Effect of Declining Lake Base Cation Concentration on Freshwater

Apr 2, 2005 - In central Ontario, SSWC critical load estimates based on data taken only 13 years apart change quite dramatically due to changing lake ...
0 downloads 0 Views 261KB Size
Download PDF
Environ. Sci. Technol. 2005, 39, 3255-3260

Effect of Declining Lake Base Cation Concentration on Freshwater Critical Load Calculations SHAUN A. WATMOUGH,* JULIAN AHERNE, AND PETER J. DILLON Environmental and Resource Studies, 1600 West Bank Drive, Trent University, Peterborough, Ontario, K9J 7B8, Canada

Steady-state critical load models have been extensively used as the scientific underpinning for air pollution control policies in Europe and are currently being applied to other parts of the world. An important assumption of steadystate models is that critical load estimates do not change through time (or time scale of interest). The most commonly used model for estimating freshwater critical loads is the steady-state water chemistry (SSWC) model. In this study we examined changes in SSWC critical load estimates for 29 lakes in south-central Ontario using data collected 13 years apart (1985-1998), during which time bulk sulfate (SO42-) deposition decreased by 35%. In lakes with the lowest base cation concentrations (150 µequiv L-1 Ca2+), the relative decline in base cation concentration was not as great, resulting in a larger increase in ANC (median increase 13.1 µequiv L-1) and an increase in the estimated CL(A) (median 5.1%). Lakes with moderate base cation concentrations (120-150 µequiv L-1 Ca2+) exhibited an intermediate response; the median ANC increased by 8.8 µequiv L-1 and the estimated CL(A) decreased by 2.2%. In central Ontario, SSWC critical load estimates based on data taken only 13 years apart change quite dramatically due to changing lake base cation concentrations, and the response appears to depend on the base status of the lakes. The changing values obtained from the SSWC model have important consequences for policy decisions regarding acceptable levels of acid deposition. The application of dynamic models that take into account changes in lake/soil chemistry appears more appropriate for estimating acceptable levels of acid deposition in the region.

Introduction Critical loads are defined as “a quantitative estimate of an exposure to one or more pollutants below which significant * Corresponding author phone: 705-748-10711, ext 1647; fax: 705748-1569; e-mail: [email protected]. 10.1021/es048607t CCC: $30.25 Published on Web 04/02/2005

 2005 American Chemical Society

harmful effects on specified sensitive elements of the environment do not occur according to present knowledge” (1). Steady-state critical load models have been extensively used as the scientific underpinning for air pollution control policies in Europe and are currently being applied to other parts of the world (2-6). An important assumption of steadystate models is that critical load values do not change through time. The time-invariance of critical loads has its limitations, certainly when considering a geological time frame, but also during shorter time periods, such as centuries or decades, one can anticipate changes in the magnitude of critical loads due to global (climate) change (7). Traditionally, surface water critical load models are used to calculate acceptable levels of sulfate (SO42-) deposition to freshwaters (8). The most common method for calculating critical loads of acidity, CL(A), for lakes is the steady-state water chemistry (SSWC) model (9-11). The SSWC model uses the so-called F factor to account for the fraction of present base cation leaching that is due to ion exchange processes in soils (10). Using this approach, Henriksen et al. (10) estimated that the number of study lakes from four regions in central Ontario that exceeded the CL(A) has decreased from 74-82% to 11-26% as bulk SO42- deposition has decreased. During the past 2 decades, however, surface soils in the region appear to have been acidifying as base cation losses exceed estimated inputs from mineral weathering and deposition (12-14). If soils are acidifying, they will be less able to ‘buffer’ a given amount of acid leaching (F factor decreases) and base cation losses from soil associated with acid leaching will decrease (15-18). Any rapid changes in soil chemistry could have a pronounced impact on SSWC critical load estimates, which has important implications for setting acceptable levels of acid deposition for freshwaters. In this study, we evaluate changes in estimated CL(A) for 29 lakes in south-central Ontario between 1985 and 1998, during which time SO42- deposition decreased by 35% (58-37 mequiv m-2 yr-1 SO42-), but nitrogen (N) deposition remained unchanged (14).

