Declining Trends of Polychlorinated Naphthalenes in Seabird Eggs

Mar 23, 2017 - Declining Trends of Polychlorinated Naphthalenes in Seabird Eggs from the Canadian Arctic, 1975−2014. Birgit M. Braune*,† and Derek...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/est

Declining Trends of Polychlorinated Naphthalenes in Seabird Eggs from the Canadian Arctic, 1975−2014 Birgit M. Braune*,† and Derek C. G. Muir‡ †

National Wildlife Research Centre, Environment and Climate Change Canada, Carleton University, Raven Road, Ottawa, Ontario, Canada K1A 0H3 ‡ Aquatic Contaminants Research Division, Environment and Climate Change Canada, Burlington, Ontario, Canada L7R 4A6 S Supporting Information *

ABSTRACT: There are relatively few studies of polychlorinated naphthalenes (PCNs) for biota in polar regions and even fewer reports of temporal trends. We determined concentrations of PCNs in eggs of thick-billed murres (Uria lomvia) collected from the Canadian high Arctic between 1975 and 2014 and calculated their associated toxic equivalents (TEQs). Concentrations of Σ67PCN decreased significantly in the murre eggs between 1975 and 2014 at an average annual rate of −14.9 pg g−1 wet weight. Although the penta- and tetra-CNs (predominantly CN-52/60 and CN-42) dominated the PCN profile, the hexa-CNs (mainly CN-66/67) accounted for the majority of the Σ67TEQ-PCN, concentrations of which also decreased significantly between 1975 to 2014. On average, Σ67TEQPCN in the murre eggs accounted for only 1.9% of the total toxicity calculated for dioxin-like compounds measured in the murre eggs. As such, the TEQ-PCN concentrations calculated for the murre eggs in this study are several orders of magnitude lower than TEQ levels associated with reproductive effects in birds. This is the first published study of temporal trends of PCNs in Canadian Arctic biota.



INTRODUCTION Polychlorinated naphthalenes (PCNs) are a group of industrial chemicals and byproducts consisting of a possible 75 congeners with physical and chemical properties similar to those of the polychlorinated dibenzo-p-dioxins (PCDDs), dibenzofurans (PCDFs), and planar polychlorinated biphenyls (PCBs). Their thermal stability and chemical inertness make them suitable for use in the electrical industry as flame retardants and dielectric fluids for transformers and capacitors, as well as being used as wood preservatives, carriers in dye production, and machine oil and rubber product additives.1 Although the manufacture of PCNs has been discontinued,2−4 emissions to the environment continue from combustion sources and through volatilization from old products containing PCNs and PCB mixtures, the latter containing PCNs as impurities.1 The physical−chemical characteristics of PCNs enable their persistence, bioaccumulation, and biomagnification in wildlife,1,2,5−7 and some CN congeners also elicit toxic effects, mediated through activation of the aryl hydrocarbon receptor (AhR), which are similar to those of the dioxin-like compounds.6,8−10 Therefore, these compounds continue to pose an environmental problem and have recently been added to the list of compounds regulated under the Stockholm Convention on Persistent Organic Pollutants (http://chm. pops.int/TheConvention/ThePOPs/TheNewPOPs/tabid/ 2511/Default.aspx), as well as the United Nations Economic Commission for Europe (UNECE) Convention on Long© 2017 American Chemical Society

Range Transboundary Air Pollution (CLRTAP) (http://www. unece.org/env/lrtap/pops_h1.html). PCNs are ubiquitous in the environment and are found in a variety of biota5−7,9,11,12 including those found in polar regions.2,13−19 Of particular interest is the fact that PCNs appear to accumulate to substantially higher levels in seabirds than in marine mammals.2,13 The widespread presence of PCNs in Arctic biota and air provides evidence of long-range atmospheric transport from source regions.2,20 However, studies of PCNs in polar regions are still relatively few,2 and studies of temporal trends for PCNs in biota are also scarce. Temporal trends, all decreasing, have been reported for total PCNs in lake trout (Salvelinus namaycush) from Lake Ontario, Canada,11 harbor seals (Phoca vitulina) from the Salish Sea,21 guillemot (Uria aalge) eggs from the Baltic Sea,22 liver of arctic cod (Cadus callarias) from Vestertana Fjord in northern Norway,23 and ringed seals (Phoca hispida) from the Canadian Arctic.24 However, no statistically significant trends were observed for a variety of marine mammal species from the Arctic and sub-Arctic.15 To our knowledge, the latter three studies by Sinkkonen and Paasivirta,23 Houde et al.,24 and Received: Revised: Accepted: Published: 3802

January 23, 2017 March 10, 2017 March 17, 2017 March 23, 2017 DOI: 10.1021/acs.est.7b00431 Environ. Sci. Technol. 2017, 51, 3802−3808

Article

Environmental Science & Technology Rotander et al.15 report the only temporal trend data available for Arctic biota. PCNs were first detected in seabird eggs from the Canadian Arctic by Muir et al.25 Subsequently, monitoring and retrospective analyses of PCNs in eggs of a Canadian Arctic seabird species were carried out. The objective of this study was to evaluate temporal trends of PCNs and their toxic equivalents (TEQs) in eggs of thick-billed murres (Uria lomvia) breeding in the Canadian high Arctic. This is the first published study of temporal trends of PCNs in Canadian Arctic biota.

