Research Trends in Mercury in Hair of Greenlandic Polar Bears (Ursus maritimus) during 1892-2001 R . D I E T Z , * ,† F . R I G E T , † E . W . B O R N , ‡ C. SONNE,† P. GRANDJEAN,§ M. KIRKEGAARD,† M. T. OLSEN,† G. ASMUND,† A. RENZONI,| H. BAAGØE,⊥ AND C. ANDREASEN# Department of Arctic Environment, National Environmental Research Institute, P.O. Box 358, Frederiksborgvej 399, DK-4000 Roskilde, Denmark, Greenland Institute of Natural Resources, P.O. Box 570, DK-3900 Nuuk, Greenland, Institute of Public Health, University of Southern Denmark, Winslowparken 17, 5000 Odense, Denmark, Department of Environmental Biology, Universita` di Siena, Via delle Cerchia 3, 53100 Siena, Italy, Vertebrate Department, Zoological Museum, Universitetsparken 15. DK-2160 Copenhagen Ø, Denmark, and Department of Archaeology, Greenland National Museum and Archives in Nuuk, P.O. Box 145 Nuuk, DK-3900, Greenland
Mercury concentrations in hair from 397 Greenland polar bears (Ursus maritimus) sampled between 1892 and 2001 were analyzed for temporal trends. In East Greenland the concentrations showed a significant (p < 0.0001, n ) 27) increase of 3.1%/year in the period 1892-1973. In Northwest Greenland, a similar (p < 0.0001, n ) 69) increase of 2.1%/year was found, which continued until 1991, when the most recent samples were obtained. In East Greenland, a significant (p ) 0.009, n ) 322) decrease of 0.8%/year was observed after 1973. Two Northwest Greenland samples from 1300 A.D. had a mean value of 0.52 mg/kg of dry weight, which can be considered as a baseline level. The Hg concentration during 1985-1991 from Northwest Greenland (mean value of 7.45 mg/kg of dry weight) was more than 14-fold higher than the assumed baseline level from 1300 A.D. from the same region (i.e., about 93% anthropogenic). Although a decrease was found in East Greenland after 1973, the concentration is still ca. 11fold higher than the baseline level (i.e., about 90% anthropogenic).
Introduction An estimated 200-300 tons of mercury (Hg) per year from various human activities at midlatitudes are transported to the Arctic by atmospheric processes, ocean currents, and rivers (1, 2). It has been discovered that global elemental Hg is being depleted from the atmosphere in the Arctic each * Corresponding author phone: +45 46301938; fax: +45 46301914; e-mail:
[email protected]. † National Environmental Research Institute, Denmark. ‡ Greenland Institute of Natural Resources, Greenland. § University of Southern Denmark, Denmark. | Universita ` di Siena, Italy. ⊥ Zoological Museum, Copenhagen, Denmark. # Greenland National Museum and Archives, Nuuk, Greenland. 1120
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 4, 2006
spring, showing that the Arctic acts as a sink for globally emitted Hg (3-6). These events emphasize the global nature of Hg pollution. Although Hg is a naturally occurring element and, as such, is always present in the environment, worldwide human activity has led to a severalfold increase in concentrations in the Arctic environment compared to pre-industrial times (2). Hg also biomagnifies in the Arctic ecosystem, and concentrations in higher trophic marine species (seabirds, seals, and whales) are relatively high, resulting in high human exposure via consumption of these animals. In some areas, Hg concentrations in the environment continue to increase (7), causing retention levels in some marine mammals, birds, and fish, where biological effects might be expected (2). The Arctic Monitoring and Assessment Program (AMAP) has demonstrated that, in certain Arctic or sub-Arctic areas (e.g., the Faroe Islands), humans can have a daily dietary