Mercury and Other Trace Metals

Using the ICP-MS technique, Ikegawa et al. (1997) determined the concentrations of 36 elements in drifting snow samples collected from July to December 1991 at Asuka Station (71° 32'S, 24° 08'E). They found peak concentrations of most elements and nssSO42- in late September to early October, and hypothesised that volcanic eruptions on Mt. Pinatubo (June 1991) and Mt. Hudson (August 1991) may have been responsible for enhanced Pb, Cd, Cu, Zn, Se and nssSO42 concentrations in the 1991 Antarctic spring. The analysis of enrichment factors suggested that Na, Mg, Ca, K and Sr were of marine origin, while Al, Fe, Mn, Rb, Cr, Ni, Ga, V and all rare earth elements were of crustal origin. A subsequent study (Ikegawa et al. 1999) on 37 element concentrations in surface snow samples collected between 1991 and 1993 during an over-snow traverse in East Queen Maud Land showed an increased fallout flux for elements such as Co, Ni and Cd at altitudes above 2,500-3,000 m a.s.l. It was suggested that the distribution of elements could reflect polar stratospheric precipitation or long-range tropospheric transport from the Southern Hemisphere. With the aim of providing an indicative picture of the elemental composition of recent Antarctic snow, Table 3 summarises data on concentrations of rare earth elements (Ikegawa et al. 1997) and of many other elements (Barbante et al. 1999; Planchon et al. 2002a). Average values measured by Veysseyre et al. (2001) in fresh snow collected at different altitudes in the French Alps are also

Table 3. Elemental composition (pg g-1) of recent snow from Antarctica and the French Alps

Element

Antarctica 1991-1993 (Ikegawa et al. 1999)

Antarctica 1989-1990

(Barbante et al. 1999; Planchon et al. 2002a)

French Alps 1999 (Veysseyre et al. 2001)

Ag

0.14

0.88±0.28

Al

130,000

29,112±28,000

Au

0.76±0.04

B

498±116

Ba

1.2

232±215

Bi

0.052

6.0±6.5

Cd

0.14

26±18

Ce

1.6±0.8

Co

4.4±2.4

Cr

1.6

Cu

2.4

79±86

Dy

0.08±0.007

Er

0.041±0.035

Eu

0.021±0.002

Fe

4,879±4,512

Gd

0.09±0.07

La

0.86±0.45

Li

229±10

Mn

4.5

413±442

Mo

19±7

Nd

0.69±0.034

Pb

1.6

472±579

Pd

0.54±0.12a

2.6±0.9

Pr

0.23±0.021

Pt

0.37±0.18a

0.50±0.04

Rh

0.04±0.01a

Sb

34±27

Sm

0.11±0.008

Sn

215±32

Th

0.68±0.063

Ti

71±87

U

0.016

5.1±7.7

V

0.4

36±35

Yb

0.097±0.190

Zn

0.8

340±247

a Data from Barbante et al. (1999)

a Data from Barbante et al. (1999)

reported to point out the different magnitude of element concentrations in remote regions and in ones impacted by human activities. These data show that only Pt, Pd and Ag concentrations in Antarctic and European snow are in the same range. Indications of atmospheric contamination by Ag and other elements such as Cr, Zn, Bi and U (in addition to the above-reported Pb and Cu contamination) were detected by Planchon et al. (2002a) in Coats Land snow deposited during the 1970s-1980s. Ag, Cr, Pb and U concentrations began to decrease at the end of the 1980s. It was suggested that natural sources, such as multiple eruptions in the Deception Island volcano, and anthropogenic activities in the Southern Hemisphere, such as increased U mining and milling operations, or mining and smelting of non-ferrous metals, contributed to the temporary increase in metal deposition. Although production of other elements such as Mn and V also increased in the same period, their enrichment factors did not significantly increase in Antarctic snow. These elements, like lithophilic Al, Ba, Fe and Ti, are rather abundant in airborne rock and soil dust and, as in the case of Cd, their natural inputs are probably so large as to mask anthropogenic ones.

