Natural Fluctuations of Mercury and Lead in Greenland Lake

Figure 1 Map over the sampling area with sampling sites in Kangerlussuaq, West ... Lake B was sampled with a gravity corer (i.d. 8.4 cm, HTH-Teknik, S...
0 downloads 0 Views 185KB Size
Download PDF
Environ. Sci. Technol. 2006, 40, 90-95

Natural Fluctuations of Mercury and Lead in Greenland Lake Sediments CAROLA LINDEBERG,* RICHARD BINDLER, AND INGEMAR RENBERG Environmental Change Assessment, Department of Ecology and Environmental Science, Umea˚ University, SE-901 87 Umea˚, Sweden OVE EMTERYD Department of Forest Ecology, Swedish University of Agricultural Sciences, SE-901 83 Umea˚, Sweden EDVARD KARLSSON National Defense Research Establishment, NBC Department, SE-901 82 Umea˚, Sweden N. JOHN ANDERSON Department of Geography, Loughborough University, Loughborough, Leicestershire LE11 3TU, U.K.

Given the current scenario of increasing global temperatures, it is valuable to assess the potential influence of changing climate on pollution distribution and deposition. In this study we use long-term sediment records from three lakes (spanning ca. 1000, 4800, and 8000 years, respectively) from the Greenland west coast to assess recent and longterm variations in mercury (Hg) and lead (Pb), including stable Pb isotopes (206Pb and 207Pb), in terms of pollution and climate influences. The temporal trends in sediment deposited from about the mid-19th century and forward are in general agreement with the history of industrial emissions at lower latitudes. Therefore, in recent sediment a possible influence from changing climate is difficult to assess. However, by using deeper sediment layers we show that changes in Greenland climate caused changes in the lake influx of material from regional aeolian activity, which resulted in large fluctuations in Hg and Pb concentrations and 206Pb/207Pb ratios. The aeolian material is primarily derived from glacio-fluvial material with low Hg and Pb concentrations and a different isotopic composition. For one of the lakes, the fluctuations in Hg concentrations (10 to 70 ng g-1) prior to the 19th century are equal to the anthropogenic increase in the uppermost layers, suggesting that when studying recent concentrations and time trends of pollution in relatively low-contaminated areas such as the Arctic, the early natural fluctuations must be considered.

in the Arctic and highly affect the ecosystem (e.g., 1). For Pb and other nonvolatile compounds, transportation toward the Arctic follows prevailing winds and weather systems, which can move these pollutants rapidly from Eurasia across the polar region to northern Canada and Alaska (2). For Pb it is possible to trace emission sources and establish transportation routes using stable isotopes (e.g., 206Pb/207Pb ratio) (2). For volatile compounds, such as elemental Hg and many organic compounds, atmospheric residence times can be longsa year or more in the case of Hg. Due to their volatility they have a temperature-driven re-evaporation to the atmosphere from contaminated soils and waters (3). Today’s climate change with increasing global temperatures may influence the accumulation of atmospherically supplied pollutants in the Arctic. Global climate models indicate that the greatest relative effects of climate change will be at high latitudes (4), and it is important, therefore, that the interaction between climate and Arctic pollutant deposition is considered. The recent temporal trends for Arctic environmental Pb and Hg are in general agreement with the history of industrial emissions (e.g., 1, 5-9), and a possible influence from changing climate is difficult to disentangle. However, in long-term records, preserved in lake sediments, peat, and glacial ice, there is a potential to assess past changes in Pb and Hg in relation to past climate changes. Arctic climate has varied during the Holocene with periods that have been warmer or colder than the present, such as the “Medieval Warm Period” and the “Little Ice Age” (10, 11). With the exception of ice records from the Greenland Summit (12-14) and some minerotrophic mires (15) there are few long-term metal records from high latitudes. Furthermore, little attention has been given to the influence of climate in these temporal metal records. A previous study noted spatial differences in Pb and Hg deposition in relation to regional climate gradients in western Greenland with higher Pb deposition nearest the coast (higher precipitation) and higher Hg deposition nearest the ice (presumably related to meteorology at the ice margin) (16). In this study we investigate Pb and Hg in lake sediments from the same area and assess long-term records of these metals in terms of natural fluctuations, driven by climatic variability, as well as pollution. The area is fairly well studied with respect to limnology, meteorology, and contemporary pollution (1619). Also, changes in Greenland Holocene climate are well reconstructed using ice cores (10, 11), lake sediments (2023), pollen records (20), and soil profiles (24). One 20-cm and two 3-m long sediment cores, covering the last ca. 1000, 4800, and 8000 years, respectively, are used to study longterm fluctuations in Pb and Hg concentrations and 206Pb/ 207Pb isotope composition. The sediment properties were described using magnesium (Mg), calcium (Ca), and carbon (C) content. In addition, the composition of aeolian inputs is assessed based on the Pb (isotopes and concentration) and Hg concentration of exposed minerogenic soil.

