Effects of acidification on the concentrations of heavy metals in running waters in Sweden

Heavy metals in running water are analysed within the Swedish Surface Water Monitoring Programme at about 80 stations. Data for selected rivers and brooks has been used to assess the effect of soil acidification on the concentrations of metals in waters. In southern Sweden, acidification shows a significant effect on Zn and Cd. In the upper parts of the drainage areas, there is an marked increased leakage of these metals from soils to waters indicated by elevated concentrations of Zn and Cd in brooks and also by high maxima during periods of low pH values. The increased leaching from forest soils also influences the concentrations in rivers in southern Sweden. During periods of high water flow, the pH values of the rivers decreases and the concentrations of Zn and Cd increases. At pH levels of 6.2–6.5 the concentrations of these metals are mostly 2–5 times higher as compared to pH levels of about 7.0. Since high concentrations of Zn and Cd coincides with high water flow, the transport of these metals has most certainly increased several times compared to the pre-industrial period. Hence, the by far most important human impact on the loading of Zn and Cd on the marine areas, surrounding southern Sweden is the increased leakage of these metals due to acidification. In northern Sweden the effects of acidification on Zn and Cd are less evident. In general, the concentrations of these metals are lower and the connection between pH and these metals are much less significant in the rivers. The links between acidification and the concentrations of Pb and Cu in the watercourses are comparatively much weaker. These two metals are more related to the content of organic matter in the waters and a possible effect of acidification is overshadowed by natural transport processes in soils and waters.     

Critical loads of acidity for surface waters

The critical load of acidity for surface waters is based on the concept that the inputs of acids to a catchment do not exceed the weathering less a given amount of ANC. The Steady State Water Chemistry (SSWC) Method is used to calculate critical loads, using present water chemistry. To ensure no damage to biological indicators such as fish species a value for ANC_limit of 20 μeq/l has been used to date for calculating critical loads. The SSWC-method is sensitive to the choice of the ANC_limit. In areas with little acid deposition the probability of acid episodes leading to fish kills is small even if the ANC_limit is set to zero, while in areas with high acidic deposition fish kills may occur at this value. Thus, the ANC_limit can be a function of the acidifying deposition to the lake, nearing zero at low deposition and increasing to higher values at higher deposition. A formulation for such an ANC_limit has been worked out, and we have tested the effect of the ANC_limit as a linear function of the deposition, assuming ANC_limit = 0 at zero deposition with a linear increase to 50 ueq/l at a deposition of 200 meq.m^?2.yr^?1. For areas with high deposition the effect of a variable ANC_limit is small, while in areas with low deposition the effect is significant. For Norway the exceeded area decreases from 36 to 30% using a variable ANC_limit instead of a fixed value of 20 μeq/l.     

Critical loads of acidity to surface waters

The critical load of acidity to surface water is based on the condition that the inputs of acids to a catchment do not exceed the weathering rate less a given amount of ANC (Acid Neutralizing Capacity). The Steady State Water Chemistry (SSWC) Method is used to calculate critical loads of acidity, using present water chemistry. To calculate the weathering, the so-called F-factor is used to estimate the part of the base cation flux that is due to soil acidification. The F-factor has been estimated empirically from historical data comparisons from Norway, Sweden, U.S.A. and Canada and is considered to be a function of the base cation concentration by the formula: F=sin(BC^*/S), where BC^* is the present base cation concentration and S the base cation concentration at which F=1. At higher values for BC^* F is set to 1. For Norway, Sweden and Finland S has been set to 400 μeq/l (ca. 8 mg Ca/l), giving F-values in the range 0.05–0.2. The importance of the F-factor in the calculations of the critical loads of acidity for Nordic surface waters was tested by calculating the magnitude of the area where the critical load of acidity is exceeded in Norway for different values of S. Similar calculations were carried out for the Finnish and Swedish lake data. Varying S from 100 μeq/l to 1200 μeq/l, the exceeded area in Norway decreases from 31,9 to 28,3%. For F=0 (S=∞, i.e. assuming no soil acidification), the exceeded area is reduced to 27,2%. For Finland and Sweden the the percent of lakes exceeded are reduced from 16,6 to 12,9% and 30 to 23,6%, respectively. For F = 0 the percent of lakes exceeded are reduced to 11,4 and 16,4, repectively. These results indicate that the F-factor is not of great importance for calculating critical load and critical load exceedances in Norway, Finland and Sweden.     

