Science of the Total Environment 533 (2015) 290–296

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Polyhalogenated compounds (PCBs, chlordanes, HCB and BFRs) in four polar bears (Ursus maritimus) that swam malnourished from East Greenland to Iceland

Walter Vetter a,⁎, Vanessa Gall a, Karl Skírnisson b a University of Hohenheim, Institute of Food Chemistry, Garbenstr. 28, D-70593 Stuttgart, Germany b University of Iceland, Keldur, Institute for Experimental Pathology, IS-112 Reykjavík, Iceland

HIGHLIGHTS GRAPHICAL ABSTRACT

• Four polar bears swam in malnourished condition from East-Greenland to Iceland. • Organohalogen compounds (PCBs, chlor- dane, PBBs, PBDEs, HCB) were quantified. • Dec-602 was tentatively identified in all tissues and an unknown compound in liver.

article info abstract

Article history: Levels of organohalogen compounds (PCBs, chlordane, PBB 153, PBDEs, HCB) were determined in adipose tissue, Received 24 March 2015 liver, kidney and muscle of four polar bears which swam and/or drifted to Iceland in extremely malnourished Received in revised form 2 July 2015 condition. Since the colonization in the 9th century polar bears have been repeatedly observed in Iceland. Accepted 3 July 2015 However, in recent years three of the animals have clearly left their natural habitat in poor condition in May or Available online xxxx June, i.e. at the end of the major feeding season. The fourth bear is believed to have drifted with melting ice to Editor: D. Barcelo North-Eastern Iceland in mid-winter. The concentrations of the POPs were within the range or higher than the typical concentrations measured in polar bears from the East Greenland population. In addition to the targeted

Keywords: compounds, we tentatively detected Dechlorane 602 and its potential hydrodechlorinated Cl11-metabolite in Polyhalogenated compounds all samples. Moreover, a polychlorinated compound which partly co-eluted with PCB 209 was detected in all POPs liver samples but not in adipose tissue, kidney or muscle. The mass spectrum of the potential metabolite did Dechlorane-602 not allow determining its structure. Polar bears are good swimmers and can reach Iceland from the ice edge of Unknown liver peak East Greenland within a few days. Potential reasons for the swims are briefly discussed. Polar bears © 2015 Elsevier B.V. All rights reserved. Malnorishment

1. Introduction

Different persistent organic pollutants (POPs) have been detected in ⁎ Corresponding author at: University of Hohenheim, Institute of Food Chemistry fl (170b) Garbenstraße 28, 70593 Stuttgart, Germany. Arctic ora and fauna due to atmospheric long range transport from in- E-mail address: [email protected] (W. Vetter). dustrialized and agricultural areas (Jones and de Voogt, 1999). As the