Materials and Methods Introduction Until recently the Arctic was considered as a pristine area due to few local emission sources. There is now increasing evidence that long-range pollutants such as persistent organic pollutants (POPs), lead (Pb), and mercury (Hg) are deposited * Corresponding author phone: +46(0)90-7867101; fax: +46(0)90-7866705; e-mail: [email protected]. 90

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 1, 2006

Study Area. Kangerlussuaq (Søndre Strømfjord) is a 170km-long fjord on Greenland’s west coast (66 to 67° N and 50 to 55° W) (Figure 1). At the head of the fjord, about 25 km from the ice margin, the small village and airport of Kangerlussuaq are located. Prior to construction of the airport during World War II the region was not permanently settled. The present-day population is about 400 and several thousand tourists annually visit the area. Mean annual precipitation in the inner part of the fjord is very low, less 10.1021/es051223y CCC: $33.50

 2006 American Chemical Society Published on Web 12/02/2005

FIGURE 1. Map over the sampling area with sampling sites in Kangerlussuaq, West Greenland. than 150 mm year-1 with a maximum in late summer. Mean summer temperature is 9 °C and mean winter temperature is -18 °C. Despite general global warming the last one hundred years, long-term climate records from Nuuk and Illulissiat indicate that western Greenland had periods of cooling during the 20th century (25). Given the low precipitation levels, surface runoff is limited and essentially confined to the spring melt period in late May and June. The exceptionally arid conditions and the ample supply of finegrained glacio-fluvial material from the outwash flood plains from the ice sheet (sandars) give rise to large amounts of aeolian material, observed both as air dust and as thin loess deposits across the landscape, up to ∼80 km from the present ice margin. The local geology is predominantly granodioritic gneiss and vegetation is characterized by a dwarf shrub tundra (17). Two of the sampled lakes are located northeast of Kangerlussuaq: lake B, 55 km away (lat 67°16′, long. 51°35′, altitude 671 m asl, lake depth 32 m), and lake G, 20 km away (lat 67°03′, long. 51°13′, altitude 338 m asl, lake depth 15 m). Lake SS16 is located nearer to the ice-sheet margin, 17 km southeast of the village (lat 66°91′, long. 50°45′, altitude 477 m asl, lake depth 12 m) (Figure 1). These lake names/ numbers are not official, but function only as site identifications in a research lake database (17). Lake B has a visible presence of chironomids in the surface sediments and lake G is tightly laminated. Close to lake SS16 exposed minerogenic soil was sampled. Sampling. The sediment cores and the minerogenic soil were collected from May 1999 to April 2001. Lake B was sampled with a gravity corer (i.d. 8.4 cm, HTH-Teknik, SE976 31 Lulea˚, Sweden) to a depth of 20 cm. In the field the first 10 cm of the core was sectioned in 0.5-cm slices and thereafter in 1-cm slices. Every slice was placed into a polypropylene container with a lid of polyethylene. For lakes G and SS16 the sediments were sampled to a depth of 3.0 and 2.9 m, respectively, using a freeze corer for the unconsolidated surface sediments (26) and a Russian peat corer (i.d. 8 cm, length 1 m) for the more consolidated sediments below. These cores were brought intact (frozen and below +4 °C, respectively) to the laboratories in Copenhagen or Umeå, where they were sub-sampled. The exposed minerogenic soil was collected in a bottle made of high-density linear polyethylene. All samples were stored cold (frozen or maximum +4 °C, respectively) and in the dark until freeze-drying and homogenizing prior to analysis. Analyses. Thermal-vapor AAS (Leco AMA 254) was used for total Hg determination. The calibration curve was based on the same matrix as the samples, i.e., analyses of different masses (5-250 mg) of certified reference sediments MESS-2 and MESS-3 (National Resources Canada, Institute for National Measurement Standards; certified values of 92 ( 9 ng Hg g-1 and 91 ( 9 ng Hg g-1, respectively). Concentrations were verified by analyses of other standard reference materials: Apple Leaves NIST-SRM 1515, Olea BCR 957, and Calcareous Loam BCR 141; analyzed values were (mean ng g-1 ( SD, n) 42 ( 4, n ) 39 (certified values 44 ( 4); 290 ( 25, n ) 21 (280 ( 20); 54 ( 10, n ) 10 (56.8 ( 4.3), respectively.

