METAL CONTAMINATION OF THE NATURAL ENVIRONMENT IN NORWAY FROM LONG RANGE ATMOSPHERIC TRANSPORT

Eiliv Steinnes (Department of Chemistry, Norwegian University of Science and Technology, N-7491 Trondheim, Norway). Eiliv.Steinnes@chembio.ntnu.no

ABSTRACT

For many heavy metals long range atmospheric transport is the most important source of contamination to the natural environment in Norway. Investigations based on aerosol studies, bulk deposition measurements, and moss analysis show that airborne transport from other parts of Europe is the major mode of supply for V, Zn, As, Se, Mo, Cd, Sn, Sb, Tl, Pb, and Bi. Elements associated with long range transport show substantial enrichment in the humus horizon of natural soils in southern Norway, in some cases to a factor of 10 or more relative to pristine areas in the north. Elements such as Pb and Cd also show enrichment in some terrestrial food chains in the south. These elements also show considerably elevated levels over background concentrations in water and sediments of small lakes in the southern part of the country. The Norwegian case is an excellent illustration of the significance of long range atmospheric transport as a source of heavy metal contamination.

INTRODUCTION

Exposure of southern Norway from long range atmospheric transport of pollutants (LRTP) has been evident for a long period of time. The issue was first strongly focused during the interdisciplinary research program "Acid precipitation - Effects on Forest and Fish", carried out during the period 1972-1980 and mainly dealing with impacts of acidifying sulfur and nitrogen compounds. Independent investigations carried out during the same period demonstrated that the problem was also evident for micropollutants such as heavy metals (Allen and Steinnes 1979, Steinnes 1980). Over the last twenty years it has become increasingly evident from many scientific investigations that terrestrial as well as aquatic ecosystems in the southern part of Norway are to a great extent contaminated with a number of heavy metals derived from source regions in other parts of Europe. In the present paper the main results from these studies are briefly reviewed.

EXPERIMENTAL

In the following the main types of samples forming basis for these studies of heavy metals and the analytical techniques employed are listed. For analyses requiring wet decomposition of the samples concentrated nitric acid was used in most cases.

Air concentrations were investigated by collecting aerosols on filters (Amundsen et al. 1992) and analyzing them by instumental neutron activation analysis (INAA) and for some elements atomic absorption spectrometry (AAS).

Bulk atmospheric deposition was studied partly by precipitation sampling on a limited number of stations (Berg et al. 1994) followed by analysis by inductively-coupled plasma mass spectrometry (ICPMS), or by analysis of terrestrial mosses collected at about 500 different sites all over the country followed by analysis using either a combination of INAA and AAS (Steinnes 1980, Steinnes et al. 1992) or more recently ICPMS (Berg et al. 1995). For most elements associated with LRTP the concentrations in moss can be converted to bulk deposition rates with a reasonable degree of accuracy (Berg and Steinnes 1997).

Natural surface soils were collected from the humus layer (normally corresponding to the F horizon) by the bulking technique (Allen and Steinnes 1979, Steinnes et al. 1989, 1997) and analyzed by AAS or radiochemical NAA. In one nationwide survey mineral soil samples from the B and C horizons were sampled at each site in addition to those from the humus layer. In this case samples were decomposed with 7M nitric acid and analyzed by inductively-coupled plasma emission spectrometry (ICPES).

Peat samples were collected along depth profiles down to 50 cm in ombrotrophic bogs and analyzed by AAS or radiochemical NAA (Hvatum et al. 1983).

Plant tissues were collected from commonly occurring species in natural ecosystems. Leaves and twigs were separated. Samples were dried, homogenized, and analyzed by AAS.

Animal tissues, in most cases liver or kidney, were obtained from a number of mammal and bird species living in terrestrial ecosystems (Frøslie et al. 1984, 1985, Kålås et al. 2000). Samples were analyzed by AAS, in some cases also by radiochemical NAA.

Surface water was sampled from 985 small lakes all over the country using clean techniqued. Samples were served with acid in the field and were not filtered (Skjelkvåle et al. 1999)

Lake sediments : The upper 1 cm of sediment was sampled in 132 lakes in southern Norway (Fjeld et al. 1994).