Relations between organic carbon and methylmercury in humic rich surface waters from Svartberget catchment in northern Sweden

Monthly sampling of a mire outlet and two tributaries, one of them originating in the mire, on the Svartberget catchment in northern Sweden was performed during one year. The concentration of total organic carbon (TOC) in the three waters was fairly high (10–40 mg/l). Methylmercury (MeHg) was analysed in the original water sample (MeHg-whl) and in the humic fraction (MeHg-hum). The MeHg-hum increased with increasing concentration of humic substances (HS; measured as absorbance at 254 nm) in the water. A seasonal variation of the MeHg-hum/TOC ratio was superimposed on a negative relationship to the water flow, which indicates that the methylation is a slow process which results in a rapid drainage of the storage during periods of high flow. A minimum of the MeHg-whl/TOC ratio observed during the spring flood was followed by a slow increase during the rest of the year.     

Operational liming of surface waters in Sweden

From 1977 to 1982, liming of acidified waters was performed for a trial period using governmental subsidies. Based on the experience gained, an operational liming program was started in 1982 by the National Environmental Protection Board of Sweden. All waters with pH of less than 6 and/or an alkalinity of less than 0.05 meq L^?1 have been eligible for subsidies. Individual water owners as well as organizations and federal agencies may apply for subsidies. Normally 85% of the costs are covered by government grants. Each county administration draws up a 5 yr liming plan and applies yearly for subsidies on a county basis. Since 1977, SEK (Swedish Kroner) 373 million have been spent on liming operations and SEK 33 million on administration and follow-up studies. A yearly budget of about SEK 110 to 130 million has been proposed for 1988 to 1991. The liming agent used has been powdered CaCO₃ with a particle size normally of 0 to 0.2 mm. At present, about 4,000 lakes have been treated. The results of the treatments are normally evaluated by analyzing water samples taken twice a year. These are analyzed for pH, alkalinity, conductivity, color, Ca and Mg. In addition, biological surveys are carried out in selected lakes and streams.     

Sources of acidity in forest-floor percolate from a maple-birch ecosystem

Ion concentrations in water collected within a forest of sugar maple and yellow birch at the Turkey Lakes Watershed near Sault Ste. Marie, Ontario were examined from 1982 to 1984 to determine sources of acidity and the extent of cation leaching from forest floor horizons. Volume-weighted concentrations and ion fluxes in throughfall and forest-floor percolate during the growing and dormant seasons were calculated. Hydrogen ion content of the forest-floor percolate decreased in relation to that of throughfall in the dormant season and increased in the growing season. Hydrogen ion deposition in throughfall could account for 100% of the flux of H⁺ through the forest floor in the dormant period, and 40% of the flux during the growing season. In forest-floor percolate, Ca²⁺ concentrations were positively correlated with those of SO₄ ^2-, NO₃ ^- and organic anions during both dormant and growing seasons. Sources of NO₃ ^- and organic anions within the ecosystem and major external inputs of NO₃ ^- and SO₄ ^2- were critical factors that influenced cation mobility in the forest floor.     

Rainfall acidity in northern Britain — Exploring the data

Daily rainfall pH measurements from a site in south Scotland were found to have a bimodal distribution. Similarly, monthly data from a network of bulk rain collectors in northern Britain gave bimodal distributions. Sites with an annual average pH of 4.7 had occasional monthly pH values of pH 4 or less. The annual average pH was a poor indicator of the frequency of such acidic periods, and may not be a good indicator of potential effects. Data from a daily and monthly collector at the same site are compared and the discrepancies discussed. Daily pH measurements gave significantly greater acidities than monthly pH measurements, but these differences were small for months with rain-weighted pH < 5.     