Replicate samples were well comparable (mean deviation 8%) and blank values were mostly well below 1% of the lowest sample concentration. Concentrations of Hg are reported as ng g-1 dry sediment. Inductively coupled plasma mass spectrometry (ICP-MS, Perkin-Elmer model ELAN 6100), following a strong acid digestion (130 °C, 2- 5 h in open Teflon vessels, HNO3 + HClO4, 10:1 v/v), was used to analyze Pb, Mg, and Ca contents in selected samples. For Pb, the certified standard, SPEX ICPMS-2 (Spex CertiPrep Certified Reference Materials) was used to make a 5-point calibration within the range of 0-500 µg/ L. Concentrations were verified by analyses of certified reference sediment IAEA-SL1 and two internal reference materials (lake sediment) that we have analyzed repeatedly over a 5 and 10-year period; analyzed values were (mean µg g-1 ( SD, n) 36 ( 2, n ) 5 (certified value 37.7 ( 7.4); 15 ( 2, n ) 5 (16 ( 1); 19 ( 3, n ) 2 (21 ( 2), respectively. Analyses of Pb masses 206 and 207 were made using dwell times of 50 ms and were corrected empirically by repeated analyses of the NIST SRM 981 reference material (fraction constant 0.2-0.7% per amu). The 206Pb/207Pb ratio of 1.217 ( 0.003 obtained for the IAEA-SL1 was the same as that reported by Farmer et al. (27). Calibration curves for Mg and Ca were made using certified high-purity ICP-MS multielement standards (ICP-MSCS-M). Concentrations were verified against reference material Corn Bran NIST-RM 8433; Ca (mean mg kg-1 ( SD, n) 431 ( 13, n ) 10 (certified value 420 ( 19); Mg 824 ( 32, n ) 10 (818 ( 30). Concentration for Pb is reported as µg g-1 dry sediment and for Mg and Ca contents is reported as mg g-1. The total carbon content of the sediment samples was determined using a PerkinElmer Elemental CHNS analyzer. Carbon content was verified against certified standard materials (PerkinElmer Cystine and Acetanilide) and replicate samples were within 2%. The C content is reported as percent (%) of dry sediment mass. Sediment Dating and Chronology. For the lakes B, G, and SS16 the sediment record was dated using a combination of radiometric analyses, which include 210Pb and 226Ra, as well as 137Cs and 241Am for lake B (28, 29), and a constantrate-of-supply (CRS) model for recent sediments, and 14C dating of older deposits (lakes G and SS16). Calendar years were obtained from the 14C-calibration tables from Stuvier et al. (30) using the 10-year terrestrial calibration curve. The series of overlapping surface and deeper cores from lakes G and SS16, respectively, are matched using depth, carbon content, and Hg and Pb.

Results Sediment and Soil Properties. The deepest sediment layer at 20 cm in lake B is between 700 and 1100 years old based on the mean recent sedimentation rate (0.008 ( 0.002 g cm-2 y-1) and an extrapolation of the 210Pb-dating. The long sediment profiles from lakes G (3.0 m) and SS16 (2.9 m) span the last 4800 and 8000 years, respectively. The total carbon content in the sediments is low and relatively constant. For lake B the carbon content is below VOL. 40, NO. 1, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

91

FIGURE 4. Lake B: Pb concentration (µg g-1), 206Pb/207Pb ratio, and Hg concentration (ng g-1).

FIGURE 2. Lake G: (a) Pb concentration (µg g-1), 206Pb/207Pb ratio, Hg concentration (ng g-1), and Ca, Mg (mg g-1), and carbon content (%) in the freeze core overlapped with the long core; (b) Pb concentration, ratio, and Hg concentration for the last 200 years. Note the overlap for the freeze core and the long core from 1910 to 1940.

Hg. Down-core Hg profiles for the investigated lake sediments are presented in Figures 2a, 2b, 3a, 3b, and 4. The Hg concentration reported is the total Hg, i.e., both mineral Hg and atmospheric deposition. In deeper sediment layers deposited before the mid-19th century, there are substantial variations in Hg concentrations in all the lakes. Lake B varies between 40 and 110 ng g-1, lake G varies between 10 and 70 ng g-1, and lake SS16 varies between 10 and 120 ng g-1, except for the four deepest layers composed of minerogenic glacial silt with a Hg concentration only between 2 and 8 ng g-1. In lake SS16 there is a generally increasing trend in concentrations throughout the core, whereas lakes G and B show no long-term trend. In lakes G and SS16 there is only a minor correlation between the total carbon content and Hg; r ) 0.3 for both lakes (n ) 98 for lake G and n ) 104 for lake SS16). In the laminated sediments of lake G a slight negative correlation between Hg concentration and Ca content is discerned (r ) 0.4, n ) 27), i.e., in CaCO3 rich sediment layers the Hg concentration is somewhat lower. For recent sediments (i.e., past 200 years) the general pattern for Hg concentrations conforms to established trends where concentrations increase during the 19th or early 20th centuries. Even though the Hg concentrations in the uppermost sediment layers in lake G are within the range of variation observed in deeper layers, the concentration increase from the beginning of the 19th century until today is about 4-fold (Figure 2b). In lakes B and SS16 the contemporary increase is about 2-fold and 3-fold, respectively (Figures 3b and 4). For these two lakes this modern increase is quite distinct from and superimposed over the concentration changes in deeper sediment layers. The Hg concentration in the exposed minerogenic soil is 7 ng g-1.