As discussed in the previous chapter, the atmospheric transport of Hg and its rapid oxidation and deposition in polar regions after spring sunrise is a critical contamination issue. Nevertheless, there are only few investigations on the dynamics of Hg in polar snow, and it is still unknown whether polar regions are important sinks in the global cycle of Hg.Vandal et al. (1993) measured Hg concentrations in 14 sections of an ice core from Dome C (77° 39'S, 124° 10'E, 3,240 m a.s.l.), covering the past 40,000 years. They found that the Hg content was less than 1 pg g-134,000 years b.p.; it increased to 2.1 pg g-1 and remained elevated during the last glacial maximum (from 28,000 to 18,000 years b.p.), decreased during the transition from the last ice age to the Holocene (17,000-13,000 years b.p.), and reached concentrations<0.5 pg g-1 during the Holocene. The emission of gaseous Hg from productive ocean regions was considered to be the main source of metal to Antarctic snow, and the threefold increase in Hg deposition during the last glacial maximum was mainly attributed to climatic conditions and enhanced primary productivity in the Southern Ocean. There is evidence (Petit et al. 1981) that aridity, higher winds and stronger poleward atmospheric transport enhanced the accumulation of aerosols in Antarctic ice. An increased input of Fe and other essential lithophilic elements in the Southern Ocean probably enhanced marine productivity in the Southern Ocean; although mechanisms are poorly known, there is evidence that marine organisms can promote the reduction of reactive Hg species to volatile Hg° (Mason et al. 1994). Research by Heumann (2001) shows that pure cultures of psychrotrophic bacteria (i.e. having a minimum growth temperature in the range 0-5 °C) from the Southern Ocean can methylate inorganic Hg2+, further contributing to the transfer of this metal from the ocean to the atmosphere. Considering the rapid equilibrium between surface oceans and the atmosphere, and the scarce sedimentation of

Hg in oceans, it seems probable that Antarctica and the other continents are the main sink for atmospheric Hg released by oceans. In samples of surface snow collected at the South Pole and other sites at different distances from the sea,Vandal et al. (1993) detected concentrations in the 0.13-0.50 pg g-1 range, with lower values occurring in samples from inland sites. Sheppard et al. (1991) reported values in the 0.1-1.5 pg g-1 range and no temporal trend in firn samples from Southern Victoria Land.

5.3.5 Persistent Organic Contaminants

The first two papers reporting the presence of DDT compounds in Antarctica were published in 1966. George and Frear (1966) failed to detect DDTs in snow and water samples collected around McMurdo Station, but they found the chemicals in Antarctic organisms at higher levels of the marine food web, such as marine mammals and seabirds. In the same year, Staden et al. (1966) reported the presence of DDT residues in Adelie penguins and a crabeater seal. The first record of DDT in Antarctic snow was published by Peterle (1969), although a subsequent review of absolute amounts and ratios of DDT compounds revealed that, due to the probable contamination of samples, the reported values were too high (Peel 1975; Risebrough 1977). However, the analysis of snow samples collected in austral summer 1975 from a pit on a permanent snowfield at Doumer Island (64° 51'S, 63° 35'W) confirmed the presence of DDT residues and PCBs (Risebrough et al. 1976). Average concentrations in surface snow of pp'-DDT (about 0.5 pg g-1), pp'-DDE (about 0.1 pg g-1) and total PCBs (about 0.15 pg g-1) increased in samples collected at 2-4 m depth (up to 4,0.27 and 1.2 pg g-1 respectively), and then decreased (at 5.5-6 m depth, the measured values were 2.1, 0.21 and 0.28 pg g-1 respectively). The distribution pattern of pp'-DDT and pp'-DDE was similar, and the quantitative ratio between the two compounds was approximately 10 along the profile. Although the samples were not dated, these results seem to indicate that enhanced deposition of DDTs and PCBs probably occurred in the decade before 1975. In the Canadian Arctic region (Agassiz Ice Cap, Ellesmere Island; Gregor et al. 1995), for instance, small amounts of PCBs were first evident in ice from 1957 to 1963, and deposition increased thereafter.

Besides the analysis of surface snow, Risebrough et al. (1976) also determined DDT and PCB concentrations in penguin eggs from the Antarctic Peninsula, and in the blubber of a leopard seal killed by a killer whale; they came to the conclusion that the atmosphere is the main pathway for DDT and PCB transport to Antarctica. However, owing to the presence of the small Chilean Station of Yelcho on Doumer Island, and given the frequent burning in the past of waste materials at the nearby and larger Palmer Station (on Anvers Island, 64° 46'S, 64° 03'W), it was supposed that a portion of PCB residues also derived from local sources.

Tanabe et al. (1983a) analysed PCBs and chlorinated hydrocarbons such as DDTs and HCHs in snow samples collected close to Syowa Station. In contrast to Risebrough and co-workers, they found no significant variations in total DDTs, HCHs or PCBs between samples of surface and deep snow. This study suggests that the deposition rate of pesticides (never used in Antarctica) and industrial compounds such as PCBs (these have been used in the region) has remained the same since the 1960s.