PCN calibration mix were well within the acceptable limits. Although SRM-1946 is not certified for PCNs, analytical precision was checked by analyzing SRM-1946 with each batch of samples analyzed. Relative standard deviations for replicate readings (n = 4) averaged 29% (range 3−97%; see Table S1). Residue concentrations were blank corrected, as needed, in addition to being recovery corrected (Supporting Information). Stable Nitrogen Analysis. As an estimate of relative trophic position, stable nitrogen isotope assays (15N/14N, expressed as δ15N) were performed on 1-mg subsamples of freeze-dried, lipid-extracted egg homogenates according to procedures described in Braune et al.27 Data Treatment. Dioxin toxic equivalents (TEQs) resulting from PCNs were calculated using toxic equivalency factors (TEFs) based on in vitro bioassays or predicted values for relative potencies expressed relative to 2,3,7,8-TCDD reported by Puzyn et al.10 For CN congeners which coeluted and had different TEF values, the average of the relevant TEF values was used. The resulting TEQs were then summed (Σ67TEQ) to provide an indication of the PCN toxic potential of the sample expressed as 2,3,7,8-TCDD toxic equivalents. As interpretation of contaminant concentrations in biota may be confounded if populations vary their diet over trophic levels through time,28 δ15N, which reflects the diet of the female prior to or during egg laying,29,30 was included in the evaluation of the temporal trends as an index of trophic position29,31 in the murres. Only those congeners for which >90% of the samples had detectable concentrations after correction for blanks were statistically analyzed. Nondetect values were set to one-half the detection limit for purposes of statistical analyses but were set to zero for calculation of Σ67PCN and reported annual means (Table S2). Regression analyses indicated no significant relationship between % lipid and wet weight concentrations of Σ67PCN, Σ67TEQ-PCN, and the 40 CN congeners which could be statistically analyzed. Nor was a significant relationship found between δ15N and wet weight concentrations of Σ67PCN, Σ67TEQ-PCN, and all but four of the 40 CN congeners which could be statistically analyzed. However, analysis of variance (ANOVA) indicated significant interannual variation in both % lipid (F10,40 = 4.10, p = 0.0006) and δ15N (F10,40 = 4.30, p = 0.0004). Therefore, temporal trends for concentrations of Σ67PCN, Σ67TEQ-PCN, and the major CN congeners were analyzed using backward stepwise regression analysis with year, % lipid, and δ15N as regressors. Residuals from the regression analyses were tested for normality using the Shapiro−Wilks’ W test, and for the ANOVA, tests for normality and sample variance in the data were performed using the Shapiro−Wilks’ W and Levene’s tests, respectively. The concentration data were loge transformed, as necessary. Changes in percent contributions of CN congeners and homologues to Σ67PCN and to Σ67TEQ-PCN over time were analyzed using the nonparametric Spearman rank correlation. All statistical tests were performed using Statistica for Windows Version 7.0 (StatSoft Inc., Tulsa, OK) with a significance level of p < 0.05. As the PCN calibration mix contained only 23 native PCN congeners, the ΣPCN concentrations based on those congeners alone were compared with Σ67PCN and found to average 67 ± 4.4% (range 50−76%) of Σ67PCN. Further, the regression analysis with year, % lipid, and δ15N as regressors returned a similar statistical result. Therefore, the results presented are based on the full suite of 67 congeners. The tabulated data are presented as arithmetic means in concentration units of pg g−1