RESULTS AND DISCUSSION

Source regions: Diurnal aerosol filter samples were collected at Birkenes, southern Norway during 1985-86 and the air trajectories for each day were used to classify the samples with respect to emission areas (Amundsen et al. 1992). For metals typically associated with episodes of polluted air from other parts of Europe (V, As, Se, Cd, Pb) the air concentrations were typically tenfold higher or more when the air came from the sectors East (Russia), South-east (Poland, DDR), South (West Germany) or South-west (UK, Benelux, France) than with air from North-west (North Antlantic) or North (mainland Norway). These findings confirmed similar conclusions reached independently for S and N compounds.

Geographical deposition patterns: In the first nationwide deposition survey in 1977 based on moss samples, comprising 26 elements (Steinnes 1980, Steinnes et al. 1992), it was shown that the fallout of V, Zn, As, Cd, Sb, and Pb was substantially higher, in most cases 10 times or more, in the southernmost part of the country than in many areas at more northerly latitudes. This strongly indicates LRTP as the dominating source of these elements in the south and is in good agreement with the above results from sector analysis of aerosols. In a factor analysis of the moss data (Schaug et al. 1990) these elements came out with a high loading in the first factor, explaining 20% of the variance in the data set. It was also clear that for other heavy metals such as Cr, Co, Ni, and Cu the deposition patterns were dominated by local point sources, situated within Norway or in north-western Russia close to the Norwegian border. These trends have been confirmed in later moss surveys (Berg et al. 1995) although the absolute deposition rates have been substantially reduced over the years for many of the LRTP-associated elements. In the more recent surveys based on ICPMS it has become evident that further elements (Mo, Sn, Ti, Bi) can be added to the "LRTP family".

Analyses of bulk precipitation samples from stations located in different parts of Norway have confirmed the geographical trends shown by moss analysis. In an extensive survey carried out in 1989-1990 at six stations (Berg et al. 1994) multivariate analysis of the data collected at Birkenes defined an LRTP associated component explaining 42% of the variation.

Soil pollution: Most of Norway is uncultivated land with podzols or similar soil types where the surface horizon maimly consists of organic matter. These humus-rich topsoils will act as a temporary sink for many metals, and in the case of Pb a residence time of the order of centuries has been indicated. In 1977 samples from the humus layer were collected all over Norway according to a network similar as for the moss surveys (Allen and Steinnes 1979, Steinnes et al. 1997) and analyzed for Cu, Zn, As, Se, Cd, Sb, and Pb. The geographical distributions were strikingly similar as those for atmospheric deposition of these elements. Except for Cu, all metals were consistently higher in the topsoil in southernmost Norway than in more northerly regions. A more detailed survey in the sothernmost four counties (Steinnes et al. 1989) revealed that the elements typical of LRTP showed maximum levels in a zone located 20-50 km from the coastline, corresponding quite closely to the highest deposition of orographic precipitation. In this zone Pb concentrations in the humus layer were around 180 ppm, as opposite to background values of about 10 ppm in areas of central and northern Norway.

High concentrations of a metal in the topsoil does not necessary imply atmospheric input. The natural content of a metal in surface soils depends on its concentration in the underlying soil, which is in most cases related to the geochemistry of the bedrock in the area. In order to find out to what extent georgaphical differences observed in the 1977 soil material might be associated with regional differences in bedrock composition, a second nationwide campaign was organized in 1985, collecting samples from the B and C (60 cm depth) horizons in addition to the humus layer. The results obtained (Njåstad et al. 1994) indicated that no geographical pattern corresponding to that observed in the humus layer exists in the underground soils for Pb, Cd, and Zn, confirming an LRTP origin of the excessive levels of these and possibly other metals in the southern topsoils. Further work indicates that these metals are being removed rather slowly from the southern topsoils (Berthelsen et al. 1995a) and that mycorrhizal fungi may be partly responsible for the slow removal e.g. of Zn and Cd (Berthelsen et al. 1995b).

Terrestrial food chains: From a study in 1982 of several higher plants growing at different sites in south and central Norway respectively it became evident that the Pb and Cd concentrations were consistently about 5 times higher in the south, and that of the essential plant nutrient Zn about twice as high. When sampling was carried out at exactly the same sites in 1992 it appeared that Pb had decreased substantially in both regions whereas Zn and Cd showed the same levels as before (Berthelsen et al. 1995a). Apparently the Pb content of vascular plants was due to surface contamination and hence decreasing parallel to the declining atmospheric deposition of lead, whereas Zn and Cd were mainly supplied by root uptake from the surface soil still contaminated from LRTP in the south.