http://dx.doi.org/10.1016/j.scitotenv.2015.07.011 0048-9697/© 2015 Elsevier B.V. All rights reserved. W. Vetter et al. / Science of the Total Environment 533 (2015) 290–296 291 top predator of the Arctic marine food web, polar bears (Ursus 2.2. Size and condition maritimus), which feed predominantly on ringed seals (Verreault et al., 2005), have been highly exposed to such pollutants as a result of All individuals were malnourished on their arrival in Iceland. The their bioaccumulation in the food chain. The natural habitats of polar lipid content of the samples was determined by isolation of the visible bears are found throughout the ice-covered seas of the circumpolar body fat reservoir and weighing it with an accuracy of one kilogram. Arctic (Stirling and Derocher, 2012). Polar bears are not native in PB-1 (male, 22.5 years old) weighed 220 kg and had a body length of Iceland, which is located south of the Arctic Circle in the North Atlantic 209 cm. The lipid content was approximately 10% of the body weight. Ocean, between North America and Europe. Sightings of polar bears in This individual arrived in Iceland on June 3, 2008 (Fig. S1, Supporting Iceland have been documented since the 9th century: at least 500 Information), i.e. at the end of the mating season and after the annual polar bears have reached the island, including 50–60 in the 20th century seal-pup hunting period on the ice north of Iceland in late winter and (Petersen and Haraldsson, 1993). Usually these polar bears traveled before the seasonal fasting from summer to fall (Polischuk et al., with pack ice drifts towards Northwest, North and Northeast Iceland. 2002). PB-2, a 14.5 year old female (142 kg body weight, 194 cm body Some were killed on arrival, but frequently the animals disappeared length) also arrived after the mating season on June 16, 2008 (Fig. S1, when they returned back to the drift ice. Some of these accounts refer Supporting Information). The annuli pattern in the cementum layer to cadavers discovered on the shore or even remains that were found suggested that this old female had given birth to cubs on three occa- in fishing nets off the coast. sions, which had most likely been successfully raised to weaning; the Between 2008 and 2011, four polar bears left their natural distribu- cubs were born when the female was six, nine and 12 years old tion range and swam to Iceland. These individuals which most likely (Skírnisson, 2009). Furthermore, a broad dark-staining incremental originated from the East Greenland population were shot when they ar- line when the female was five year old suggested that it had spent the rived in Iceland. At the time of their arrival, the East Greenland ice edge winter in a maternity den and given birth to cubs that were not raised was only 110 to 170 km away from Iceland (Skírnisson, unpublished (Skírnisson, 2009). Despite arriving shortly after the seasonal seal- data). Polar bears are good swimmers, and are able to travel this hunting period, the female had no fat reservoir left in the body, so that distance in a few days. For instance, a satellite-tracked female polar adipose tissue could not be analyzed. Immediately after its arrival the bear was found to have swum 687 km in nine days at an average exhausted old female laid down in a colony of eider ducks close to the speed of 3 km/h (Durner et al., 2011). At such a speed, the polar bears sea shore and did not attempt to move, hunt or eat. Being close to in this study would have made the journey to Iceland in 1.5–2.4 days. death, it was shot 30 h after its arrival in Iceland. Autopsy revealed Dissections revealed the polar bears to be in very poor condition with water in the lungs (evidence of heart failure). Her condition at the unusually little, if any, lipid reservoirs left. At least two of these onset of the swim (to Iceland) remains unknown but prolonged starva- malnourished animals are believed to have left their natural habitat in tion before leaving its habitat and/or during the long stay in water prior East Greenland already in poor body condition. Depending on the time to the arrival in Iceland is regarded to have depleted all fat reserves. spent during the swim to Iceland some lipid depletion would have PB-3 was a subadult female (4.5 years old, 138 kg body weight, occurred in transit as polar bears usually do not hunt and feed in the 173 cm body length) shot in North-Eastern Iceland on January 27, open sea (Durner et al., 2011). 2010 (Fig. S1, Supporting Information). Before arriving in Iceland it is In this study we analyzed persistent organic pollutants (POPs) in believed this young female had been located close to the ice edge adipose, liver, kidney and muscle tissue of the four polar bears. Since north of Iceland. The animal is thought to have drifted, perhaps on an they had not consumed any food in Iceland we compared the POP pat- iceberg into the Gulf Stream where the ice melted whereon prevailing tern in the polar bears in order to determine whether the concentra- wind and water currents from the northeast would have aided its swim- tions and compound ratios were characteristic for polar bears already ming and drifting journey to northwest Iceland (Skírnisson and examined from the neighboring East-Greenland population. We also Sigurðardóttir, 2010). The body fat reservoir of this animal was ~6% of screened the samples for new contaminants and considered some the body weight, probably not unusual for this time of year, being a potential reasons that might have influenced the decision of the polar few months prior to the seasonal feeding on the ice. PB-4 was a sub- bears to swim away from their natural range. adult, 3.5 year old female (95 kg body weight, 193 cm body length) which was detected and shot in North-West Iceland on May 2, 2011 (Fig. S1, Supporting Information). This animal's body fat reservoir was 2. Materials and methods ~5% of its body weight. Samples of these animals were compared with those of a 6.5 year old 2.1. Arrival of polar bears in Iceland and sampling young male (260 kg body weight, 205 cm body length), killed in its natural habitat at the ice-edge north of Iceland in June 1993 and later Four polar bears (PB-1 to PB-4) were shot soon after their swim to dissected at the Institute for Experimental Pathology in Iceland. Lipid Iceland in 2008, 2010, and 2011. The polar bears were dissected and reservoir contributed ~25% to the body weight (Klobes et al., 1998). processed by specialists from the Institute for Experimental Pathology, This apparently healthy individual did not show the same degree of Reykjavík, Iceland. Samples for measurements of polyhalogenated com- malnourishment and was considered to be typical for healthy polar pounds were taken from the liver, skeletal muscles, kidney and subcuta- bears at the ice edge north of Iceland. It was the most recent polar neous and mesenteric fat (if present), wrapped in aluminium foil and bear studied in Iceland prior to the individuals analyzed in this study. kept frozen at −18 °C until analysis. Age determination was performed by counting growth layer groups in the cementum region of I1 2.3. Sample cleanup procedure (Skírnisson, 2009). Upon arrival of the polar bears in Iceland the distance to the ice edge east of Greenland varied between 110 and Samples