FIGURE 3. Lake SS16: (a) Pb concentration (µg g-1), 206Pb/207Pb ratio, Hg concentration (ng g-1), and Ca, Mg (mg g-1), and carbon content (%) in the freeze core overlapped with the long core; (b) Pb concentration and ratio, and the Hg concentration for the last 200 years. 6% by mass and for both lakes G and SS16 the mean carbon content is 14%, except for an increase in the uppermost sediment layers (Figures 2a and 3a) and a decrease in the four deepest layers in lake SS16, where the minerogenic glacial silt deposited 5500-6000 BC has a low carbon content. Changes in within-lake processes are evidenced by the occurrence of laminations in lake G, which are mostly composed of light layers with CaCO3. The Ca and Mg contents in the sediment cores from lakes G and SS16 show substantial fluctuations (Figures 2a and 3a). For the exposed minerogenic soil material, mostly consisting of aeolian input to the area around lake SS16, the Ca content is 4 mg g-1 and the Mg content is 3 mg g-1. 92

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 1, 2006

Pb. Down-core Pb profiles for the investigated lake sediments are presented in Figures 2a, 2b, 3a, 3b, and 4. The Pb extracted in the strong-acid digestion is a combination of mineral Pb and atmospheric deposition. Prior to the mid19th century the Pb concentrations in the lakes are low, with variations between 1 and 1.5 µg g-1 in lake B, 0.5 and 1.5 µg g-1 in lake G, and between 0.7 and 1.9 µg g-1 in lake SS16. During the same time period the 206Pb/207Pb ratios vary substantially: between 1.29 and 1.45 in lake B, between 1.28 and 1.57 in lake G, and between 1.24 and 1.40 in lake SS16. Notably, lakes G and SS16 have high isotope ratios in sediment deposited between 1000 BC and AD 500. In lake B, and to some extent also in the longer cores, high isotope ratios are also observed about AD 1600. The Hg and Pb concentrations are positively correlated in sediment deposited before the mid-19th century: r ) 0.9 in lake B (n ) 15), r ) 0.6 in lake G (n ) 23), and r ) 0.5 in lake SS16 (n ) 27). There is no correlation between the total carbon content and Pb concentration: r ) 0.05 in lake G (n ) 45) and r ) 0.06 in lake SS16 (n ) 38). In the laminated lake G the Pb concentration correlates slightly positively with the Ca and Mg content: r ) 0.5 (n ) 27) and r ) 0.4 (n ) 27), respectively.

In recent sediments from lakes B, G, and SS16, Pb concentrations increase between 2 and 5 times the values in deeper sediment layers to maximum values of 2.9, 3.8, and 8.7 µg g-1 in sediments dated to about the 1980s, 1960s, and 1920s, respectively (Figures 2b, 3b, and 4). The 206Pb/207Pb ratio declines simultaneously to low values of about 1.20 in lakes G and SS16 and 1.24 in lake B. The concentration and isotope ratio in the exposed minerogenic soil are 1.1 µg g-1 and 1.43, respectively.

Discussion Due to the substantial input of long-range transported Hg and Pb in the Arctic region (e.g., 1, 5-9, 16, 18), the higher Hg and Pb concentrations and decreasing 206Pb/207Pb ratio observed in the uppermost sediment layers in this investigation are mainly results of anthropogenic emissions at lower latitudes. The temporal trends for Hg and Pb for the period after the Industrial Revolution in the mid-19th century agree generally with anthropogenic emission trends from both Europe and North America (31-33). A former investigation showed that the anthropogenic Pb deposited in lakes at the Greenland west coast mainly came from emissions in Eurasia (18). The Pb concentrations found in that study as well as in this are much lower than concentrations observed in other lake sediments from high latitude regions, such as the Canadian Arctic with concentrations over 20 µg g-1 in recently deposited sediment layers (8) and Greenland marine sediments with concentrations between 10 and 17 µg g-1 (7). This is likely caused by the very low precipitation in the area, resulting in low Pb deposition. Even though pre-industrial pollution is well established, anthropogenic emissions cannot be the reason for the large fluctuations in Hg and Pb concentrations and 206Pb/207Pb ratios in older sediments (Figures 2a, 3a, and 4). Although Martinez-Cortizas et al. (34) found indications of Hg pollution as early as 2500 years ago in a Spanish peat record, more widespread evidence of long-range Hg pollution is only dated to about 200-500 years ago (9, 35-37). For Pb early (ca AD 0) atmospheric deposition has been reported for Europe (e.g., 15, 33, 38) and Greenland (12-14) but this early Pb pollution was still very small (e1 µg m-2 yr-1) (14) and not sufficient to cause substantial concentration changes in lake sediment (18, 33, 38). Additional supporting evidence for the nonanthropogenic source for the Pb deposited in the investigated sediments prior to the Industrial Revolution is the increasing 206Pb/207Pb ratio during the period from 1000 BC to AD 500 in lakes G and SS16 and a small increase observed about AD 1600 in lake B. These ratio increases are opposite the declining trend caused by historical pollution, because the atmospheric Pb pollution from Europe had low 206Pb/207Pb ratios (1.161.19) (14, 33, 38). This is also the ratio observed in modern Arctic air affected by anthropogenic emissions (2). Therefore, to explain the observed pre-industrial fluctuations in Hg and Pb concentrations, and particularly the Pb isotopic composition in deeper sediment layers, a mechanism other than pollution must be sought. It is well-known that the carbon content can influence the Hg and Pb concentrations of sediments. The poor correlation between Hg and Pb concentration and total carbon content in the older sediments in lakes G and SS16 may be due to calcium carbonate formation; at least in lake G with weak correlations between the Ca content and the Hg and Pb concentrations. Variations in within-lake processes could contribute to changes in concentration either through dilution or enrichment due to changes in productivity, or enhanced scavenging in the water column and sedimentation with precipitates formed in the water column. With the dating resolution of our study we are not able to determine to what extent changes in concentration are influenced by variations in the sedimentation rate within the lakes. Still, these