Materials and Methods Study Region. The chemistry of freshwaters in central Ontario is related in general terms by their location relative to the boundary of the Precambrian Shield. Lakes and streams on the shield generally contain soft waters (conductivity typically ranges between 15 and 40 µS cm-1) that are low in nutrients as well as major ions. Many contain appreciable levels of dissolved organic carbon (DOC), and organic anions often form a major component of the charge balance. The climate of the study area is north temperate, with long-term average precipitation ranging between about 8001100 mm, about one-quarter to one-third of which falls as snow. The mean monthly air temperatures for January and July are -10 and 19 °C, respectively, and the long-term annual average temperature is approximately 5 °C. On the Precambrian Shield, the land cover is largely mixed forest, with deciduous forests dominating where the soil is thicker, e.g., outwash plains and deeper glacial tills, and conifers dominating where the soils are thin. Small wetlands are ubiquitous throughout the entire region, covering an estimated 10% of the total area. All lakes in the current study are located in south-central Ontario (District of Muskoka and the Counties of Haliburton and Nipissing) on the Precambrian Shield (Figure 1). The majority of the 29 study lakes are headwater lakes and are sensitive to the deposition of strong acids because of their geological setting (10). The study lakes range in size from 5.4 VOL. 39, NO. 9, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3255

FIGURE 1. Location of the 29 study lakes in the District of Muskoka and counties of Haliburton and Nipissing, south-central Ontario, Canada (see Table 1 for lake names).

TABLE 1. Selected Characteristics [lake ID, name, location, area, mean depth, lake-to-catchment ratio (r), runoff (Q), flushing rate, 1985 concentrations for dissolved organic carbon (DOC), base cations (BC ) Ca2+ + Mg2+ + K+), and sulfate (SO42-), and the Change in Concentrations between 1985 and 1998 for Base Cations (∆BC) and Sulfate (∆SO42-)] for the 29 Study Lakes in South-Central Ontario

ID

lake name

location

BH BW BC BY BK CB CIE CI CL CN DE FN GF HP HE HY LH LD LC MY MO PC PRE PR RCE RCM SE WR WD

Basshaunt Lake Bigwind Lake Blue Chalk Lake Brandy Lake Buck Lake Chub Lake Cinder Lake East Cinder Lake West Clear Lake Crosson Lake Dickie Lake Fawn Lake Gullfeather Lake Harp Lake Healey Lake Heney Lake Leech Lake Leonard Lake Little Clear Lake Mckay Lake Moot Lake Plastic Lake Poker Lake East Poker Lake West Red Chalk Lake East Red Chalk Lake Main Solitaire Lake Walker Lake Westward Lake median

45° 07′ N, 78° 28′ W 45° 03′ N, 79° 03′ W 45° 12′ N, 78° 56′ W 45° 06′ N, 79° 31′ W 45° 23′ N, 79° 00′ W 45° 13′ N, 78° 59′ W 45° 04′ N, 78° 56′ W 45° 04′ N, 78° 56′ W 45° 11′ N, 78° 43′ W 45° 05′ N, 79° 02′ W 45° 09′ N, 79° 05′ W 45° 10′ N, 79° 15′ W 45° 06′ N, 79° 01′ W 45° 23′ N, 79° 08′ W 45° 05′ N, 79° 11′ W 45° 08′ N, 79° 06′ W 45° 03′ N, 79° 06′ W 45° 04′ N, 79° 27′ W 45° 24′ N, 79° 00′ W 45° 03′ N, 79° 10′ W 45° 09′ N, 79° 10′ W 45° 11′ N, 78° 50′ W 45° 03′ N, 78° 56′ W 45° 03′ N, 78° 56′ W 45° 11′ N, 78° 57′ W 45° 11′ N, 78° 57′ W 45° 23′ N, 79° 01′ W 45° 23′ N, 79° 05′ W 45° 29′ N, 78° 45′ W

lake area (ha)

mean depth (m)

r

47.3 111.0 52.4 108.0 40.3 34.4 50.1 26.9 88.4 56.7 93.6 85.8 65.9 71.4 122.0 21.4 82.0 195.0 10.9 121.5 46.2 32.1 5.4 15.3 13.1 44.1 124.0 68.2 63.3 56.7