were lyophilized in a Lyovac GT2 (Leybold-Heraeus, 170 km. PB-1, PB-3 and PB-4 most likely swam within a few days to Cologne, Germany) apparatus for at least 24 h. Aliquots were supple- Iceland (Skírnisson, unpublished data). By contrast, histological exami- mented with the internal standard perdeuterated α-HCH (α-PDHCH). nation of PB-2 produced evidence that this individual had been in the It followed focused open-vessel microwave-assisted extraction (FOV- sea for a considerably longer period of time, even weeks. Specifically, MAE) in a modified Star 2 system (CEM, Kamp-Lintfort, Germany) in contrast to the other polar bears PB-2 showed long-term, hairless (Vetter et al., 1998). Samples were loaded into the extraction tubes friction wounds under the forelegs. Marked thickening of the skin (300 mL), 90 mL cyclohexane/ethyl acetate (46:54, w/w) and the suggested that the healing process had been underway for some time internal standard α-PDHCH (adjusted to give a final concentration of (Sigurðardóttir, unpublished). ~20pg/μL) were added and the extraction was placed in the microwave 292 W. Vetter et al. / Science of the Total Environment 533 (2015) 290–296 system and equipped with a water trap and a refluxer according to more from PCB 163 (Klobes et al., 1998). Therefore, a detailed evalua- Batista et al. (2001). The system was programmed as follows: step 1: tion of PCB congener ratios could not be carried out. In the following, within 10 min to 88 °C, hold time 10 min; step 2: within 5 min to only the predominant of co-eluting PCB congeners is listed and 95 °C, hold time 30 min. The extracts were filtrated into 100 mL discussed. amber glass flasks and concentrated by rotary evaporation (30 °C, The PCB pattern was dominated by PCB 153, which contributed 29% or 170 mbar) to b5 mL. The sample extract was transferred into 5 mL more to the PCB content. With a few exceptions, the abundance of PCBs calibrated flasks and made up to 5 mL. The bulk (4.5 mL) of the decreased in the order PCB 153 N PCB 180 N PCB 170 N PCB 138 N PCB sample solution was subjected to gel-permeation chromatography 99 (Fig. 1). Remarkable exceptions were in the muscle of PB-2 where with 5 mL/min of cyclohexane/ethyl acetate (46:54, w/w). Then, the PCB 180 contributed only b1% to sumPCBs. Moreover, all tissues of PB-4 solvent was changed to iso-octane. The sample was further purified by showed a higher prevalence of PCB 138 in the PCB pattern than the sam- adsorption chromatography (1 cm i.d. column) with 3 g silica gel ples from the other three polar bears (~30% vs. 8–15% contribution to deactivated with 30% water (w/w) and topped with Na2SO4 (Bendig sumPCBs, Fig. 1). In the liver of PB-4, PCB 138 was even more abundant et al., 2012). The sample was eluted with 60 mL n-hexane. Then, the vol- than PCB 153. PCB 194 (concentration ~50% of PCB 170), PCB 209 and fur- ume was reduced and made up to exactly 5 mL. Aliquots were analyzed ther congeners were also detected but not quantified in this study. by GC/ECD and GC/ECNI-MS. Organohalogen reference standards were Norstrom et al. (1988) noted that six PCB congeners (i.e. PCB 99, PCB from LGC Promochem, Dr. Ehrenstorfer, Wellington, or synthesized in 153, PCB 138, PCB 180, PCB 170 and PCB 194) represented ~93% of the our group. PCB content. These PCBs were also the most abundant congeners in our samples. Polar bears are able to metabolize many of the PCBs (Norstrom 2.4. Gas chromatography with electron capture detection (GC/ECD) et al., 1988; Bytingsvik et al., 2012) which are persistent in other mam- mals (Norstrom and Muir, 1994). GC/ECD measurements were performed with a Hewlett-Packard Based on lipid weight, the concentrations of PCBs were highest in liver (HP) 5890 series II plus system equipped with a GC-PAL auto sampler (Table 1). Except for PB-2, which did not have any body lipids left (no (CTC Analytics, Zwingen, Switzerland). Helium (5.0) was used as the analysis possible), adipose tissue was second in prevalence (amounting carrier gas. One microliter was splitless injected (injector temperature to 37–54% of the concentration in liver lipids) followed by muscle or kid- 280 °C) onto a 30 m length × 0.25 mm i.d., 0.25 μm film thickness ney. Higher lipid-based PCB concentrations in liver than in adipose tissue BGB-1 capillary column (BGB-Analytik, Boeckten, Switzerland). The were also observed in the Icelandic polar bear (PB-1993) from 1993 GC oven program started at 80 °C (2 min), then at 10 °C/min to 300 °C (~50% higher, Klobes et al., 1998) and especially in samples from East- (hold time: 20 min). The detector temperature was set to 300 °C and Greenland analyzed in 1999–2001 (4.7 fold higher, Gebbink et al., nitrogen 5.0 was used as the makeup gas. 2008a) when these literature results were transferred from fresh weight to lipid weight (Table 1). 2.5. Gas chromatography with electron capture negative ion mass The PCB concentration of 21,000 ng/g lipids in adipose tissue of the spectrometry (GC/ECNI–MS) only adult male individual in our study (PB-1) was about two fold higher than the average of the samples from the East Greenland population A 7890A GC, equipped with a 7693A auto sampler and a 5975C mass (2006 and 2010) analyzed by Dietz et al. (2013a) (Table 1). Of the 28 spectrometer (Agilent, Waldbronn, Germany) was used for the mea- adult males studied by Dietz et al. (2013a) in 2006–2010, only one indi- surements. The carrier gas helium (5.0) was transported with a flow vidual had similarly high sum-PCB concentrations as PB-1. The PCB con- rate of 1.2 mL/min. Injections (1 μL) were made in splitless mode with centration in PB-1 was also ~2.5 higher than in the male polar bear a pressure pulse of 25 psi/min. Methane (5.5) was used as the modera- from 1993 (Klobes et al., 1998)(Table 1). Since the new millennium, tion gas at a flow rate of 2.0 mL/min. The transfer line, ion source and only one polar bear from East-Greenland showed a higher PCB level of quadrupole temperatures were set at 300 °C, 150 °C and 150 °C. A DB- ~25,000 ng/g lipid weight (Dietz et al., 2013a). 5MS column (Agilent, Folsom/USA) of 30 m length, 0.25 mm i.d., With regard to the PCB concentrations in our subadult females, PB-3 0.25 μm film thickness was installed in the GC oven. The solvent delay showed a concentration within the 95% confidence interval of juvenile was 8 min. In the full scan mode (m/z 50–800), the GC oven program females reported by Dietz et al. (2013a) for 2006–2010, whereas those was the following: 50 °C (1 min), then at 10 °C/min to 300 °C of PB-4 were ~four fold higher than mean values reported in the litera- (14 min). In the selected ion monitoring (SIM) mode, the GC oven ture (Table 1). Based on a lipid content, the PCB concentration in PB-4 program started at 60 °C. After 2 min, the temperature was raised at was twice as high as the highest PCB concentration reported by Dietz 10 °C/min to 300 °C (14 min). Six SIM methods were used for quantifi- et al. (2013a) in their comprehensive study on PCBs in polar bears cation and screening purposes (see Supporting Information). Sample, from East Greenland from 2000 on. Throughout the sampling period chemical, and instrumental blanks were analyzed and measured regu- larly. None of the compounds reported in this study were detected in these blanks.

