, PCB and tDDT in by-caught harbour porpoises (Phocoena phocoena). Mar Pollut Bull 2007;54:1663–72. Burnham KP, Anderson DR. Model selection and multimodel inference. 2nd ed. New York: Springer; 2002. Buser H-R. DDT, a potential source of environmental tris(4-chlorophenyl)methane and tris(4-chlorophenyl)methanol. Environ Sci Technol 1995;29:2133–9. Castonguay M, Hodson PV, Couillard CM, Eckersley MJ, Dutil J-D, Verreault G. Why is recruitment of the American eel declining in the St. Lawrence River and Gulf? Can J Fish Aquat Sci 1994;1:479–88. COSEWIC (Committee on the Status of Endangered Wildlife in Canada). Beluga whale St. Lawrence Estuary population, last designation, May 2004. http://www.speciesatrisk. gc.ca/search/speciesDetails_e.cfm?SpeciesID=102, 2004. [Accessed April, 2013]. Desforges J-PW, Ross PS, Loseto LL. Transplacental transfer of polychlorinated biphenyls and polybrominated diphenyl ethers in arctic beluga whales (Delphinapterus leucas). Environ Toxicol Chem 2012;31:296–300. DFO (Department of fisheries and Oceans). Recovery strategy for the beluga whale (Delphinapterus leucas) St. Lawrence Estuary population in Canada. Species at risk act, recovery strategy series. Ottawa: Fisheries and Oceans Canada; 2012 [88 pp. + X pp. Available at http://www.dfo-mpo.gc.ca/species-especes/species-especes/ belugaStLa-eng.htm]. French TD, Campbell LM, Jackson DA, Casselman JM, Scheider WA, Hayton A. Long-term changes in legacy trace organic contaminants and mercury in Lake Ontario salmon in relation to source controls, trophodynamics, and climate variability. Limnol Oceanogr 2006;51:2794–807. French TD, Petro S, Reiner EJ, Bhavsar SP, Jackson DA. Thirty-year time series of PCB concentrations in a small invertivorous fish (Notropis hudsonius): an examination of post-1990 trajectory shifts in the lower Great Lakes. Ecosystems 2011;14:415–29. Frouin H, Lebeuf M, Hammill M, Fournier M. Transfer of PBDEs and chlorinated POPs from mother to pup during lactation in harp seals Phoca groenlandica. Sci Total Environ 2012;417–418:98–107. Fry B. Using stable isotopes to monitor watershed influences on aquatic trophodynamics. Can J Fish Aquat Sci 1999;56:2167–71. Geraci JR, Lounsbury VJ. Marine mammals ashore: a field guide for strandings. 2nd ed. Baltimore, MD: National Aquarium in Baltimore; 2005 [372 pp.]. Gouteux B, Lebeuf M, Muir DCG, Gagné JP. Levels and temporal trends of toxaphene congeners in beluga whales (Delphinapterus leucas) from the St. Lawrence Estuary, Canada. Environ Sci Technol 2003;37:4603–9. Hammill MO, Measures LN, Gosselin J-F, Lesage V. Lack of recovery in the St. Lawrence Estuary beluga. DFO Can Sci Advis Sec Res Doc 2007;026. [28 pp. Available at http://www.dfo-mpo.gc.ca/CSAS/Csas/DocREC/2007/RES2007_026_e.pdf]. Hebert CE, Weseloh DVC. Adjusting for temporal change in trophic position results in reduced rates of contaminant decline. Environ Sci Technol 2006;40:5624–8. Hickey JP, Batterman SA, Chernyak SM. Trends of chlorinated organic contaminants in great lakes trout and walleye from 1970 to 1998. Arch Environ Contam Toxicol 2006;50:97–110. Hickie BE, Kingsley MCS, Hodson PV, Muir DCG, Béland P, Mackay D. A modelling-based perspective on the past, present, and future polychlorinated biphenyl contamination of the St Lawrence beluga whale (Delphinapterus leucas) population. Can J Fish Aquat Sci 2000;57(Suppl. 1):101–12. Hobbs KE, Muir DCG, Michaud R, Béland P, Letcher RJ, Norstrom R. PCBs and organochlorine pesticides in blubber biopsies from free-ranging St. Lawrence River Estuary beluga whales (Delphinapterus leucas), 1994–1998. Environ Pollut 2003;122:291–302. Hoguet J, Keller JM, Reiner JL, Kucklick JR, Bryan CE, Moors AJ, et al. Spatial and temporal trends of persistent organic pollutants and mercury in beluga whales (Delphinapterus leucas) from Alaska. Sci Total Environ 2013;449:285–94. Ikonomou MG, Addison RF. Polybrominated diphenyl ethers (PBDEs) in seal populations from eastern and western Canada: an assessment of the processes and factors controlling PBDE distribution in seals. Mar Environ Res 2008;66:225–30. Jardine TD, Kidd KA, Fisk AA. Applications, considerations, and sources of uncertainty when using stable isotope analysis in ecotoxicology. Environ Sci Technol 2006;40: 7501–11. Kelly JF. Stable isotopes of carbon and nitrogen in the study of avian and mammalian trophic ecology. Can J Zool 2000;78:1–27.