Sources of sulphate and acidity in wetlands and lakes in Nova Scotia

There is a declining gradient of wet SO₄ deposition from south to north in Nova Scotia with the highest values being in the south, along with a localized increase around the Halifax metropolitan area, due to local SO₄ emission. Edaphic conditions such as drainage from soils containing gypsum or drainage on disturbed rocks containing pyrite, provide additional SO₄ to surface waters.Acidity is usually absent in the former (pH > 7.0) and very high in the latter (as low as pH 3.6). By contrast peaty, organic drainages release water low in SO₄ during the growing season but they release high amounts of organic anions (A?), consequently, these waters maintain decreased pH values, usually < 4.5. A study of over 80 wetlands and lakes during the ice free period in Nova Scotia showed that sea salt corrected SO₄ concentrations range from 45 ueq L^?1 in the south end of the province, ～30 ueq L^?1 in the Kejimkujik area and < 17 ueq L^?1 in the northern areas with values > 85 ueq L^?1 in the Halifax area, reflecting the atmospheric deposition pattern of SO₄ The SO₄ concentrations may be > 2000 ueq L^?1 in drainages containing gypsum, > 700 ueq L^?1 in drainages over pyrite bearing socks but < 20 ueq/L^?1 in streams draining bogs. The SO₄ concentrations change considerably during the non-growing season when the ground is saturated with water or frozen, and the runoff is high (snow and rain often alternate in winter). Under such conditions SO₄ concentration drops in the two former cases and increases in bog drainages, accompanied with a considerable drop in (A?) concentrations. Care should be taken when interpreting SO₄ concentrations in surface waters in Nova Scotia with respect to atmospheric SO₄ deposition.     

Increasing contributions of nitrogen to the acidity of surface waters in Norway

The acidification of lakes in southern Norway is largely due to sulfate. Recent data from regional surveys of lakes and from monitoring stations indicates that nitrate concentrations have increased in many lakes and rivers in southernmost Norway. The ratio of NO₃ to NO₃ + SO₄ is still low for most areas, but is 0.54 on an equivalent basis in lakes and rivers in the area of high runoff in southwestern Norway. Here, concentrations of nitrate in the runoff are lower than for sites in Central Europe, but are higher than those in North America. The sites showing increases in NO₃ also increased in Al. Further increases in nitrate as a mobile anion which may be due to decreased uptake in the watershed, will contribute to acidification in the same manner as sulfate.     

Output of methylmercury from a Catchment in northern Sweden

The runoff output of methylmercury (MeHg) from the 50 ha Svartberget Catchment in northern Sweden was studied during 1993. These outputs are compared to those from the Gårdsjön Catchment in the southwestern part of Sweden. Although the wet deposition of MeHg is several times higher in southern Sweden the output of MeHg per unit area from the two catchments was comparable (0.12 g /km_* ²yr). Furthermore, the concentration of MeHg (0.4 ng/L) in the Svartberget Catchment was more than twice that from the Gårdsjön Catchment. These results suggest that the contemporary wet deposition of MeHg in itself is not a good indicator of runoff concentration or output per unit area. DOC transport and catchment characteristics such as wetland area, or possibly other forms of contemporary atmospheric deposition may all be more important for explaining MeHg output from the terrestrial ecosystem.     

The role of weathering and forestry in determining the acidity of Lakes in Sweden

The long term acidity level of a lake is determined by the balance between acidity input to the catchment and the generation of alkalinity in the catchment. If the input of acidity through biomass net production and the production of alkalinity through weathering of minerals can be estimated, then the steady-state acidity level can be calculated for the lake under a certain acid deposition rate. Such a calculation has been carried out for 8 lakes ranging from acid to neutral. For lakes with the most sensitive soils in the catchment, the critical acid deposition load that will permit the lake to stay neutral, may be less than zero acidity, indicating that the forest growth is contributing to the acidification of very sensitive system under the present forest managements methods.     