intake of Hg that exceeds the threshold levels for negative effects on children’s neurological development (2). In the future, effects are likely to be more apparent in Arctic areas where Hg continues to increase. Apart from a few data primarily on Boreal species (8), no long-term (10-year or more) data on concentrations of Hg in soft tissues exist. Data from Greenland (9, 10) indicate that there is still a need for longer time series studies on Hg in soft tissues (8, 11). However, hard tissues (e.g., bone, teeth, and hair) have the potential for an extension of backdated Hg data series. Data on Hg spanning several centuries exist in human teeth from Norway (12, 13), human and seal hair from Greenland (14), and human hair from Canada (15). Similar studies have been performed by Outridge et al. (16, 17), who compared Hg concentrations in the teeth of recently collected beluga (Delphinapterus leucas) and walrus (Odobenus rosmarus) to historical archaeological samples. Work on laboratory rats and, e.g., polar bear hair and beluga teeth, indicates that hair, teeth, and internal organs are all correlated with the intake of organic and inorganic Hg (1821). Thus, examination of historic samples of hard tissue such as hair or teeth reflects changes in intake and softtissue exposure over time. Most of the previous studies compare only two periods based on relatively few samples. In the present study, we determined whether a time trend of Hg concentrations could be detected in hair samples from a total of 399 Greenland polar bears (Ursus maritimus) collected during the period 1892 and 2001. For comparison with “prepollution” conditions, two polar bear hair samples dated at 1300 A.D. from Northwest Greenland were included in the study.
Material and Methods Samples. A total of 436 polar bear hair samples, collected in Greenland during the period 1300-2001 A.D., were analyzed for Hg concentrations. Of these, only 399 were used in the final time trend analysis because of sublimate contamination of 37 museum samples (see Table 1). The material was composed of 345 individuals from East Greenland (EG) and 91 from Northwest Greenland (NWG). The recent samples were either from “The Polar Bear Expeditions” lead by Dr. Vibe in EG (n ) 58), 1973-1974, or from samples obtained by local hunters in Greenland, 1984-2001 (EG, n ) 264; NWG, n ) 62). Historical samples (n ) 37), later discarded from this time trend analysis because of contamination, was collected between 1850 and 1947 from traditional Inuit clothing and other Inuit artifacts held at the National Museum 10.1021/es051636z CCC: $33.50
2006 American Chemical Society Published on Web 01/11/2006
TABLE 1. Numbers of Observations, Means, Ranges, and Median Concentrations of Hg (mg/kg of dry weight) during 10-Year Intervals of Hair Samples of Polar Bears from Northwest and East Greenland, between 1885 and 2001 and during 1300 A.D median
1300 TTFa
1.36 0.858 0.391 1.61 6.04 4.25 5.36 4.90
2.6 2.1 0.8 3.1 13.9 7.4 11.2 10.5
Northwest Greenland (n ) 69) 2 0.518 0.479 0.558 0.518 2 1.53 0.430 2.62 1.53 2 6.37 2.11 10.6 6.37 1 3.46 3.46 3.46 3.46 62 7.45 2.96 14.2 7.61
1.0 3.0 12.3 6.7 14.4
region/year
n
1885-1894 1905-1914 1925-1934 1945-1954 1965-1974 1975-1984 1985-1994 1995-2004
2 3 1 2 58 20 147 97
1300 1915-1924 1935-1944 1955-1964 1985-1994
mean
min
max
East Greenland (n ) 330) 1.36 0.940 1.78 1.07 0.719 1.63 0.391 0.391 0.391 1.61 1.03 2.18 7.22 1.69 21.6 3.87 1.78 5.36 5.82 1.13 24.2 5.42 0.791 17.7
a Time trend factor (1300 TTF) was calculated in relation to the presumed baseline concentrations obtained from two samples from Nuulliit, Northwest Greenland, from 1300 A.D.