processes that would influence the accumulation rates cannot explain the large fluctuations in Pb isotope ratios observed in the deeper layers of lakes G and SS16, and supply from outside the lake must be assumed. In a study of Hg fluxes in northern Alaskan lakes, Fitzgerald et al. (9) suggest that local soil erosion contributes a substantial portion (15-65%) of the Hg load in the lakes, despite permafrost and small catchment/lake area ratios (similar to our Greenland lakes). They used the Mg content in the catchment soils and in the sediments as a normalizer for the soil-derived Hg in the sediments. Although we did not measure the soil Mg content, no correlation between Mg content and Hg concentration was observed in the sediment from lakes G and SS16; r ) 0.3 (n ) 27) and r ) 0.01 (n ) 20), respectively. Subtracting a Mg-related fraction, using the deepest sample in lake G and SS16 respectively, also does not remove the observed variations in Hg. In contrast to the Alaskan lakes, fluvial erosion in our study catchments is limited due to a lack of inlet streams and restricted overland flow. For Pb and Hg there are relatively important natural sources, such as volcanism, hot spring activity, and fracturing and faulting (3, 39). Although ash-layers from volcanic eruptions are identified in the Greenland ice (40), volcanic activity can be discounted as an explanatory variable for the fluctuations in Hg and Pb concentrations. The 206Pb/207Pb ratios measured in volcanic rocks are low, as observed in Icelandic lava (1.16 to 1.24) (41) and Sicilian lava (1.19 to 1.28) (42), and should not give rise to a prolonged increase in 206Pb/207Pb ratios, such as that observed between 1000 BC and AD 500 in lake SS16 (Figure 3a). For Hg, volcanic eruptions are suggested as the reason for early accumulation fluctuations in peat from the Swiss Jura Mountain (43). However, the few volcanic signals in the Greenland ice during the investigated time period (44) indicate that volcanic eruptions affecting Greenland were distinct events. The amount of Hg emitted from a single eruption and the time resolution of our sediments are too low to result in the observed fluctuations. Instead of long-range sources for these early Hg and Pb fluctuations, we suggest that regional aeolian activity has been an important process for our study sites. Extensive aeolian activity exists today in several valleys (e.g., Sandflugtdalen “sand-dune valley”) in the Kangerlussuaq area proximal to the ice sheet, but also past activity is well described for this region (17, 45). Aeolian transport deflates material from the sandy outwash plains at the ice sheet margin and deposits it on both the lakes and their catchments. This glacial material may have been transported within the ice sheet over long distances, and can therefore have a totally different mineral and chemical composition compared to local lake catchment geology. This material has an important influence on the Hg and Pb concentrations and the Pb isotope composition in the lake sediment, especially in times before the deposition from anthropogenic emissions greatly increased during the last hundred years. Mostly the enhanced modern Hg and Pb concentrations and Pb isotope composition are superimposed over and conceal the signal from aeolian material, but in lake G the modern Hg concentration is within the Hg fluctuations from the aeolian activity. The lake input of aeolian material is supported by the composition in present-day aeolian dust measured in the exposed minerogenic soil, which has a high 206Pb/207Pb ratio (1.43) and a low Hg concentration (7 ng g-1), and by Hg and Pb concentration and Pb isotope ratio in other surface soil samples from the same area (unpublished data). Local aeolian activity rarely affects the Greenland ice accumulation at the summit of the ice sheet (3200 m asl), where anthropogenic Pb is detected in ice layers from AD 0 to 500 (13, 14), because the local prevailing wind direction from the ice over the adjacent tundra would limit aeolian transport eastward and VOL. 40, NO. 1, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