Estimation of background sulphate concentrations in natural surface waters in Sweden

In order to measure the extent of acidification the background, ‘preindustrial’ conditions must be known. An equation for the estimation of background concentrations of sulphate in surface water in Norway was proposed by Henriksen. When applied on data from the Swedish lake survey in 1990 it was found that the calculated background concentrations exceeded those measured for about one-third of the lakes. The proposed revision is based on a background concentration in precipitation and an estimated contribution from weathering, the latter associated with base cations. Three different approaches were tested to establish the contribution from weathering; geochemical ratios or groundwater chemistry data as a basis, and historical data on denudification. The weathering calculated from groundwater chemistry data seems to give the best estimate of the background sulphate concentration in surface water. Organic matter as source or sink of sulphur is discussed and considered negligible.     

Implications of natural acidification for mitigation strategies in northern Sweden

Northern Sweden has been regarded as unaffected by acid deposition, but many surface waters in the region fall within the definition of acid surface water (pH < 6.0, alkalinity < 50 mmol_c m^?3) permanently or during episodes. Approximatly 100 MSEK in spent annually on liming in northern Sweden. This paper summarizes our conclusions from a workshop on natural versus anthopogenic acidification held in February 1995. It was shown that organic substances have a key role in determining the acidity of surface waters in the region, although anthropogenic effects are documented in some coastal systems and in the southern mountain range. Sulfide oxidation occurs by the coast. It appears clear that many surface waters that were naturally acidic have been limed to unnatural pH levels. New criteria to screen liming candidates should be developed, and one such model based on water chemistry data is proposed.     

Can forest-soil liming mitigate acidification of surface waters in Sweden?

Acidification of surface waters and forest soils is severe in large parts of southern Sweden. The shallow groundwaters are also affected. Large scale liming of surface waters and streams is in operation, often combined with wetland liming to limit the effects of acid episodes, e.g. at snow melt. Acid episodes are perhaps the most severe problem in limed surface waters and in many as yet well buffered waters, because of temperature-layered acid inflow, often superficial. As a result of some investigations, a large scale forest liming programme covering 6.500–10.000 km² was recently suggested. The main objectives of this forest liming programme are to retard cation depletion and to prevent nutrient imbalance and forest decline in acidified areas. This paper deals with the effects of forest soil liming on streams and surface waters. The response of water chemistry is very dependent on hydrological and soil properties. Although pH itself may be little affected by liming, the acidity (or negative ANC) decreases, inorganic Al-species decrease and the Al/BC-ratio increases in the runoff water. Especially interesting is that this is also true during acid episodes. This means that toxicity for acid sensitive biota decreases. These results indicate that large scale liming may have beneficial effects on surface water chemistry. Furthermore, as surface waters are expected to respond to smaller decreases in acid deposition than do forests soils, forest soil liming may allow less frequent liming of lakes. Consequently, forest soil liming in combination with the anticipated emission reductions may have very beneficial results on surface waters in certain areas of Sweden.     

Distributions of anuran populations in central Ontario relative to habitat acidity

Adult anuran densities in 20 ponds in central Ontario were determined by using daytime visual censuses and by counting males calling during the breeding season. The pH of ponds varied from 4.55 to 6.37 and alkalinity from ?1.9 mg L^?1 to 23.2 CaCO₃. Bullfrog (Rana catesbeiana), small (2 to 3 em snout-vent length) green frog (R. clamitans), and spring peeper (Hyla crucifer) densities were lower in more acidic ponds even though pH levels were not low enough to cause acute toxicity to eggs or larvae. The wood frog (R. sylvatica) was the only species whose density increased in more acidic ponds. In a census of another 40 ponds for wood frog egg masses (an estimate of the number of females breeding in the pond), egg mass density was reduced as pond pH and alkalinity decreased even though the pH levels were not low enough to be acutely toxic to eggs or larvae (pH ranged from 4.44 to 6.63). These results suggest that long before the H⁺ concentration in pond habitats increases to levels that cause acute toxicity, chronic sublethal effects in conjunction with indirect ecological effects of pond acidification (eg. changes in habitat structure, food resources, predator-prey relationships) may be affecting anuran populations.     