MATERIALS AND METHODS Sampling. Eggs of thick-billed murres were collected from the Prince Leopold Island Migratory Bird Sanctuary (74°02′ N, 90°05′ W) in Lancaster Sound, Nunavut, Canada (Figure S1) in 1975 (n = 9), 1987 (n = 9), 1993 (n = 15), 1998 (n = 15), 2003 (n = 15), 2005 (n = 15), 2006 (n = 15), 2008 (n = 15), 2010 (n = 15), 2012 (n = 15), and 2014 (n = 15). As this species lays a single egg, independence among samples was ensured. All eggs were taken under appropriate research and collection permits. Eggs were kept cool in the field and shipped to the National Wildlife Research Centre (NWRC) in Ottawa, Canada, where egg contents were homogenized and stored frozen at −40 °C in acetone−hexane rinsed glass vials. Samples collected in 2014 were analyzed within six months of collection, whereas earlier samples were retrieved from the National Wildlife Specimen Bank at NWRC for retrospective analysis. Egg homogenates were analyzed for PCNs as well as ratios of stable nitrogen isotopes (15N/14N, expressed as δ15N). Chemical Analysis. Egg homogenates were analyzed for PCNs as pooled (composite) samples with each pool consisting of equal aliquots of three individual egg samples (i.e., a collection of 15 eggs was analyzed as five pooled samples comprising three eggs each; a collection of nine eggs was analyzed as three pooled samples comprising three eggs each). Extraction of egg homogenates followed US EPA Method 1668C26 with minor modifications. Briefly, samples were ground with anhydrous sodium sulfate followed by Soxhlet extraction using dichloromethane. Extracts were cleaned on acid/silica gel (15 cm of 45% w/w H2SO4 on silica gel and 2 cm of neutral silica gel) and then on basic alumina columns. Samples were then concentrated to 25 μL final extract volume in nonane. All samples were spiked with 13C12-labeled PCN internal standards (listed in the Supporting Information) prior to extraction and with injection internal standards (13C-PCBs) in the final vialed extract prior to instrumental analysis. Recoveries of the 13C-internal standards averaged 65% ± 13% (range 34−95%). Samples were analyzed for PCNs by isotope dilution gas chromatography-high resolution mass spectrometry in electron ionization mode (GC-EI-HRMS) at ≥10,000 resolution as per US EPA method 1668C for PCB congeners26 including the use of two chlorine isotope masses and the ion abundance ratio criterion of within 15% of theoretical value. A total of 67 congeners (individuals and coeluting groups) were quantified relative to a suite of 23 available native PCN congeners representing all homologue groups. The results are therefore inherently recovery corrected. Further method details are provided in the Supporting Information. Method blanks, NIST SRM-1946 (Lake Superior lake trout) as well as a laboratory control sample (clean matrix spiked with extraction and native standards) were run with each batch of samples for quality control. Recovery of all congeners in the 3803

DOI: 10.1021/acs.est.7b00431 Environ. Sci. Technol. 2017, 51, 3802−3808

Article

Environmental Science & Technology

Σ67PCN concentrations, adjusted for δ15N, averaged −18.2 pg g−1 ww.

wet weight (ww) for PCNs and TEQs and ‰ for δ15N (Table S2).



RESULTS PCNs were found in measurable concentrations in all sample pools of thick-billed murre eggs collected between 1975 and 2014. CN-16, CN-49, CN-51, CN-53/55, and CN-56 were not detected in any of the samples. Mean annual concentrations of Σ67PCN ranged from 364 ± 22.2 to 995 ± 52.8 pg g−1 ww (Table S2). The predominant CN congeners were CN-42 and CN-52/ 60, which together comprised 42−57% of Σ67PCN annually, followed by CN-57, CN-58, and CN-66/67, which added an additional 20−24% to Σ67PCN. From 1975 to 2014, the pentaCNs (mainly CN-52/60) accounted for 41−50% of Σ67PCN followed closely by the tetra-CNs (primarily CN-42) contributing 36−45% to Σ67PCN (Figure 1A). The hexa-CNs

Figure 2. Mean annual concentrations (pg g−1 wet wt) of total polychlorinated naphthalenes (Σ67PCN) in eggs of thick-billed murres from Prince Leopold Island, Nunavut, Canada, 1975−2014. For the years 1975 and 1987, n = 9 eggs per year analyzed as three pools comprising three eggs each; for the years 1993−2014, n = 15 eggs per year analyzed as five pools comprising three eggs each.

The calculated total PCN toxic equivalent concentrations (Σ67TEQ-PCN) also showed a significant decrease (p < 0.001) in the murre eggs, from an annual mean of 0.89 ± 0.05 pg g−1 in 1975 to 0.20 ± 0.005 pg g−1 ww in 2014 (Figure 3). There

Figure 1. Percent contributions of (A) mono- to octa-chlorinated naphthalenes to total polychlorinated naphthalenes (Σ67PCN) and (B) toxic equivalents (TEQs) of mono- to octa-chlorinated naphthalenes to total TEQs (Σ67TEQ) in eggs of thick-billed murres from Prince Leopold Island, Nunavut, Canada, 1975−2014.

Figure 3. Mean annual concentrations (pg g−1 wet wt) of total toxic equivalents (Σ67TEQ) for polychlorinated naphthalenes in eggs of thick-billed murres from Prince Leopold Island, Nunavut, Canada, 1975−2014. For the years 1975 and 1987, TEQ calculations were based on n = 9 eggs per year analyzed as three pools comprising three eggs each; for the years 1993−2014, TEQ calculations were based on n = 15 eggs per year analyzed as five pools comprising three eggs each.

(primarily CN-66/67) accounted for 5−10%, and the tri-CNs (mainly CN-15) accounted for 2−7% of Σ67PCN. Total heptaand octa-CNs each contributed trivial amounts (