3. Results and discussion

GC/ECD and GC/ECNI–MS measurements enabled the detection of ~50 compounds per sample. The POP patterns in the samples were similar, and the most abundant peaks originated from PCBs or oxychlordane. Throughout this article, concentrations of polyhalogenated compounds are reported on a lipid weight basis.

3.1. PCB residues in the polar bears

Up to 19 PCB congeners were detected by GC/ECNI–MS in the sam- ples. However, several of these PCBs co-elute with other PCB congeners on the capillary columns used in our study. For instance, in polar bears the peak at the retention time of PCB 138 may originate by 25% or Fig. 1. Percentage distribution of major PCB congeners in tissues of the polar bears. W. Vetter et al. / Science of the Total Environment 533 (2015) 290–296 293

Table 1 PCB levels (ng/g lipid weight) in tissues of polar bears stranded in Iceland (2008–2011) and literature data.

PB-1 PB-2 PB-3 PB-4 Mean Sample 1993 Gebbink Dietz et al., 2013b, 22.5 years, male 14.5 years, female 4.5 years, female 3.5 years, female value (Klobes et al., 1998) et al., 2008a 2006–10 ⁎ ⁎⁎ Adipose tissue 21,000 – 7,890 36,100 21,700 7,860 4,850 7,630–12,650 ⁎ ⁎ Liver 56,400 20,500 14,500 74,400 41,500 15,130 28,400 Kidney 6,150 10,300 4,590 25,800 11,700 –– Muscle 15,200 8,400 5,500 25,000 13,500 ––

⁎ Data transferred from fresh weight to lipid weight by using the average lipid content of 90% in adipose tissue and 11% in liver as reported by Gebbink et al. (2008a). ⁎⁎ All age classes: adult males: 12,650 ng/g lipid weight PCBs (95% confidence interval: 11,030–14,270 ng/g lipid weight); adult females: 7,630 ng/g lipid weight PCBs (95% confidence interval: 5,740–9,510 ng/g lipid weight); juveniles: 8,470 ng/g lipid weight PCBs (95% confidence interval: 7,020–9,930 ng/g lipids) (Gebbink et al., 2008a).

1983–2010, only two subadults from the mid-1980s had higher PCB of the ion abundance ratio of m/z 79 to m/z 161 revealed that the first concentrations (Dietz et al., 2013a). Dietz et al. (2013a) noted a decline peak in our sample mainly originated from PBB 153 while BDE 154 of PCB concentrations in subadult polar bears from 1983–2010 by ~4% only played a minor role. This is in agreement with the fact that BDE per year. Moreover, PB-2 and PB-4 exceeded the highest sumPCB 154 was not among the PBDE congeners reported in polar bears from concentration detected in 2006 in adipose tissue of 140 polar bears East Greenland (Dietz et al., 2013b). GC/ECNI–MS analysis of the most from all over the Arctic of 19,700 ng/g (Muir et al., 2006). It is worth abundant isotope peak of the molecular ion of PBB 153 (i.e. m/z 627.5) noting that average PCB levels in polar bears were highest in the East and BDE 154 (i.e. m/z 643.5) in two samples verified that the peak Greenland population, followed by Svalbard in the east (European in the chromatograms originated from PBB 153 (Fig. S2, Supporting Arctic) and different regions such as Baffin Island in the Western Arctic Information). Moreover, the measurement of m/z 627.5 ([M + 6]− of (Muir et al., 2006). hexaBBs) enabled us to detect three further peaks originating from The high PCB concentrations, especially in two of our samples are like- hexabromobiphenyl isomers. Based on the work of von der Recke ly due to the mobilization of body lipids during the fast. Polischuk et al. and Vetter (2008a), these peaks represented the co-eluting pair PBB (2002) studied the effect of fasting for 47–68 days when the body fat of 132/PBB 146, as well as PBB 133 and PBB 149 (Fig. S2, Supporting polar bears decreased from ~28% to ~23%, i.e. less threatening than in Information). The hexabromobiphenyl pattern detected in the polar the present samples where the body lipids contributed 0–10% to the bears is a strong indicator for the exposure being to technical body weight. Polischuk et al. (2002) further noted that PCB concentra- hexabromobiphenyl (which was primarily applied in the U.S.) and not tions in adipose tissue were increasing during the fasting period, depen- with decabromobiphenyl (which was primarily applied in Europe, von dent of gender and age, as well as the constitution of the individual der Recke and Vetter, 2008b). Accordingly, the PBB contamination while the total PCB load in the body remained unchanged (in contrast mainly originated from North America (because transformation of to chlordane in males and DDT, HCH in all bears). Accordingly, PCBs decabromobiphenyl leads to the accumulation of PBB 154 and were neither excreted nor metabolized during this period (Polischuk PBB 155 which were not detected in the polar bear samples), and et al., 2002). In agreement with this study we did not notice differences this verified that the polar bears did not originate from the in the PCB patterns in most samples except for the liver of PB-4 where more distant Svalbard population (distance between Svalbard and PCB 138 was more abundant than PCB 153 and the extremely low contri- Iceland N1000 km). However, contributions of the other PBB congeners bution of PCB 180 in the muscle of PB-2 (Fig. 1). In the latter case, it seems were low and only PBB 153 was quantified in this study. In addition, no probable that PCB 180 had been metabolized during starvation because response was detected for PBB 138 (which is in agreement with marine PB-2 was extremely malnourished and did not have any body lipids left. mammals) (von der Recke and Vetter, 2008a). Thus, the second peak in In the remaining samples, the contribution of the body lipids to the total question which was detected in some chromatograms could be traced weight was 5–10%. This is 15–20% less than the normal case of 25% back to BDE 153, although m/z 161 could not be detected in all samples body lipids (based on the apparently healthy individual killed north of due to the low abundance of the compound. Iceland in 1993 (Klobes et al., 1998) and in agreement with Polischuk In agreement with other studies, BDE 47 (65–82% of the PBDE et al., 2002). Accordingly, about 20–40 kg fat had been mobilized from contamination in polar bears from East Greenland and Svalbard, Muir the adipose tissue reservoirs during the fast. A similar loss of weight et al., 2006) was the predominant PBDE congener in polar bears follow- was reported for a female between inspections in August 23 and October ed by BDE 99 or BDE 153 (Muir et al., 2006; Gebbink et al., 2008a). As 2008 (Durner et al., 2011). with PCBs, the highest PBDE concentration was found in the subadult With an average PCB load of 21.7 mg/kg lipids (Table 1), the amount female PB-4 (concentration of BDE 47: 250–1030 ng/g lipid weight). released and otherwise distributed in the polar bear's body was The PBDE concentrations in polar bears from East Greenland were ~430–870 mg PCBs without adjusting for an increase in concentration higher than (average 70 ng/g lipid weight PBDEs, Muir et al., 2006)or due to starvation. Assuming the PCB concentration in remaining lipids on the same level (average 43 ng/g lipid weight PBDEs, McKinney increased by 40% (typical values ~30–50% for females and ~20% for et al., 2011) as those in Svalbard (average 50 ng/g lipid weight PBDEs, males during a 3–5 month fast from summer to fall, Polischuk et al., Muir et al., 2006, 43 ng/g lipid weight PBDEs, McKinney et al., 2011). 2002), the amount released from adipose tissue would still be ~250– Lower PBDE concentrations were measured in the east and other popu- 500 mg PCBs. lations in the Western Arctic (average 22 ng/g lipid weight PBDEs, Muir et al., 2006) except the South Hudson Bay (78 ng/g lipid weight, 3.2. Residues of brominated flame retardants in the polar bears McKinney et al., 2011). Between 35–132 ng/g lipid weight PBDEs were determined in adipose tissue samples from East Greenland in PBDEs were determined by GC/ECNI–MS–SIM using m/z 79 for 1999–2001 (Muir et al., 2006). In particular, PB-4 exceeded these quanti fication (formed by virtually all polybrominated compounds) PBDE concentrations in adipose tissue by more than six-fold (Table 2). and m/z 161 (restricted to polybrominated diphenyl ethers and their Samples PB-1 and PB-3 were slightly above the average PBDE levels in derivatives) for verification (Vetter, 2001). All samples showed peaks samples from the East Greenland population. at the retention times of BDE 47, BDE 153 and BDE 154 (and partly In PB-1, PB-2 and PB-3, PBB 153 was higher concentrated than BDE 99 and BDE 100). However, with the present GC setup BDE 154 sumPBDEs (Table 2). Likewise, the highest levels on a lipid basis were co-eluted with PBB 153 (Covaci et al., 2003; McKinney et al., 2011) determined in liver, followed by adipose and muscle tissue. The kidney and BDE 153 with PBB 138 (Vetter and Rosenfelder, 2008). Evaluation samples showed the lowest content of PBB 153 (Table 2). It followed in 294 W. Vetter et al. / Science of the Total Environment 533 (2015) 290–296