7.7 10.7 8.5 3.5 10.9 8.9 10.1 4.8 12.4 9.2 5.0 3.5 4.8 13.3 2.8 3.3 6.3 6.9 8.1 5.2 2.7 7.9 6.9 6.3 5.7 16.7 13.3 6.2 20.5 6.9

0.06 0.21 0.33 0.03 0.20 0.13 0.07 0.15 0.28 0.10 0.19 0.06 0.06 0.12 0.16 0.19 0.24 0.28 0.02 0.14 0.05 0.25 0.03 0.13 0.02 0.07 0.31 0.22 0.28 0.14

to 195 ha, with mean depth between 2.7 and 20.5 m and lake to catchment area ratios (r) between 0.02 and 0.33 (Table 1). Critical Loads. The SSWC model has been described in detail by a number of authors (8, 19, 20). Briefly, the model uses current base cation concentrations in lakes to estimate inputs in preindustrial times via the F factor. This preindustrial base cation input determines the critical load of the lake. The current application follows the most recent 3256

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 9, 2005

BC SO42∆BC ∆SO42flushing DOC (µequiv (µequiv (µequiv (µequiv L-1) L-1) Q (m) rate (yr) (mg L-1) L-1) L-1) 0.429 0.456 0.496 0.514 0.568 0.518 0.431 0.431 0.470 0.468 0.510 0.514 0.472 0.592 0.481 0.511 0.469 0.519 0.565 0.470 0.508 0.458 0.423 0.423 0.493 0.493 0.570 0.582 0.502 0.493

1.0 4.9 5.7 0.2 3.9 2.2 1.6 1.6 7.5 1.9 1.9 0.4 0.6 2.6 1.0 1.2 3.3 3.8 0.3 1.6 0.3 4.3 0.5 1.9 0.3 2.5 7.3 2.4 11.3 1.9

4.2 3.1 1.8 12.4 2.7 4.5 5.2 5.2 1.6 4.0 4.9 8.9 5.1 3.8 6.0 2.9 4.0 3.1 2.5 5.1 6.6 2.1 6.4 5.5 2.8 2.4 2.1 3.3 1.5 4.0

291.3 204.0 198.7 315.0 216.7 182.2 176.7 180.4 172.1 165.2 185.9 190.5 192.8 231.3 182.4 159.1 197.4 175.7 243.7 220.3 136.0 141.5 217.6 213.4 217.0 205.5 205.3 219.4 167.7 197.4

151.5 147.5 134.1 113.3 140.9 152.9 133.4 128.5 167.2 143.4 144.2 125.7 154.0 157.3 121.5 147.7 157.0 147.0 137.6 150.2 91.8 132.6 112.4 122.8 135.7 139.6 139.8 138.4 132.9 139.6

0.0 -32.6 -11.7 26.5 -15.7 -30.4 -23.9 -19.9 -26.0 -31.4 2.2 -3.6 -38.1 -17.5 1.1 -31.0 -32.7 -6.7 39.2 -31.6 4.4 -25.7 -49.9 -43.4 -20.7 -24.9 -11.7 28.5 -19.8 -19.9

-25.5 -33.3 -21.2 22.6 -24.8 -31.2 -17.8 -7.7 -31.3 -32.3 -34.2 -9.6 -32.7 -28.9 -9.6 -50.3 -43.4 -29.9 -1.5 -35.7 11.2 -18.7 -24.0 -27.0 -20.7 -22.6 -19.0 -14.3 -29.3 -24.8

published description (8, 10); CL(A) for surface waters (expressed as an annual flux, mequiv m-2 yr-1) is calculated as

CL(A) ) Q × ([BC]0 - [ANC]limit)

(1)

where Q is the mean annual runoff (m yr-1), [BC]0 is the preacidification base cation (BC ) Ca2+ + Mg2+ + K+ + Na+)

concentration in lake water and [ANC]limit is the preselected damage-threshold level for sensitive indicator organisms. In this study, an [ANC]limit of 40 µequiv L-1 was chosen to match that used by Henriksen et al. (10) for lakes in south-central Ontario. The [BC]0 is estimated from the present base cation concentration minus the long-term changes in base cation concentration, inferred from the inputs of strong acid anions through the F factor