A study of metals in lamb livers from 10 livestocks in different parts of Norway (Frøslie et al. 1985) revealed an almost perfect match between the Pb contentration in liver and atmospheric deposition of Pb (R=0.935). In another study by these authors dealing with wild rumnants (Frøslie et al. 1994) rather clear south-north gradients were observed in moose liver (Pb, As) and kidney (Cd). Similarly higher concentrations of Pb were found in the southern part of Norway than farther north in livers from species such as hare and black grouse (Kålås et al. 2000). Studies on other animal species have failed to show similar connections, indicating that the feeding habits may play an important role in this respect.

Small lakes: Results from a recent survey of small lakes in Norway indicate a clear influence of LRTP on the geographical distribution of a number of metals in surface water (Skjelkvåle et al. 1999). This may partly be due to contribution from current and previous deposition of these metals (Zn, Cd, Pb), partly to release of the metals due to effects of acidification associated with LRTP (Be, Co). Lake sediments also carry distinct information about supply of metals to Norwegian lakes from LRTP (Fjeld et al. 1994). Results from a study of sediments from 132 lakes showed considerable enrichment of Pb, Hg, and Cd in the near-surface sediment, clearly indicating a connection between these elements and LRTP.

Retrospective studies: In ombrotrophic bogs the peat layers down to a certain depth have no contact with the ground water and hence is supplied with chemical substances only from the atmosphere. Studies of cores from such bogs (Hvatum et al. 1983) show that the deposition of elements such as Zn, As, Cd, Sb, and Pb in southern Norway during pre-industrial times probably was no higher than 1-2% of that observed in recent decades. In the surface peat the geographical patterns of these elements were generally similar to their atmospheric deposition patterns.

Overall conclusion: The results of the extensive studies on heavy metals briefly reviewed in this paper show a high degree of consistency and clearly demonstrate a strong influence of LRTP. The Norwegian case is an excellent illustration of the significance of long range atmospheric transport as a source of heavy metal contamination.

REFERENCES

Allen RO, Steinnes E (1979), In: Proc. Intl. Conf. Heavy Metals in the Environ. Edinburgh, CEP Consultants, pp. 271-274.

Amundsen CE, Hanssen JE, Semb A and Steinnes E (1992), Atmos. Environ. 26A: 1309-1324.

Berg T, Steinnes E (1997), Environ. Pollut. 98: 61-71.

Berg, T, Røyset, O, Steinnes, E (1994), Atmos. Environ. 28: 3519-3536.

Berg, T, Røyset, O, Steinnes, E, Vadset M (1995), Environ. Pollut. 88: 67-77.

Berthelsen BO, Steinnes E, Solberg W, Jingsen L (1995a), J. Environ. Qual. 24: 1018-1026.

Berthelsen BO, Olsen R, Steinnes E (1995b), Sci. Total Environ. 170: 141-149.

Fjeld E, Rognerud S, Steinnes E (1994), Can. J. Fish. Aquat. Sci. 51: 1708-1720.

Frøslie A, Norheim G, Rambæk JP, Steinnes E (1984), Acta Vet. Scand. 25: 333-345.

Frøslie A, Norheim G, Rambæk JP, Steinnes E (1985), Bull. Environ. Contam. Toxicol. 34: 175-182.

Hvatum OØ, Bølviken B, Steinnes E (1983), Ecol. Bull. (Stockholm) 35: 351-356.

Kålås JA, Steinnes E, Lierhagen S (2000), Environ. Pollut. 107: 21-29.

Njåstad O, Steinnes E, Bølviken B, Ødegård M (1994), National Survey of Element Composition in Natural Soil. Trondheim, Report NGU 94.327, Geological Survey of Norway (in Norwegian).

Skjelkvåle BL et al. (1999), Heavy Metal Surveys in Nordic Lakes; Harmonized Data for Regional Assessment of Critical Limits. Oslo, Report SNO 4039-99, Norwegian Institute of Water Research.

Steinnes E (1980), J. Radioanal. Chem. 58: 387-391.

Steinnes E, Solberg W, Petersen HM, Wren, CD (1989), Water, Air, Soil Pollut. 45: 207-218.

Steinnes E, Allen RO, Petersen HM, Rambæk JP, Varskog P (1997), Sci. Total Environ. 205: 255-266.