Table 2 Concentration of different POPs (ng/g lipid weight) in polar bears stranded in Iceland (2008–2011).

PB-1 PB-2 PB-3 PB-4

Fat/liver/kidney/muscle Fat/liver/kidney/muscle Fat/liver/kidney/muscle Fat/liver/kidney/muscle

PBDEs 94/300/15/68 –/81/35/45 74/90/45/53 830/1,470/520/410 PBB 153 260/590/75/160 –/170/77/86 65/94/42/53 180/160/110/180 Oxychlordane 4,330/66,500/1,230/2,950 –/10,000/370/350 2,690/36,000/1,260/1,530 8,370/95,900/6,000/5,400 trans-chlordane 4.1/8.8/n.d./2.9 –/1.4/n.d./n.d. 2.4/12/n.d./0.8 3.8/6.5/3.0/2.9 cis-Chlordane 4.9/39/n.d./3.9 –/21/n.d./n.d. 3.3/12/n.d./0.8 15/74/11/9.0 HCB 330/400/220/160 –/400/220/160 270/170/130/100 1,840/1,500/1,220/350 ⁎ %cis- of oxychlordane 0.11%/0.13%/0.006%/––/–/0.02%/– 0.12%/0.05%/0.003%/– 0.18%/0.17%/0.08%/0.18%

⁎ Percentage of cis-chlordane of the concentration of oxychlordane. Gebbink et al. (2008) noted that cis-chlordane represented 15% of oxychlordane in adipose tissue of polar bears. abundance BDE 47 (PB-1 and PB-2) which reached ~1/3 of the PBB 153 we found that the peak in both studies did not originate from chlordene level and BDE 153 (~1/10 of PBB 153). PB-4 was different in that the plus. Subsequent GC/ECNI–MS analysis of m/z 608 through m/z 617 in level of PBDEs in adipose tissue exceeded that of PBB 153. the SIM mode revealed that this compound had a Cl12-isotope pattern In samples from East Greenland (2005–2008) average PBDE levels starting at m/z 608 (Fig. 2b,c).Thisisindeed2uhigherthanM− of (43 ng/g lipids) were comparable with the average level of PBB 153 chlordene plus. Starting with the maximum number of carbon atoms, (39 ng/g lipids) in samples from East Greenland whereas in Svalbard, possible molecular compositions of the compound in our samples

PBDEs (44 ng/g lipids) were on average two fold higher than PBB 153 were C15H8Cl12,C14H20Cl12,C14H4Cl12O, C13H16Cl12O, or C12H12Cl12O2. (21 ng/g lipids) (McKinney et al., 2011). McKinney et al. (2011) also The only plausible (known) composition within these possibilities suggested that PBB 153 was the most abundant single BFR compound was found to be C14H4Cl12O which is the molecular formula of in polar bears. This was also found in our individuals except for PB-4. Dechlorane 602. Dechlorane plus and related compounds (Dechlorane This sample showed a much higher PBDE level than any polar bear 602, Dechlorane 603 and Dechlorane 604) were detected in various ma- reported in the literature (Muir et al., 2006; Gebbink et al., 2008a,b). rine and terrestrial samples (Feo et al., 2012; Shen et al., 2010). Frequently, Dechlorane 602 was the most prominent one (e.g. in fish 3.3. Chlordane and HCB residues in the polar bears from the Great Lakes (Shen et al., 2010) and peregrine falcon eggs from Canada (Guerra et al., 2011)) although there is little information Oxychlordane was two orders of magnitude higher concentrated than cis-andtrans-chlordane (Table 2). Typically, cis-chlordane domi- nated over trans-chlordane. According to literature, adipose tissue of male adults (2006–2010) from East Greenland contained on average 840 ng/g lipids oxychlordane and well as 13 and 1.8 ng/g lipids cis- and trans-chlordane (Dietz et al., 2013a). Concentrations in juveniles were on the same level, and those of female adults were slightly higher (Dietz et al., 2013a). Levels of cis- and trans-chlordane in our samples were within the literature range while oxychlordane was up to tenfold higher concentrated (Table 2). Chlordane levels in adipose tissue were reported to increase during normal seasonal fasting, but not more than by a factor of 2 (de Wit et al., 2010). The very high oxychlordane levels in our samples indicated that this trend was continuing when body lipids were further decreasing. This was also apparent from the concentration of cis-chlordane relative to oxychlordane. In our adipose tissue samples (PB-1, PB-3, PB-4), the concentration of cis-chlordane amounted to 0.11–0.18% of oxychlordane which is ~ one order of magnitude less than the value of 1.55% calculated from concentrations reported by Dietz et al. (2013a). In the other tissues, the relative contri- bution of cis-chlordane was even lower. The ratio of cis-chlordane to oxychlordane could thus be useful to estimate the length of fasting of polar bears. The samples also contained several further chlordane- related compounds which were not quantified by us. HCB concentrations were lowest in muscle and highest in adipose tissue or liver (Table 2). In PB-1, HCB ranged from 160–400 ng/g lipids but in PB-4, HCB levels were tenfold higher (Table 2).