[BC]0 ) [BC]t - F × ([SO42-]t + [NO3-]t - [SO42-]0 -

[NO3-]0) (2)

where [NO3-] is the nitrate concentration in lake water and the subscripts 0 and t refer to preacidification and present concentrations, respectively. The F factor represents the fraction of present base cations due to soil acidification. Within a region the F factor can vary both spatially and temporally depending on catchment characteristics and acid deposition. Operationally the F factor is an estimated from observed lake chemistry; as such, temporal variation of the F factor may be caused by changes in lake chemistry data, which can vary from year-to-year. These variations might greatly affect model estimates, but this process cannot be considered in steady-state models. In general, the F factor is estimated as a function of base cation concentration (8, 21)

(

)

π ×Q×[BC]t 2 F ) sin S

(3)

where S is the base cation flux at which F ) 1. For Norway, S has been estimated to be 400 mequiv m-2 yr-1, and this value was used for both time periods in the present study. If Q × [BC]t > 400, then F is set to 1. A range of values have been suggested for S (e.g., 200-400 µequiv L-1 (21)); depending on the hydrological characteristics of the region, different values of S may have a considerable impact on critical load calculations (6, 8, 10). Sensitivity analysis of the F factor, for the current study region and elsewhere, has been described previously (6, 10, 22). Obviously, critical load estimates for lakes will vary depending upon the formulation of the F factor. However the focus of this study is not the evaluation of the F factor per se, but whether SSWC critical load estimates for lakes vary between two sampling periods using the same model parameterization. The [SO42-]0 is estimated from the relationship between [SO42-]t and [BC]t for lakes in regions little affected by acid, and [NO3-]0 is assumed to be zero. For Norway the following relationship exists (8):

[SO42-]0 ) 8 + 0.17 × [BC]t

(4)

This equation suggests that there is an atmospheric background contribution of [SO42-] equal to 8 µequiv L-1 and a geological contribution that is proportional to the concentration of base cations. A number of different forms of the relationship have been estimated for different regions (8); however, sensitivity analysis has shown the different forms to have little influence on the number of exceeded lakes (6, 8, 10). Similar to the F factor, in this study we use the same relationship for preindustrial SO42- for all lakes during both summary periods. Input Data. Chemical data for lakes in south-central Ontario have been collected during the course of a number of different research projects and lake surveys since the late 1970s. The current study focuses on a subset of the regional data set that includes only those lakes with 10+ years of monitoring data during the period 1980-1999 (29 lakes; see Figure 1, Table 1). Lake chemistry at the start and end of the

monitoring period was described using three-year annual averages centered on 1985 and 1998 (the number of available data-years for the 1998 average ranged from one to three). All lakes were analyzed for alkalinity, conductivity, pH, and major cations and anions. Chemical analyses were carried out at Ontario Ministry of the Environment laboratories; methods are described in detail by the Ontario Ministry of the Environment (23). Further details are given by Henriksen et al. (10). Sodium and chloride (Cl-) have been excluded from all calculations due to road salt contamination in some of the lakes, although inputs of Na+ and Cl- in deposition are low due to the large distance to the nearest ocean. Road salt is primarily NaCl, and as such, lake chemistry was not corrected for sea salt fractions of base cations (Mg2+, Na+, K+) or SO42-. Runoff data (m yr-1) were interpolated from measured 30year annual average runoff at all long-term hydrologic gauging stations throughout Ontario to produce a provincial runoff map (Cumming Cockburn Ltd., unpublished map). The 30year annual average runoff is used for all lakes (Table 1). Continuous flow measurements were available for a subset of the lakes (six) and indicate very little difference in runoff between the periods for which the chemistry was averaged. In the present study we classified the acid sensitivity of lakes based on their (1985) Ca2+ concentration (dominant base cation), with lakes having >150 µequiv L-1 Ca2+ in 1985 to be representative of acid-insensitive lakes. Lakes with 150 µequiv L-1). Asterisks indicate lakes that have DOC > 6 mg L-1. a major factor delaying the chemical improvement of these lakes in response to declining SO42- deposition. Similar examples of declining base cation concentrations in lakes have been reported in other regions of Europe and eastern North America (27-30). Although there was a general tendency for base cation concentrations in lakes to decrease, there was considerable variation among lakes. Specifically, the change in lake base cation concentration between the two sampling periods appeared to vary depending upon the base status of the lakes. Similar varying chemical responses of lakes to declining acid inputs have been reported (31). Change in Lake ANC, F factor, and CL(A) Estimates (1985-1998). Between 1985 and 1998, the median ANC for the “insensitive” lakes increased by 13.1 µequiv L-1 (Figure 2). In contrast, the median ANC for the “sensitive” lakes increased by only 1.6 µequiv L-1. Lakes classified as “moder3258