93

upward to the ice summit. Major dust storms in the area today are associated with cold katabatic winds flowing off the ice sheet, which are funneled along the main west-east oriented valleys. Moreover, the altitude at the summit of the ice sheet means it is more exposed to long-range atmospheric transported pollution than the investigated lakes situated at lower altitudes. Soil dust input is also suggested to be the reason for fluctuations in Hg and Pb concentrations during the lateglacial and Holocene period in an ice core from Antarctica (12), and during the Holocene for peat bog cores from the Swiss Jura Mountains (46). In these investigations, a colder and drier climate was suggested to increase soil dust input. This is likely the case also for the Greenland lake sediments because the period with increased 206Pb/207Pb ratios about 1000 BC to AD 500 in lake SS16 and G coincides with lower temperatures in climate reconstructions from the Greenland ice cores (GRIP and Dye 3; 11) and lake sediments from Greenland’s south and east coast (20, 22). For Greenland, this period is also suggested to have been more arid (20, 23) and pollen records show contemporary reductions in plant cover (20). Even though Willemse et al. (45) suggest that during this colder and drier period the silt influx to lakes was low, it is highly probable that these factors would have resulted in enhanced aeolian activity and input of aeolian material to the lakes. Lower 206Pb/207Pb ratios are observed in all the investigated lakes contemporary with the “Medieval Warm Period” (10), which in Greenland was about AD 1000. Higher Pb isotope ratios are again observed in lakes B and G and to some extent in lakes SS16 during the “Little Ice Age”, a low-temperature period AD 1350 to 1700 (10, 44). In addition, in lake B a lower ratio is observed around a period that could agree with the interstadial warmer period between AD 1500 and 1550 (44). In this study we show that variations in input from regional aeolian activity results in Pb and Hg fluctuations in lake sediment records from the Greenland west coast. For Hg the fluctuations are substantial. The aeolian input has been more pronounced in periods with lower temperatures. Although anthropogenic Hg and Pb contributions to the investigated area are unambiguous both during the 19th and especially the 20th centuries, the natural fluctuations are as high as the recent anthropogenic-driven changes in one of the two lakes where a several thousand year long sediment record was studied. One implication of this is that a limited number of samples from pre-industrial sediments may not fully characterize natural conditions and thus interpretations of recent influences may be flawed, especially in relatively lowcontaminated areas such as the Arctic.

Acknowledgments This research was funded by grants from the Centre for Environmental Research (CMF) in Umeå, the Swedish Natural Science Research Council (NFR/VR), and the Danish Natural Science Research Council (SNF). Thanks also to the four anonymous reviewers who have greatly improved this manuscript, to Peter Appleby and Thorbjørn Joest Andersen for radiometric analyses, Birgitta Olsson for the Pb analyses, and Leif Olsson and Johan Rydberg for carbon analyses, and to all other colleagues who have contributed.

Literature Cited (1) AMAP. Assessment Report: Arctic Pollution 2002. Arctic Monitoring and Assessment Programme: Oslo, Norway, 2002. (http:// amap.no). (2) Sturges, W. T.; Hopper, J. F.; Barrie, L. A.; Schnell, R. C. Stable lead isotope ratios in Alaskan Arctic aerosols. Atmos. Environ. 1993, 27A, 2865-2871. (3) Schroeder, W. H.; Munthe, J. Atmospheric mercury, an overwiew. Atmos. Environ. 1998, 32, 809-822. 94