Atmospheric mercury in northern Wisconsin: Sources and species

The atmospheric chemistry, deposition and transport of mercury (Hg) in the Upper Great Lakes region is being investigated at a near-remote sampling location in northern Wisconsin. Intensive sampling over two years and various seasons has been completed. A multi-phase collection strategy (gas-, particle- and precipitation-phases) was employed to gain insight into the processes controlling concentrations and chemical/physical speciation of atmospheric Hg. Additional chemical and physical atmospheric determinations (e.g. ozone, particulate constituents, meteorology) were also made during these periods to aid in the interpretation of the Hg determinations. For example, correlations of Hg with ozone, sulfur dioxide and synopticscale meteorological features suggest a regionally discernible signal in Hg. Comparison to isosigma backward air parcel trajectories confirms this regionality and implicates the areas south, southeast and northwest of the site to be sources for Hg. Particle-phase Hg (Hg_p) was found to be approximately 40% in an oxidized form, or operationally defined as “reactive”. However, this was quite variable from year-to-year. Hg_p and other particle constituents (esp. sulfate) show significant correlation and similarity in behavior (concentration ratios in precipitation and in particles). These observations are part of the growing evidence to support the hypothesis that precipitation-phase Hg arises in large part from the scavenging of atmospheric particulates bearing Hg. Observed concentrations of rain and particle-Hg fit broadly the theoretical expectations for nucleation and below-cloud scavenging. Significant increases in the Hg/aerosol mass ratio appear to take place during transport. Enrichment of aerosols is taken as evidence of gas/particle conversion which could represent the step linking gas-phase Hg with rain. The refined budget indicates ca. 24% of total deposition is from summer particle dry deposition, and that this deposition also contributes ca. 24% of all reactive Hg deposition. Additionally, almost all (86%) deposition (wet and dry) occurs during the summer months.     

Sources and behavior of anthropogenic radionuclides in the Ottawa river waters

Multiphase radionuclide measurements on Ottawa River waters are reported for the period October 1984 to March 1986. Numerous radionuclides are present in detectable amounts in raw, drinking and centrifuged waters as well as in suspended sediment and water filtration plant floc samples. The sediment/water partitioning behavior of these radionuclides is also reported. The prevailing low particle flux allows rapid migration of radionuclides through the system. Most of the ⁹⁰Sr in the river derives from its fallout inventory in the watershed rather than from the Chalk River Nuclear Laboratories (CRNL). On the other hand, fallout and CRNL-derived ¹³⁷Cs appear to contribute equally to the prevailing levels of this radionuclide in the river. The data also suggest that the prevailing levels of the naturally-occurring ²²⁶Ra impart more dose than ⁹⁰Sr insofar as the radiological quality of the drinking water is concerned. Many of the radionuclides are removed by the floc during filtration of the receiving waters. Though their removal efficiencies cannot be precisely defined from the currently available measurements, the results show that processes other than simple association with suspended sediments may also play significant role in the retention of some radionuclides by the floc.     

Subcatchment output of mercury and methylmercury at Svartberget in northern Sweden