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 9, 2005

ate” exhibited an intermediate response with a median ANC increase of 8.8 µequiv L-1 (Figure 2). The lack of increase in ANC, particularly in the sensitive lakes, is due to the decline in lake base cation concentration that has accompanied the decline in SO42- concentration in lakes. The variable response of base cations in lakes among the different sensitivity classes adopted in this study has also resulted in substantial changes in the F factor and estimated CL(A) for the lakes in the region (Figure 2). The F factor for insensitive lakes increased slightly (median, 3.6%), whereas the median F factor for moderate and sensitive lakes decreased by 6.9 and 14.0%, respectively (Figure 2). The estimated CL(A) for insensitive lakes increased slightly (median, 5.1%), whereas the median CL(A) for moderate and sensitive lakes decreased by 2.2 and 14.6%, respectively (Figure 2). Clearly, the change in base cation concentration in lakes during just a 13-year period has had a substantial impact on the estimated critical load of the

lakes using the SSWC model. Depending on which data set (1985 vs 1998) is used, acceptable acid deposition levels for the region (based on the most sensitive lake in the study) differs by almost 25%. Possible Causes of Declining Base Cations in Lakes. A reduction in SO42- deposition and hence concentration in lakes should result in a proportional decrease in base cation concentrations due to the decline in ionic strength. However, the decrease in lake base cation concentrations in the sensitive lakes was greater than expected from declining SO42alone, resulting in a general decrease in the F factor and the critical load and a lower than expected increase in lake ANC (Figure 2). Assuming mineral weathering rates to be relatively invariant during the study period, declining base cation concentrations in lakes may be a result of soil acidification or retention of base cations in soils. The rationale for each argument is outlined below. Loss of Base Cations from Soil. Changes in lake base cation concentrations may be a result of continuing soil acidification in the surrounding catchments. The effect of soil acidification on base concentrations in lakes is expected to be greatest in lakes dominated by shallow soils with low weathering rates (sensitive lakes in this study). Long-term mass balance studies conducted in the region have suggested that soils, particularly the upper horizons, have continued to acidify during the 1980s and 1990s despite declining SO42- deposition (12-14). These results have also been supported by repeated field measurements (13-14, 18). In the sensitive lakes in this study the median increase in ANC during the 13-year study period was only 1.6 µequiv L-1, supporting the contention that soil acidification is responsible for declining base cation concentrations in sensitive lakes. In lakes that are considered less sensitive to acid deposition, which in this region implies greater soil overburden, the response of lakes to declining acid deposition may be different. In these lakes base cation inputs are derived from a combination of upper and deeper soil horizons (groundwater). Water draining the upper soil horizons should respond in a similar manner as shallow soils, but the base cation input from groundwater will be less affected by declining acid leaching. The net effect is that the lake ANC should increase in response to declining acid deposition. In the present study the median ANC of insensitive lakes increased by 13.1 µequiv L-1. It is worth noting that there is also considerable variation within each of the classes adopted in this study, which may be due to a number of factors. For example, base cation inputs from weathering are generally much higher than deposition in the region, and therefore, the classification based on Ca2+ concentrations used to separate the lakes is obviously influenced by other factors such as lake to catchment area, as well as any small differences in soil mineralogy. Second, DOC concentrations vary among lakes and can form a major component of the charge balance (Table 1). The SSWC model does not take into account differences in DOC concentration, and in some lakes DOC will have a considerable influence on base cation inputs into lakes and, hence, the estimated CL(A) influence on base cation concentrations in lakes. For example, lakes in the region with high DOC levels have been shown to have lower SO42concentrations than low DOC lakes due to the formation of organo-S complexes (24). Similarly, in lakes with high DOC, SO42- will comprise a smaller component of the anion charge. Combined, these features will mean that high DOC lakes will be less responsive to decreases in S deposition and will not behave in a fashion similar to low DOC lakes. In this study we have highlighted lakes that contain >6.0 mg L-1 DOC, where CL(A) calculations may be most affected (Figure 2). Again, this is an arbitrary distinction because DOC concentrations in lakes may also be affected by other parameters