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 1, 2006

(4) Houghton, J. T.; Ding, Y.; Griggs, D. J.; Noguer, M.; van der Linden, P. J.; Dai, X.; Maskell, K.; Johnson, C. A. Climate Change 2001: The Scientific Basis; Cambridge University Press: Cambridge, U. K., 2001. (5) Rosman, K. J. R.; Chisholm, W.; Boutron, C. F.; Candelone, J. P.; Go¨rlach, U. Isotopic evidence for the source of lead in Greenland snows since the late 1960s. Nature 1993, 362, 333-335. (6) Boutron, C. F.; Vandal, G. M.; Fitzgerald, W. F.; Ferrari, C. P. A forty year record of mercury in central Greenland snow. Geophys. Res. Lett. 1998, 25, 3315-3318. (7) Riget, F.; Dietz, R.; Johansen, P.; Asmund, G. Lead, cadmium, mercury and selenium in Greenland marine biota and sediments during AMAP phase 1. Sci. Total Environ. 2000, 245, 3-14. (8) Outridge, P. M.; Hermanson, M. H.; Lockhart, W. L. Regional variations in atmospheric deposition and sources of anthropogenic lead in lake sediments across the Canadian Arctic. Geochim. Cosmocim. Acta 2002, 66, 3521-3531. (9) Fitzgerald, W. F.; Engstrom, D. R.; Lamborg, C. H.; Tseng, C. M.; Balcom, P.; Hammerschmidt, C. Modern and historic atmospheric mercury fluxes in northern Alaska: Global sources and Arctic depletion. Environ. Sci. Technol. 2005, 39, 557-567. (10) Meese, D. A.; Gow, A. J.; Grootes, P.; Mayewski, P. A.; Ram, M.; Stuvier, M.; Taylor, K. C.; Waddington, E. D.; Zielinski, G. A. The accumulation record from GISP2 core as an indicator of climate change throughout the Holocene. Science 1994, 266, 16801682. (11) Dahl-Jensen, D.; Mosegaard, K.; Gundestrup, N.; Clow, G. D.; Johnsen, S. J.; Hansen, A. W.; Balling, N. Past temperatures directly from the Greenland ice sheet. Science 1998, 282, 268271. (12) Boutron, C. F.; Candelone, J. P.; Hong, S. Past and recent changes in large-scale tropospheric cycles of lead and other heavy metals as documented in Antarctic and Greenland snow and ice: A review. Geochim. Cosmochim. Acta 1994, 58, 3217-3225. (13) Hong, S.; Candelone, J. P.; Patterson, C. C.; Boutron, C. F. Greenland ice evidence of hemispheric lead pollution two millennia ago by Greek and Roman Civilizations. Science 1994, 265, 1841-1843. (14) Rosman, K. J. R.; Chisholm, W.; Hong, S.; Candelone, J. P.; Boutron, C. F. Lead from Carthaginian and Roman Spanish Mines Isotopically Identified in Greenland Ice dated from 600 BC to 300 AD. Environ. Sci. Technol. 1997, 31, 3413-3416. (15) Shotyk, W.; Goodsite, M. E.; Roos-Barraclough, F.; Frei, R.; Heinemeier, J.; Asmund, G.; Lohse, C.; Hansen, T. S. Anthropogenic contributions to atmospheric Hg, Pb and As accumulation recorded by peat cores from southern Greenland and Denmark dated using the 14C “bomb pulse curve”. Geochim. Cosmocim. Acta 2003, 67, 3991-4011. (16) Bindler, R.; Renberg, I.; Appleby, P. G.; Anderson, N. J.; Rose, N. L. Mercury accumulation rates and spatial patterns in lake sediments from West Greenland: a coast to ice margin transect. Environ. Sci. Technol. 2001, 35, 1736-1741. (17) Anderson, N. J.; Harriman, R.; Ryves, D. B.; Patrick, S. T. Dominant factors controlling variability in the composition of West Greenland lakes. Arctic, Antarct., Alpine Res. 2001, 33, 418425. (18) Bindler, R.; Renberg, I.; Anderson, N. J.; Appleby, P. G.; Emteryd, O.; Boyle, J. Pb isotope ratios of lake sediments in West Greenland: inferences on pollution sources. Atmos. Environ. 2001, 35, 4675-4685. (19) Malmquist, C.; Bindler, R.; Renberg, I.; van Bavel, B.; Karlsson, E.; Anderson, N. J.; Tysklind, M. Time trends of selected persistent organic pollutants in lake sediments from Greenland. Environ. Sci. Technol. 2003, 37, 4319-4324. (20) Wagner, B.; Melles, M.; Hahna, J.; Niessen, F.; Hubberten, H. W. Holocene climate history of Geographical Society Ø, East Greenland- evidence from lake sediments. Palaeogeogr., Palaeoclimatol., Palaeoecol. 2000, 160, 45-68. (21) Anderson, N. J.; Fritz, S. C.; Gibson, C. E.; Hasholt, B.; Leng, M. J. Lake-catchment interactions with climate in the low Arctic of southern West Greenland. Geol. Greenl. Survey Bull. 2002, 191, 144-149. (22) Kaplan, M. R.; Wolfe, A. P.; Miller, G. H. Holocene environmental variability in southern Greenland inferred from lake sediments. Quat. Res. 2002, 58, 149-159. (23) Anderson, N. J.; Leng, M. J. Increased aridity during the early Holocene in West Greenland inferred from stable isotopes in laminated-lake sediments. Quat. Sci. Rev. 2004, 23, 841-849. (24) Willemse, N. Arctic Natural Archives; Netherlands Geographical Studies 272; Utrecht University: The Netherlands, 2000. (25) Box, J. E. Survey of Greenland instrumental temperature records: 1873-2001. Int. J. Climatol. 2002, 22, 1829-1847.