The subcatchments of two tributaries and the headwater mire of the 50 ha Swartberget catchment were studied. Monthly sampling was conducted during 1993 on the two tributaries and at the outlet of a mire. This was complemented by more intensive sampling during spring flood and a rain-driven episode at the end of July. Samples were analyzed for total methylmercury (MeHg), total mercury (Hg-tot) and TOC. The MeHg and TOC content of water were also fractionated into humic and non-humic components. Outputs budgets based on continuous flow monitoring and monthly volume weighted average concentrations of MeHg/ Hg were calculated for the subcatchments of the two tributaries and the mire. There was a pronounced decline in MeHg concentrations at all sampling locations during the spring flood. A clear pattern of the seasonal variation in the MeHg outputs was evident at all three sampling locations. Minimum MeHg concentrations occurred during spring flood and increased during the summer to peak levels in the autumn before tapering off during the winter (except for at the mire). The mire had larger increases in MeHg concentrations during the summer and did not decline during the winter. The warmth and increased soil humidity may promote the biogeochemical processes, including methylation, demethylation which make MeHg available for export by runoff from the soil. Variations in Hg-tot concentrations were differed from MeHg in streams that there were increases in Hg-tot concentrations with flow during both spring flood and the July rain event in runoff from the forested tributaries. The highest outputs of both humic and non-humic MeHg occurred during summer from all subcatchments. Those high outputs extended into the autumn at the mire. The largest monthly output of Hg-tot was during the spring flood period and the next largest was during the period of summer rainfall. Despite the similarity in mineralogy and atmospheric deposition on the two tributary subcatchments, there was ca 30% larger ouput of MeHg per unit area from the tributary which had deeper riparian peats. The output of Hg-tot, however, was higher on the catchment with the shallower riparian peats. The difference in the geometry of the riparian zone may contribute to these differences in output. The annual output concentrations at the mire outlet of MeHg was 0.65 ng /L and of Hg-tot was 4.04 ng/L. In the main tributary MeHg was 0.42 ng/L and Hg-tot was 3.64 ng/L. In the Västrabäcken tributary, the mean MeHg concerntration was 0.25 ng/L and the Hg-tot concerntration was 4.02 ng/L. Among the three subcatchments the largest annual output fluxes of MeHg were from the mire, 0.16 g/km2^*yr, the next from Kallkällbäcken below the mire, 0.12 g/km2^*yr, and the lowest from Västrabäcken, 0.08 g/km2^*yr.     

The natural acidity of some unpolluted waters in southeastern Alaska and potential impacts of acid rain

Natural, unpolluted waters draining surface soils and bog soils near Petersburg, Alaska have low pH (≤4.7) due to the presence of organic acids. Nearby streams are influenced by organics as well, but are only slightly acid (pH≥6). Projections of the effects of acid and sulfate inputs (acid rain) using a worst-case model indicate the greatest potential declines in pH would occur in the most acid waters, whereas very little change would occur in near neutral waters.     

Critical loads of acidity to surface waters in the Vosges massif (north-east of France)

Acid clearwater fishless streams have been identified in the Vosges mountains. In order to evaluate the relatipnships between acidifying factors (such as atmospheric deposition), buffering factors (such as bedrock and soil type), and surface water acidity, an exhaustive survey of streamwater acidity in the Vosges mountains (N-E France) was performed. A network of 11 measurement stations of atmospheric deposition was used to estimate and map deposition over the whole massif (total area 5000 km²). Data on bedrock, soil, superficial deposits, and vegetation were collected from published studies. Sensitive areas as well as acidifying environment factors were derived from the corresponding maps. Over the whole massif, 19% of streams showed baseflow alkalinity below 30 eq.r¹ and 7.5 % were identified as acid (pH < 5.4).=" acid=" streams=" occur=" on=" the=" north-western=" side=" of=" the=" massif=" on=" quartzrich=" sandstone=" and=" acid=" granites.=" in=" each=" of=" these=" areas,=" we=" could=" clearly=" point=" out=" on=" one=" hand,=" the=" negative=" influence=" of=" conifer=" vegetation=" and=" glacial=" soil=" abrasion=" or=" induration,=" and=" on=" the=" other=" hand=" the=" buffering=" effect=" of=" moraine=" deposits.=" a=" corresponding=" range=" of=" critical=" loads=">< 0.2=" to=" 2.0=" keq.=">¹ yr¹) for surface water was calculated using the Steady State Water, Chemistry method (SSWC).