such as the pH of the lake (32). Other uncertainties not accounted for in this study include potential disturbances such as logging, insect defoliation, fire, and any developments that occurred during the study period and influenced base cation or acid anion concentrations in lakes, although during either of the summary periods there is no reason to believe that any of these disturbances affected any of our lake classes preferentially. Retention of Base Cations in Soil. The counter argument for declining base cation concentrations in lakes is that base cations are currently being retained in the soils. Stoddard et al. (31) showed that the increase in ANC was greater in acidsensitive lakes in the eastern United States compared with less sensitive lakes in response to declining SO42-, suggesting the response was due to retention of base cations in soil. To evaluate this hypothesis, we have followed the approach described by Kirchner and colleagues (15-17), who determined that the relationship between acid leaching and base cations (also hydrogen and aluminum) in streams draining acid forest soils is linear and very predictable and that a decrease in acid leaching will bring about a predictable decrease in base cation leaching (15-17). Similar findings have been shown to exist for catchments in the study region (16-18). Soils with smaller base cation pools will buffer a smaller amount of the acid leaching, leading to lower ANC values in water (sensitive lakes in this study). For a given decrease in acid leaching (and assuming no soil acidification), base cation leaching should decrease in a predictable fashion (15-18). However, because soils with lower base cation pools buffer a smaller proportion of the acidity, the change in ANC (base cations - acid anions) should be greater than in wellbuffered systems. As a result, acid sensitive lakes should show a greater increase in ANC compared with less sensitive lakes for a given decrease in acid leaching, provided there is minimal change in soil chemistry. Previous studies have reported that an improvement in the ANC in response to declining acid deposition was most notable in acid-sensitive lakes (31). In the present study, however, the increase in ANC was actually greatest in the least sensitive lakes (Figure 2). Furthermore, if soils were retaining base cations, the exchangeable base cation pool in soils should increase, which should lead to an even greater increase in lake ANC levels, particularly in the acid-sensitive lakes. The fact that there has been generally minimal improvement in water quality (ANC) of the acid-sensitive lakes (Figure 2) is difficult to reconcile with increasing base cation levels in soil, particularly in light of the expected response of these lakes to declining acid leaching. It would also imply that the combined base cation inputs by weathering and deposition are greater than currently measured in lakes. These base cation weathering rates would be much higher than estimated for soils in the region (12-14, 33) and are also inconsistent with dynamic model assessments for the region that indicate soils are still acidifying (34). Regardless of the exact cause of base cation decrease in lakes, the large change in base cation concentration has resulted in substantial changes in the critical load estimated by the SSWC model during just a 13-year period. The extent of changes in base cation concentrations also appear to be affected by the acid sensitivity of the lakes, with the least acid sensitive lakes considered in this study showing the greatest improvement in ANC. In Canada, regional critical load estimates for lakes are currently determined using the SSWC model, although dynamic models have been applied in a few localized areas. A major disadvantage of using a simple steady-state model is the inability to capture the temporal variation of the F factor, which can greatly affect model estimates. Due to the large difference in critical loads obtained by the SSWC, the application of a dynamic model VOL. 39, NO. 9, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3259

that takes into account temporal changes in soil chemistry appears to be a more appropriate method of estimating acceptable levels of acid deposition to lakes.