3.4. Qualitative detection of further polyhalogenated compounds in the polar bears

According to GC/ECNI–MS measurements, the samples contained further organochlorine compounds. Three remarkable compounds were detected in the higher retention time range, i.e. between octachlorobiphenyls and PCB 209. The most prominent one in the GC/ECNI–MS chromatograms showed the most abundant isotope Fig. 2. GC/ECNI–MS (a) full scan and (b) SIM mass spectrum (ten subsequent masses, m/z 608–617) of a polychlorinated compound tentatively identified as Dechlorane 602 peaks at m/z 612/614 (Fig. 2a). A compound with a similar mass spec- in the liver of a polar bear (PB-2) due to the perfect match with (c) the calculated isotope trum was recently described by Bendig et al. as chlordene plus pattern of Dechlorane 602. Note that ions above m/z 617 were not recorded in SIM mode (Bendig et al., 2013). However, by means of a chlordene plus standard and are thus missing in the mass spectrum shown in (b). W. Vetter et al. / Science of the Total Environment 533 (2015) 290–296 295 on its production and use (Feo et al., 2012). Re-inspection of the GC/MS less than PCB 209). A second fragment ion was detected at m/z 342 data of Bendig et al. (2013) verified that the peak in their samples was and featured a tetrachloro pattern but further details could not be also Dechlorane 602. found (Fig. 3d). Its exclusive presence in liver samples indicated that Slightly earlier than Dechlorane 602, we noted a further small peak the “liver peak” was most likely the metabolite of a polyhalogenated whose GC/ECNI–MS spectrum was similar to Dechlorane 602 except compound. The unresolved structure makes this compound an impor- that the molecular ion was shifted to lower mass and corresponded tant candidate for follow-up studies. Last but not least, it should be with one Cl substituent less. This polychlorinated compound (Fig. S3) noted that halogenated natural products (HNPs) were not detected in was tentatively identified as a hydrodechlorination product of any sample. Since bioaccumulative HNPs have been detected in many

Dechlorane 602 (C14H5Cl11O). regions worldwide (Vetter, 2006), it is rather likely that the polar bear A further unknown polyhalogenated compound, which was detect- is capable of metabolizing these compounds. ed in all liver samples, eluted slightly earlier than PCB 209. Most remark- ably, this “liver peak” was only detected in liver but not in adipose tissue 3.5. Comparison of concentrations of different organohalogen compounds and kidney (Fig. 3a–c). Despite its long retention time, the highest iso- in the polar bear samples from Iceland tope cluster that could be detected was topped by m/z 379 (i.e. N100 u In three out of four liver samples, concentrations of oxychlordane exceeded those of the PCBs (Tables 1,2). Only in PB-2, oxychlordane (10,000 ng/g lipids) was lower concentrated than sumPCBs but still higher than PCB 153 (6,590 ng/g lipids). Dominance of oxychlordane over PCBs in liver was also found in the polar bear sample from 1993 (Klobes et al., 1998). According to Dietz et al. (2013a), the concentra- tions of PCB 153 and oxychlordane have moderately declined in adipose tissue between the 1980s and the 2010s but at a comparable ratio, so that the ratio between both substance classes has remained relatively constant. Chlordane was produced and used in high quan- tities in the USA but not in Europe and Japan (WHO, 1994). The high concentrations of oxychlordane support our hypothesis that the four polar bears most likely originated from the East Greenland popula- tion. By contrast, polar bears from Svalbard are characterized by higher burdens of PCBs compared to oxychlordane (Klobes et al., 1998). HCB, cis-andtrans-chlordane as well as polybrominated compounds were much lower concentrated than oxychlordane and PCBs. According to literature, PBDE levels in polar bears are usually 2–3 orders of magnitude lower than PCBs (McKinney et al., 2011). This was also found in our samples except for PB-4 which showed a comparably high burden of PBDEs (3–4.5% of PCBs). This may be due to a different origin or a particularly strong effect due to the emaciated status of this individual. Within subpopulations, relatively high PBDE concentrations were determined in Southern and Western Hudson Bay N East Greenland and Svalbard N Eastern Canadian Arctic N Western Arctic (McKinney et al., 2011). However, this could not be substantiated by additional data except that the animal reached Iceland on the North-West while the other animals entered Iceland in the North or North-East.