(26) Renberg, I. Improved methods for sampling, photographing and varve counting of varved lake sediments. Boreas 1981, 10, 255-258. (27) Farmer, J. G.; Eades, L. J.; Atkins, H.; Chamberlain D. F. Historical trends in the lead isotopic composition of archival Sphagnum mosses from Scotland (1838-2000). Environ. Sci. Technol. 2002, 36, 152-157. (28) Appleby, P. G.; Nolan, P. J.; Gifford, D. W.; Godfrey, M. J.; Oldfield, F.; Anderson, N. J.; Battarbee, R. W. 210Pb dating by low background gamma counting. Hydrobiologia 1986, 141, 2127. (29) Appleby, P. G.; Richardson, N.; Nolan, P. J. Self-absorption corrections for well-type germanium detectors. Nucl. Instrum. Methods Bull. 1992, 71, 228-233. (30) Stuiver, M.; Reimer, P. J.; Braziunas, T. F. High-precision radiocarbon age calibration for terrestrial and marine samples. Radiocarbon 1998, 40, 1041-1083. (31) Pirrone, N.; Keeler, G. J.; Nriagu, J. O. Regional differences in the worldwide emissions of mercury to the atmosphere. Atmos. Environ. 1996, 30, 2981-2987. (32) Pacyna, E. G.; Pacyna, J. M. Global emission of mercury from anthropogenic sources in 1995. Water, Air, Soil Pollut. 2002, 137, 149-165. (33) Bra¨nnvall, M. L.; Bindler, R.; Emteryd, O.; Renberg, I. Four thousand years of atmospheric lead pollution in northern Europe: a summary from Swedish lake sediments. J. Paleolimnol. 2001, 25, 421-435. (34) Martinez-Cortizas, A.; Pontevedra-Pombal, X.; Garcı´a-Rodeja, E.; No´voa-Mun ˜ oz, C.; Shotyk, W. Mercury in a Spanish peat bog: Archive of climate change and atmospheric metal deposition. Science 1999, 284, 939-942. (35) Bindler, R.; Olofsson, C.; Renberg, I.; Frech, W. Temporal trends in mercury accumulation in lake sediments in Sweden. Water, Air, Soil Pollut. 2001, 1, 343-355. (36) Bindler, R. Estimating the natural background atmospheric deposition rate of mercury utilizing ombrotrophic bogs in southern Sweden. Environ. Sci. Technol. 2003, 37, 40-46. (37) Roos-Barraclough, F.; Shotyk, W. Millennial-scale records of atmospheric mercury deposition obtained from ombrotrophic

(38) (39) (40) (41)

(42)

(43)

(44) (45) (46)

and minerotrophic peat lands in the Swiss Jura Mountains. Environ. Sci. Technol. 2003, 37, 235-244. Renberg, I.; Wik Persson, M.; Emteryd, O. Pre-industrial atmospheric lead contamination detected in Swedish lake sediments. Nature 1994, 368, 323-326. Nriagu, J. A global assessment of natural sources of atmospheric trace metals. Nature 1989, 338, 47-49. Ram, M.; Gayley, R. I. Long-range transport of volcanic ash to the Greenland ice sheet. Nature 1991, 349, 401-404. Thirlwall, M. F.; Gee, M. A. M.; Taylor, R. N.; Murton, B. J. Mantle components in Iceland and adjacent ridges investigated using double-spike Pb isotope ratios. Geochim. Cosmocim. Acta 2004, 68, 361-386. Monna, F.; Aiuppa, A.; Varrica, D.; Dongarra, D. Pb isotope composition in lichens and aerosols from eastern Sicily: Insights into the regional impact of volcanoes on the environment. Environ. Sci. Technol. 1999, 33, 2517-2523. Roos-Barraclough, F.; Martinez-Cortizas, A.; Garcı´a-Rodeja, E.; Shotyk, W. A 14500 year record of the accumulation of atmospheric mercury in peat: volcanic signals, anthropogenic influences and a correlation to bromine accumulation. Earth Planet. Sci. Lett. 2002, 6334, 1-18. Hammer, C. U.; Clausen, H. B.; Dansgaard, W. Greenland ice sheet evidence of post-glacial volcanism and its climate impact. Nature 1980, 288, 230-235. Willemse, N. W.; Koster, E. A.; Hoogakker, B.; Tatenhove, F. G. M. A continuous record of Holocene eolian activity in West Greenland. Quat. Res. 2003, 59, 322-334. Shotyk, W.; Krachler, M.; Martinez-Cortizas, A.; Cheburkin, A. K.; Emons, H. A peat bog record of natural, pre-anthropogenic enrichments of trace elements in atmospheric aerosols since 12 370 14C yr BP, and their variation with Holocene climate change. Earth Planet. Sci. Lett. 2002, 199, 21-37.

Received for review June 27, 2005. Revised manuscript received October 21, 2005. Accepted November 3, 2005. ES051223Y

VOL. 40, NO. 1, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

95