Acknowledgments This work was funded by grants from the Natural Sciences and Engineering Council of Canada (NSERC) and Ontario Power Generation (OPG).

Literature Cited (1) Nilsson, J.; Grennfelt, P. (Editors). Critical Loads for Sulphur Nitrogen; Nordic Council of Ministers: Copenhagen, 1988. (2) Skeffington, R. A. Use of critical loads in environmental policy making: A critical appraisal. Environ. Sci. Technol. 1999, 33, 245-252. (3) Henriksen, A.; Skjelkvale, B. L.; Mannio, J.; Wilander, A.; Harriman, R.; Curtis, C.; Jensen, J. P.; Fjeld, E.; Moiseenko, T. Northern European Lake Survey, 1995. Finland, Norway, Sweden, Denmark, Russian Kola, Russian Karelia, Scotland and Wales. Ambio 1998, 27, 80-91. (4) Duan, L.; Hao, J.; Xie, S.; Du, K. Critical loads of acidity for surface waters in China. Sci. Tot. Environ. 2000, 246, 1-10. (5) Snucins, E.; Gunn, J.; Keller, B.; Dixit, S.; Hindar, A.; Henriksen, A. Effects of regional reductions in sulphur deposition on the chemical and biological recovery of lakes within Killarney Park, Ontario, Canada. Environ. Monit. Assess. 2001, 67, 179-194. (6) Aherne, J.; Curtis, C. J. Critical loads of acidity for Irish lakes. Aquat. Sci. 2003, 65, 21-35. (7) Posch, M. Impacts of climate change on critical loads and their exceedances in Europe. Environ. Sci. Policy 2002, 5, 307-317. (8) Henriksen, A.; Posch, M. Steady-state models for calculating critical loads of acidity for surface waters. Water Air Soil Pollut.: Focus 2001, 1, 375-398. (9) Henriksen, A.; Kamari, J.; Posch, M.; Wilander, A. Critical loads of acidity: Nordic surface waters. Ambio 1992, 21, 356-363. (10) Henriksen, A.; Dillon, P. J.; Aherne, J. Critical loads of acidity for surface waters in south-central Ontario, Canada: Regional application of the Steady-State Water Chemistry (SSWC) Model. Can. J. Fish. Aquat. Sci. 2002, 59, 1287-1295. (11) Posch, M.; de Smet, P. A. M.; Hettelingh, J.-P.; Downing, R. J. (Editors). Modelling Mapping of Critical Thresholds in Europe, Status Report 2001; National Institute of Public Health and the Environment (RIVM): Bilthoven, 2001. (12) Watmough, S. A.; Dillon, P. J. Base cation and nitrogen budgets for seven forested catchments in central Ontario, 1983-1999. For. Ecol. Manage 2003, 177, 155-177. (13) Watmough, S. A.; Dillon, P. J. Base cation and nitrogen budgets for a mixed hardwood catchment in south-central Ontario. Ecosystems 2003, 6, 675-693. (14) Watmough, S. A.; Dillon, P. J. Major element fluxes from a coniferous catchment in central Ontario, 1983-1999. Biogeochemistry 2004, 67, 369-398. (15) Kirchner, J. W.; Lydersen, E. Base cation depletion and potential long-term acidification of Norwegian catchments. Environ. Sci. Technol. 1995, 29, 1953-1960. (16) Kirchner, J. W.; Dillon, P. J.; LaZerte, B. D. Predicted response of stream chemistry to acid loading tested in Canadian catchments. Nature 1992, 358, 478-482. (17) Kirchner, J. W.; Dillon, P. J.; LaZerte, B. D. Predictability of geochemical buffering and runoff acidification in spatially heterogeneous catchments. Wat. Resour. Res. 1993, 29, 38913901. (18) Watmough, S. A.; Dillon, P. J. Calcium losses from a forested catchment in south-central Ontario, Canada. Environ. Sci. Technol. 2003, 37, 3085-3089.