4. Conclusions

Starvation had led to a strong depletion of the lipid reserves of the polar bears. During fasting, when almost all lipid stores are remobilized, polyhalogenated compounds can suddenly become bioavailable and re-circulate in the body (Fuglei et al., 2007). As a consequence, they can be redistributed in sensitive vital organs such as brain, kidneys and liver, i.e. tissues where toxic effects were much more likely to occur and where the risk of negative health effects is high (Fuglei et al., 2007; Daley et al., 2009). In the present cases alone, ~250–500 mg PCBs were likely released from adipose tissue during the fasting. Typically, 96% of the PCBs are stored in adipose tissue (Gebbink et al., 2008b). Remobilization from lipids may explain the relatively high concentrations of PCBs and other contaminants in the tissues of the polar bears (see above). Three of the polar bears left their environment in May or June in a period of the year when they should be well-nourished and in good body condition after successful late-winter seal hunting and consumption. Two of the bears (PB-1 and PB-4) most likely swum Fig. 3. GC/ECNI–MS (a) full scan chromatograms of (a) liver, (b) adipose tissue and to Iceland within a few days. At least these individuals were already (c) kidney of a polar bear (PB-2) indicating the presence of an unknown compound labeled “LP” (partly co-eluting with PCB 209) detected in liver only along with (d) its malnourished when they left their habitat. Starving polar bears in GC/ECNI–MS spectrum. May and June 2008–2011 may indicate that there is a lack of sufficient 296 W. Vetter et al. / Science of the Total Environment 533 (2015) 290–296 food sources in their natural habitat. Stirling and Derocher (2012) Fuglei, E., Bustnes, J.O., Hop, H., Mørk, T., Björnfoth, H., van Bavel, B., 2007. Environmental contaminants in arctic foxes (Alopex lagopus) in Svalbard: Relationships with feeding reported that polar bears are nowadays exposed to shorter hunting ecology and body condition. Environ. Pollut. 146, 128–138. periods due to the climate change. This could be one reason for the Gebbink, W.A., Sonne, C., Dietz, R., Kirkegaard, M., Riget, F.F., Born, E.W., Muir, D.C.G., poor body condition of the polar bears in spring. From 2012 until June Letcher, R.J., 2008a. Tissue-specific congener composition of organohalogen and metabolite contaminants in East Greenland polar bears (Ursus maritimus). Environ. 2015, polar bears have not been spotted in Iceland. It can, however, Pollut. 152, 621–629. not be ruled out that further malnourished polar bears went into the Gebbink, W.A., Sonne, C., Dietz, R., Kirkegaard, M., Born, E.W., Muir, D.C.G., Letcher, R.J., sea but did not reach Iceland. These animals would most likely have 2008b. Target tissue selectivity and burdens of diverse classes of brominated and drowned and would have been overlooked. These scenarios indicate chlorinated contaminants in polar bears (Ursus maritimus) from East Greenland. Environ. Sci. Technol. 42, 752–759. that the East Greenland population of polar bears may be more endan- Guerra, P., Fernie, K., Jiménez, B., Pacepavicius, G., Shen, L., Reiner, E., Eljarrat, E., Barceló, gered than anticipated. A recent simulation of the field metabolic rate D., Alaee, M., 2011. Dechlorane plus and related compounds in peregrine falcon – indicated that even relatively small reductions in feeding opportunities (Falco peregrinus) eggs from Canada and Spain. Environ. Sci. Technol. 45, 1284 1290. Jones, K.C., de Voogt, P., 1999. Persistent organic pollutants (POPs): the state of science. could affect the survival of the polar bears in East Greenland (Pavlova Environ. Pollut. 100, 209–221. et al., 2014). Klobes, U., Vetter, W., Glotz, D., Luckas, B., Skírnisson, K., Hersteinsson, P., 1998. Levels and enantiomeric ratios of organochlorine hydrocarbons in livers of arctic fox (Alopex lagopus) and adipose tissue and liver of a polar bear (Ursus maritimus) sampled in Acknowledgments Iceland. Int. J. Environ. Anal. Chem. 69, 67–81. McKinney, M.A., Letcher, R.J., Aars, J., Born, E.W., Branigan, M., Dietz, R., Evans, T.J., Gabrielsen, G.W., Peacock, E., Sonne, C., 2011. Flame retardants and legacy contami- Ólöf Guðrún Sigurðardóttir, Vilhjálmur Svansson at the Institute nants in polar bears from Alaska, Canada, East Greenland and Svalbard, 2005–2008. for Experimental Pathology, Keldur, Univeristy of Iceland, and Þorvaldur Environ. Int. 37, 365–374. Þór Björnsson, Icelandic Institute for Natural History helped with dissec- Muir, D.C.G., Backus, S., Derocher, A.E., Dietz, R., Evans, T.J., Gabrielsen, G.W., Nagy, J., Norstrom, R.J., Sonne, C., Stirling, I., Taylor, M.K., Letcher, R.J., 2006. Brominated tions of the polar bears and by taking the samples. We are also grateful flame retardants in polar bears (Ursus maritimus) from Alaska, the Canadian Arctic, to Carolin Hauler and Christoph Gallistl (University of Hohenheim, East Greenland, and Svalbard. Environ. Sci. Technol. 40, 449–455. Institute of Food Chemistry) for GC/MS measurements. Norstrom, R.J., Muir, D.C.G., 1994. Chlorinated hydrocarbon contaminants in Arctic marine mammals. Sci. Total Environ. 154, 107–128. Norstrom, R.J., Simon, M., Muir, D.C.G., Schweinsburg, R.E., 1988. Organochlorine contam- Appendix A. Supplementary data inants in arctic marine food chains: identification, geographical distribution, and temporal trends in polar bears. Environ. Sci. Technol. 