of Copenhagen, Copenhagen, Denmark. Another seven samples from EG (1892-1948) and four from NWG (19201939) were obtained from polar bear skins kept at the Zoological Museum in Copenhagen, Copenhagen, Denmark. Other historical samples came from the Greenland National Museum and Archives in Nuuk, Greenland, where samples from four polar bears (EG, 1950; NWG, 1300-1960) were obtained from traditional Inuit clothing, two of which came from an excavation of an Inuit dwelling in NWG (Nuulliit) and radiocarbondated to ca. 1300 A.D. All other dating refers to the year the sample was obtained in Greenland. Laboratory Analyses of Hair Samples. Department of Arctic Environment. The majority of the Hg analyses were performed at the Department of Arctic Environment (DAE) laboratory, Roskilde, Denmark (which is DANAK certified). All hair specimens, approximately 1 g each, were stored in a polyethylene plastic bag at room temperature or in a freezer until analysis. Before the hair samples were wet-ashed, they were rinsed in a standard detergent (RBS 35; Bie & Berntsen A-S), containing a sodium hydroxide/sodium hypochlorite solution, to clean the sample and prevent possible external contamination, according to the method of Hansen et al. (14). Polyethylene gloves were used to transfer approximately 0.5 g of hair to the tarred Teflon liner of an Anton Paar Multiwave 3000 microwave sample preparation system. The digestion and Hg analyses performed atomic absorption spectrometry (AAS; hydride generation and the flow injection analyses) have previously been described by Asmund et al. (22). The detection limit was 0.005 mg/kg of dry weight. Analytical quality was ensured by repeated analyses and by frequent analysis of various certified reference materials [TORT-2 (lobster hepatopancreas), DORM-2, and Dolt-3] supplied by the National Research Council of Canada (Marine Analytical Chemistry Standards Program). The DAE laboratory participated in the international intercomparison exercises conducted by the International Council for the Exploration of the Sea (ICES), EEC (QUASIMEME), and the Department of Fisheries and Oceans, Winnipeg, Canada (22). This article documents that the 95% confidence interval is better than 12.5% for concentrations larger than 0.1 µg/g. University of Southern Denmark. The polar bear hair from the 1973-1974 Polar Bear Expeditions to EG was analyzed at the Institute of Public Health, University of Southern Denmark, by a procedure developed for human hair analysis
(23). For determination of hair Hg concentrations at the University of Southern Denmark, an accurately weighed hair sample of 0.01-0.1 g was placed in a PTFE-lined digestion vessel, before microwave digestion and preparation, as previously described (23). Mercury analysis in duplicate was performed by flow-injection cold-vapor atomic absorption spectrometry (Perkin-Elmer model 5100 instrument with FIAS-200 and AS-90 attachments). The total analytical imprecision was estimated to be 12.6% and 5.8% at Hg concentrations of 1.56 and 11.5 µg/g, respectively. The accuracy of the Hg determinations in human hair was ensured by using the certified reference material CRM 397 (BCR, Brussels, Belgium) and powdered hair HH-1 (IAEA, Vienna, Austria) as quality-control materials. University of Sienna. Finally, previously published Hg data on 64 of our 330 polar bears from EG sampled between 1978 and 1989 were obtained from the University of Sienna, Sienna, Italy (18). Documentation on the quality analysis of this laboratory is given under the laboratory comparison section. All data are presented on a dry weight basis. Laboratory Comparison. Three laboratories (National Environmental Research Institute, Denmark; University of Southern Denmark; and University of Sienna, Italy) were involved in the analyses of Hg in the East Greenland polar bear hair samples, and as a result, the intercomparability was important. A subsample of 10 specimens from 19731974 was therefore reanalyzed by the National Environmental Research Institute, and the results were not significantly different (P ) 0.8184, paired t-test) from the earlier analysis of these samples made by the University of Southern Denmark. The results from University of Sienna did not allow for a direct comparative study, as no additional samples were available, but the previously published results were not statistically different from the NERI data covering the same time span. It was therefore concluded that the results from the three laboratories were of high and comparable quality and that the results were not influenced by differences in analytical methods. Data Analysis. Prior to the statistical