3260

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 9, 2005

(19) Sverdrup, H.; de Vries, W.; Henriksen, A. Mapping Critical Loads; Nordic Council of Ministers: Copenhagen, 1990. (20) Werner, B.; Spranger, T. (Eds.). Manual on Methodologies Criteria for Mapping Critical Levels/Loads Geographical Areas Where They Are Exceeded; Federal Environmental Agency (Umweltbundesamt): Berlin, 1996. (21) Brakke, D. F.; Henriksen, A.; Norton, S. A. A variable F-factor to explain changes in base cation concentrations as a function of strong acid deposition. Verh. Internat. Verein. Limnol. 1990, 24, 146-149. (22) Forsius, M.; Ka¨ma¨ri J.; Posch, M. Critical loads for Finnish lakes: Comparison of three steady-state models. Environ. Pollut. 1992, 77, 185-193. (23) Ontario Ministry of the Environment. Handbook of Analytical Methods for Environmental Samples; Ontario Ministry of the Environment: Rexdale, 1983. (24) Eimers, M. C.; Dillon, P. J. Climate effects on sulphate flux from forested catchments in south-central Ontario. Biogeochemistry 2002, 61, 337-355. (25) Eimers, M. C.; Dillon, P. J.; Watmough, S. A. Long-term (18year) changes in sulphate concentrations in two Ontario headwater lakes and their inflows in response to decreasing deposition and climate variations. Hydrol. Process. 2004, 18(14), 2617-2630. (26) Devito, K. D.; Hill, A. R.; Dillon, P. J. Episodic sulphate export from wetlands in acidified headwater catchments: Prediction at the landscape scale. Biogeochemistry 1999, 44, 187-203. (27) Stoddard, J. L.; Jeffries, D. S.; Lu ¨ kewille, A.; Clair, T. A.; Dillon, P. J.; Driscoll, C. T.; Forsius, M.; Johannessan, M.; Kahl, J. S.; Kellogg, J. H.; Kemp, A.; Mannio, J.; Monteith, D. T.; Murdoch, P. S.; Patrick, S.; Rebsdorf, A.; Skjelkvåle, B. L.; Stainton, M. P.; Traaen, T.; Van Dam, H.; Webster, K. E.; Wieting, J.; Wilander, A. Regional trends in aquatic recovery from acidification in North America and Europe. Nature 1999, 401, 575-578. (28) Clair, T. C.; Ehrman, J. M.; Ouellet, A. J.; Brun, G.; Loackerbie, D.; Ro, C. Changes in Freshwater Acidification Trends in Canada’s Atlantic Provinces: 1983-1997. Water Air Soil Pollut. 2002, 135, 335-354. (29) Driscoll, C. T.; Postek, K. M.; Kretser, W.; Raynal, D. J. Longterm trends in the chemistry of precipitation and lake water in the Adirondack region of New York, USA. Water Air Soil Pollut. 1995, 85, 583-588. (30) Keller, W.; Dixit, S. S.; Heneberry, J. Calcium declines in northeastern Ontario lakes. Can. J. Fish. Aquat. Sci. 2001, 58, 2011-2020. (31) Stoddard, J.; Kahl, J. S.; Deviney, F.; DeWalle, D.; Driscoll, C. T.; Herlihy, A.; Kellogg, J.; Murdoch, P.; Webb, J.; Webster, K. Response of Surface Water Chemistry to the Clean Air Act Amendments of 1990; U.S. Environmental Protection Agency: Research Triangle Park, 2003. (32) Gennings, C.; Molot L. A.; Dillon, P. J. Enhanced photochemical loss of organic carbon in acidic waters. Biogeochemistry 2001, 52, 339-354. (33) Kirkwood, D. E.; Nesbitt, H. W. Formation and evolution of soils from an acidified watershed: Plastic Lake, Ontario, Canada. Geochim. Cosmochim. Acta. 1991, 55, 1295-1308. (34) Aherne, J.; Dillon, P. J.; Cosby, B. J. Acidification and recovery of aquatic ecosystems in south-central Ontario, Canada: Regional application of the MAGIC model Hydrol. Earth Sys. Sci. 2003, 7, 561-573.

Received for review September 7, 2004. Revised manuscript received March 1, 2005. Accepted March 2, 2005. ES048607T