22, 1063–1071. Pavlova, V., Nabe-Nielsen, J., Dietz, R., Svenning, J.-C., Vorkamp, K., Rigét, F.F., Sonne, C., Supplementary data to this article can be found online at http://dx. Letcher, R.J., Grimm, V., 2014. Field metabolic rate and PCB adipose tissue deposition doi.org/10.1016/j.scitotenv.2015.07.011. efficiency in East Greenland polar bears derived from contaminant monitoring data. PLoS ONE 9, e104037. Petersen, A.E., Haraldsson, T., 1993. Villt íslensk spendýr. In: Hersteinsson, P., References Sigbjarnarson, G. (Eds.), Landvernd and HIN, Reykjavík, pp. 74–78 (In Icelandic). Polischuk, S.C., Norstrom, R.J., Ramsay, M.A., 2002. Body burdens and tissue concentra- Batista, A., Vetter, W., Luckas, B., 2001. Use of focused open vessel microwave-assisted tions of organochlorines in polar bears (Ursus maritimus) vary during seasonal fasts. extraction as prelude for the determination of the fatty acid profile of fish — a Environ. Pollut. 118, 29–39. comparison with results obtained after liquid–liquid extraction according to Bligh Shen, L., Reiner, E.J., MacPherson, K.A., Kolic, T.M., Sverko, E., Helm, P.A., Bhavsar, S.P., and Dyer. Eur. Food Res. Technol. 212, 377–384. Brindle, I.D., Marvin, C.H., 2010. Identification and screening analysis of halogenated Bendig, P., Blumenstein, M., Hägele, F., Vetter, W., 2012. Hydrodebromination of norbornene flame retardants in the Laurentian Great Lakes: dechloranes 602, 603, decabromodiphenyl ether (BDE-209) in cooking experiments with salmon fillet. and 604. Environ. Sci. Technol. 44, 760–766. J. Agric. Food Chem. 60, 8521–8527. Skírnisson, K., 2009. Age determination and predicted life history of two polar bears Bendig, P., Hägele, F., Vetter, W., 2013. Widespread occurrence of polyhalogenated Ursus maritimus which swam to Iceland in June 2008. Natturufroeingurinn 78, compounds in fat from kitchen hoods. Anal. Bioanal. Chem. 405, 7485–7496. 39–45 (in Icelandic with English summary). Bytingsvik, J., Lie, E., Aars, J., Derocher, A.E., Wiig, Ø., Jenssen, B.M., 2012. PCBs and Skírnisson, K., Sigurðardóttir, Ó.G., 2010. Examinations on a female polar bear shot at OH-PCBs in polar bear mother-cub pairs: a comparative study based on plasma levels Ósland in Thistilfjörður. Report Sent to the Icelandic Institute for Natural History in 1998 and 2008. Sci. Total Environ. 417–418, 117–128. and The Environment Agency of Iceland (13 pp., In Icelandic). Covaci, A., Voorspoels, S., de Boer, J., 2003. Determination of brominated flame retardants, Stirling, I., Derocher, A.E., 2012. Effects of climate warming on polar bears: a review of the with emphasis on polybrominated diphenyl ethers (PBDEs) in environmental and evidence. Glob. Chang. Biol. 18, 2694–2706. human samples — a review. Environ. Int. 29, 735–756. Verreault, J., Gabrielsen, G.W., Chu, S., Muir, D.C.G., Andersen, M., Hamaed, A., Letcher, R.J., Daley, J.M., Leadley, T.A., Drouillard, K.G., 2009. Evidence for bioamplification of nine 2005. Flame retardants and methoxylated and hydroxylated polybrominated (PCB) congeners in yellow perch (Perca flavascens) eggs diphenyl ethers in two Norwegian arctic top predators: glaucous gulls and polar during incubation. Chemosphere 75, 1500–1505. bears. Environ. Sci. Technol. 39, 6021–6028. de Wit, C.A., Herzke, D., Vorkamp, K., 2010. Brominated flame retardants in the Arctic Vetter, W., 2001. AGC/ECNI–MS method for the identification of lipophilic anthropogenic environment — trends and new candidates. Sci. Total Environ. 408, 2885–2918. and natural brominated compounds in marine samples. Anal. Chem. 73, 4951–4957. Dietz, R., Rigét, F.F., Sonne, C., Born, E.W., Bechshøft, T., McKinney, M.A., Letcher, R.J., Vetter, W., 2006. Marine halogenated natural products of environmental and food 2013a. Three decades (1983–2010) of contaminant trends in East Greenland polar relevance. Rev. Environ. Contam. Toxicol. 188, 1–57. bears (Ursus maritimus). Part 1: legacy organochlorine contaminants. Environ. Int. Vetter, W., Rosenfelder, N., 2008. Gas chromatographic retention data of environmentally 59, 485–493. relevant polybrominated compounds. Anal. Bioanal. Chem. 392, 489–504. Dietz, R., Riget, F.F., Sonne, C., Born, E.W., Bechshoeft, T., McKinney, M.A., Letcher, R.J., Vetter, W., Weichbrodt, M., Hummert, K., Glotz, D., Luckas, B., 1998. Combined 2013b. Three decades (1983–2010) of contaminant trends in East Greenland polar microwave-assisted extraction and gel permeation chromatography for the determi- bears (Ursus maritimus). Part 2: brominated flame retardants. Environ. Int. 59, nation of chlorinated hydrocarbons in seal blubber and cod livers. Chemosphere 37, 494–500. 2439–2449. Durner, G.M., Whiteman, J.P., Harlow, H.J., Amstrup, S.C., Regehr, E.V., Ben-David, M., von der Recke, R., Vetter, W., 2008a. Congener pattern of hexabromo biphenyls in marine 2011. Consequences of long-distance swimming and travel over deep-water pack biota from different proveniences. Sci. Total Environ. 393, 358–366. ice for a female polar bear during a year of extreme sea ice retreat. Polar Biol. 34, von der Recke, R., Vetter, W., 2008b. Anaerobic transformation of polybrominated 975–984. biphenyls with the goal of identifying unknown hexabromobiphenyls in Baltic cod Feo, M.L., Barón, E., Eljarrat, E., Barceló, D., 2012. Dechlorane Plus and related compounds liver. Chemosphere 71, 352–359. in aquatic and terrestrial biota: a review. Anal. Bioanal. Chem. 404, 2625–2637. WHO, 1994. Chlordane. Environ. Health Crit. 34, 1–82.