More recently, Fuoco and Ceccarini (2001) have determined concentrations of selected PCB congeners in samples of recent snow and firn, collected during the austral summer 1993-1994 in seven snowfields throughout northern Victoria Land. The pattern of PCB isomers in snow (Fig. 41) showed a predominance of lower chlorinated congeners and, although in this study PCB-18 was not considered, the pattern in Fig. 41 roughly reflects that reported for atmospheric PCBs at Signy Island and King George Island (Fig. 36). The numerous studies performed on PCB congener patterns in Arctic air and snow show that their distribution is very similar, with lighter tri-chlorinated PCB homologues largely dominating the atmosphere and snow of northern sites (Macdonald et al. 2000). The sampling sites of surface snow throughout northern Victoria Land were located at different altitudes (from sea level to 3,000 m a.s.l.) and at varying distances from the sea, but total PCB concentrations measured by Fuoco and Ceccarini (2001) showed no significant spatial variations (range 0.28-0.73 pg g-1; mean=0.52 pg g-1). Samples from a 2.5-m-deep pit at the Hercules Névé collected in summer 1993-1994 and 1994-1995 showed higher total PCB concentrations (about 1 pg g-1) in the deepest samples (presumably deposited in 1987) than in surface snow (about 0.65 pg g-1). This result seems to corroborate previous findings by Risebrough et al. (1976), and agrees with the general decreasing trend in POP concentrations in the atmosphere of Antarctica and the sub-Antarctic islands during the 1980s and 1990s (Fig. 35).

Fig. 41. Mean concentrations (pg l-1) of PCB isomers in samples of snow and firn from seven snow-fields in northern Victoria Land. (Data from Fuoco and Ceccarini 2001)

Fig. 41. Mean concentrations (pg l-1) of PCB isomers in samples of snow and firn from seven snow-fields in northern Victoria Land. (Data from Fuoco and Ceccarini 2001)

Estimates of spatio-temporal variations in the rate of POP atmospheric deposition based on snow and firn data may contain large uncertainties due to meteorological conditions - which make it very difficult to select sampling sites representative of local atmospheric deposition, especially in terrains with blowing and drifting snow - and to the possible post-depositional release of some contaminants. As deeper snow retains its burden of more volatile POPs such as HCH better than shallow snow, sampling sites in gullies or hummocks can produce significantly different results. Moreover, under the same conditions of atmospheric contamination, the scavenging of atmospheric POPs may be enhanced by lower temperatures. For instance, higher average concentrations of PCBs and DDTs have been measured in Arctic snowfall than in Lake Superior (Macdonald et al. 2000). Temperature variations may also affect the post-depositional loss of more volatile compounds, and this process may be accentuated under warmer climatic conditions.

Despite the difficulty in making reliable comparisons between POP concentrations in snow from regions with different climatic and environmental conditions, Table 4 reports the results of research performed on snow from Antarctica and the Arctic (Macdonald et al. 2000; Melnikov et al. 2003), Alps and Pyrenees (Carrera et al. 2001). Comparisons among the data in Table 4 are complicated not only by the different sampling years and very different climatic and environmental conditions, but also by the different approaches adopted for calculating the total amount of PCBs and HCHs (e.g. sum of dif-

Table 4. Ranges of POP concentrations (pg g-1) in snow from Antarctica, the Arctic, Alps and Pyrenees

Region

Period

ÂDDT

2hch

ÂPCB

Reference

Antarctica

Doumer Island

1975

0.5-4.3

-

0.03-1.2

Risebrough et al. (1976)

Syowa Station

1982

0.16-1.0a

1.5-4.9a

0.16-1.0a

Tanabe et al. (1983a)

Northern Victoria Land

1994

0.28-0.73 Fuoco and Ceccarini (2001)

Arctic

Canada

19911992

<0.01-0.42

0.02-3.7

0.9-13

Macdonald et al. (2000)

Ob-Yenisey River watershed

19921993

0.5-0.7

1.2-3.4

0.4-0.6

Melnikov et al. (2003)

Europe

Alps and Pyrenees

19971998

n.d.-330

0.49-1.1

0.22-2.2

Carrera et al. (2001)

a Concentrations expressed as pg dm-3; n.d., not detectable a Concentrations expressed as pg dm-3; n.d., not detectable ferent number of congeners; quantified as Aroclor mixtures; sum of a-HCH and g-HCH). However, in the first half of the 1990s, average PCB concentrations in Victoria Land snow (about 0.5 pg g-1) corresponded to the mean value measured by Melnikov et al. (2003) in snow from the Ob-Yenisey River watershed, and were about 4 times lower than the average 4.1 pg g-1 reported by Macdonald et al. (2000) for Canadian Arctic snow.

Was this article helpful?

0 0

Post a comment