analyses, individual Hg concentrations were log-transformed to obtain normality and the homoscedasticity of the variance. Analysis of the variance was used when testing for possible differences in mean concentrations between laboratories, between regions, or between sampling periods. Tukey’s post hoc test was used to test for differences between group means. Linear regression analyses were used to test for age dependency and correlation studies of Hg concentrations in polar bear hair. Both the parametric Pearson’s correlation and the nonparametric Spearman correlation analysis were applied. To test for temporal trends linear regression analyses were applied using individual log-transformed Hg concentrations as the dependent variable and year as the explanatory variable. Potential Contamination of Historical Samples. When using museum specimens, there is a possibility that the artifacts have been contaminated by mercury-containing preservation agents. Mercury, however, has not been used as often as arsenic, because of its high toxicity to humans (25, 26). The primary source for Hg is sublimate (mercuric chloride, HgCl2), but because of the toxicity, sublimate was phased out at the end of the 19th century. The hair Hg patterns from the three archives (i.e., the National Museum, the Zoological Museum, and the Greenland National Museum and Archives) were therefore evaluated with respect to comparability. The Hg concentrations from the National Museum [arithmetic mean (AM) ) 1325 mg/kg of dry weight, standard deviation (SD) ) 2779 mg/kg of dry weight, geometric mean (GM) ) 63.4 mg/kg of dry weight, range ) 2.70-11639 mg/kg of dry weight, n ) 37] were surprisingly variable and significantly higher (t-test: p < 0.0001, t ) 7.44) VOL. 40, NO. 4, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1121
than results from the other two sources. The results from the Zoological Museum (AM ) 2.21 mg/kg of dry weight, SD ) 2.90 mg/kg of dry weight, GM ) 1.39 mg/kg of dry weight, range ) 0.391-10.6 mg/kg of dry weight, n ) 11) from 1892 to 1948 and the results from the Greenland National Museum and Archives (AM ) 1.38 mg/kg of dry weight, SD ) 1.41 mg/kg of dry weight, GM ) 0.99 mg/kg of dry weight, range ) 0.479-3.46 mg/kg of dry weight, n ) 2/4) from 1300 to 1960 were of the same order of magnitude, showed limited variability, and were therefore thought to reflect mercury exposures in vivo at the time of collection. There are no records that Hg compounds were ever used as preservatives at the Zoological Museum (Jeppe Møhl, personal communication, 2003), whereas sublimate had been seen in the shelves, even in the 1950s, of the National Museum, Copenhagen (Ole Schmidt, personal communication, 2003). This information on use, the comparison of levels, and the variability of samples collected before 1960 all suggested that the National Museum samples had been contaminated by sublimate, whereas this was not the case for the Zoological Museum and Greenland National Museum and Archives. Therefore, the 37 samples from the National Museum were excluded from further data treatment. Age and Seasonability of Hair Samples. As no ages exist for a large proportion of the samples and as no systematic age differences in samples were expected over time, no attempt was made to conduct normalization relative to age. Previous investigations also omitted such normalization, because they did not find an age-related accumulation of Hg in polar bear hair (18). The lack of age-related accumulation is in accordance with the general acceptance that hair Hg reflects the blood levels in the time period the hair is growing (27). Also, the blood concentration reflects the recent intake and the elimination half-life of methylmercury in blood, which would be no more than a few months for a large mammal such as the polar bear (28). Information was also lacking on sampling season for some of the old samples. However, as explained by Born et al. (18), the winter fur is grown from September to October, whereas part of this fur is molted at the end of May. We have previously shown that no significant difference can be detected between summer fur (June-August) and winter fur (September-May) (18) of Greenland polar bears. Hence, the season of sampling was not believed to be a source of error for the time trend comparisons.
FIGURE 1. Mercury in East Greenland polar bear hair (n ) 27) between 1892 and 1973 showing a significant (p < 0.0001) increasing trend of ca. 3.1%/year (slope ) 0.031).
TABLE 2. Results of Linear Regression Analysis Using Individual log-Transformed Hg Concentrations in Polar Bear Haira period
r
p
N
1892-1973 1973-2001
East Greenland 0.031