386

M. Lebeuf et al. / Science of the Total Environment 485–486 (2014) 377–386

Lebeuf M. La contamination du béluga de l'estuaire du Saint-Laurent par les polluants organiques persistants en revue. Rev Sci Eau 2009;22:199–233. Lebeuf M, Nunes T. PCBs and OCPs in sediment cores from the Lower St. Lawrence Estuary, Canada: evidence of fluvial inputs and time lag in delivery to coring sites. Environ Sci Technol 2005;39:1470–8. Lebeuf M, Gouteux B, Measures L, Trottier S. Levels and temporal trends (1988–1999) of polybrominated diphenyl ethers in beluga whales (Delphinapterus leucas) from the St. Lawrence Estuary, Canada. Environ Sci Technol 2004;38:2971–7. Lebeuf M, Noël M, Trottier S, Measures L. Temporal trends (1987–2002) of persistent, bioaccumulative and toxic (PBT) chemicals in beluga whales (Delphinapterus leucas) from the St. Lawrence Estuary, Canada. Sci Total Environ 2007;383:216–31. Lemieux Lefebvre S, Michaud R, Lesage V, Berteaux D. Identifying high residency areas of the threatened St. Lawrence beluga whale from fine-scale movements of individuals and coarse-scale movements of herds. Mar Ecol Prog Ser 2012;450:243–57. Lesage V. Trends in the trophic ecology of St Lawrence beluga (Delphinapterus leucas) over the period 1988–2012, based on stable isotope ratios. DFO Can Sci Advis Sec Res Doc 2013;126. [iv + 26 p.]. Martineau D, Béland P, Desjardins C, Lagacé A. Levels of organochlorine chemicals in tissues of beluga whales (Delphinapterus leucas) from the St. Lawrence Estuary, Québec, Canada. Arch Environ Contam Toxicol 1987;16:137–47. Minagawa M, Wada E. Stepwise enrichment of 15N along food chains: further evidence and the relation between delta 15N and animal age. Geochim Cosmochim Acta 1984;48:1135–40. Muir DGC, Ford CA, Stewart REA, Smith TG, Addison RF, Zinck ME, et al. Organochlorine contaminants in belugas, Delphinapterus leucas, from Canadian waters. In: Smith TG, St Aubin DJ, Geraci JR, editors. Advances in research on the beluga whale, Delphinapterus leucas, 224. Can Bull Fish Aquat Sci; 1990. p. 165–90. Muir DGC, Koczanski K, Rosenberg B, Béland P. Persistent organochlorines in beluga whales (Delphinapterus leucas) from the St. Lawrence River Estuary — II. Temporal trends, 1982–1994. Environ Pollut 1996a;93:235–45. Muir DGC, Koczanski K, Rosenberg B, Béland P. Persistent organochlorines in beluga whales (Delphinapterus leucas) from the St. Lawrence River Estuary — I. Concentrations and patterns of specific PCBs, chlorinated pesticides and polychlorinated dibenzo-p-dioxins and dibenzofurans. Environ Pollut 1996b;93:219–34.

Myers RA, Hutchings JA, Barrowman NJ. Why do fish stocks collapse? The example of cod in Atlantic Canada. Ecol Appl 1997;7:91–106. NIST (National Institute of Standards & Technology). Certificate of analysis, standard reference material 1945, organics in whale blubber; 2007 [10 pp.]. Raach M, Lebeuf M, Pelletier É. PBDEs and PCBs in the liver of the St Lawrence Estuary beluga (Delphinapterus leucas): a comparison of levels and temporal trends with the blubber. J Environ Monit 2011;13:649–56. Rigét F, Bignert A, Braune B, Stow J, Wilson S. Temporal trends of legacy POPs in Arctic biota, an update. Sci Total Environ 2010;408:2874–84. S.P.S.S. Inc. SYSTAT 10 for Windows. Chicago: SPSS Inc.; 2000. Savenkoff C, Castonguay M, Chabot D, Hammill MO, Bourdages H, Morissette L. Changes in the northern Gulf of St. Lawrence ecosystem estimated by inverse modelling: evidence of a fishery-induced regime shift? Estuar Coast Shelf Sci 2007;73:711–24. Sergeant DE. Levels of mercury and organochlorine residues in tissues of sea mammals from the St. Lawrence estuary. Conseil International pour l'Exploration de la Mer (CIEM); 1980 [C.M./E:55]. Stern GA, Macdonald CR, Armstrong D, Dunn B, Fuchs C, Harwood L, et al. Spatial trends and factors affecting variation of organochlorine contaminants levels in Canadian Arctic beluga (Delphinapterus leucas). Sci Total Environ 2005;351–352:344–68. Stewart REA, Campana SE, Jones CM, Stewart BE. Bomb radiocarbon dating calibrates beluga (Delphinapterus leucas) age estimates. Can J Zool 2006;84:1840–52. Stockholm convention. Secretariat of the Stockholm Convention. http://chm.pops.int/ TheConvention/ThePOPs/tabid/673/Default.aspx, 2001. [Accessed February, 2014]. Vladykov VD. Études sur les mammifères aquatiques. IV. Nourriture du marsouin blanc (Delphinapterus leucas) du fleuve Saint-Laurent. Province de Québec: Département des Pêcheries; 1946. p. 17 [119 pp.]. Ward J, Mohapatra SP, Mitchell A. An overview of policies for managing polybrominated diphenyl ethers (PBDEs) in the Great Lakes basin. Environ Int 2008;34:1148–56. Wells RS, Tornero V, Borrell A, Aguilar A, Rowles TK, Rhinehart HL, et al. Integrating lifehistory and reproductive success data to examine potential relationships with organochlorine compounds for bottlenose dolphins (Tursiops truncatus) in Sarasota Bay, Florida. Sci Total Environ 2005;349:106–19.