Modeling Future Acidification and Fish Populations in Norwegian

Norwegian Institute for Water Research, Gaustadalleen 21, 0349 Oslo, Norway, and Department of Environmental Sciences, University of Virginia, ...
0 downloads 0 Views 4MB Size
Download PDF
Environ. Sci. Technol. 2010, 44, 5345–5351

Modeling Future Acidification and Fish Populations in Norwegian Surface Waters THORJØRN LARSSEN,† BERNARD J. COSBY,‡ ESPEN LUND,† AND R I C H A R D F . W R I G H T * ,† Norwegian Institute for Water Research, Gaustadalleen 21, 0349 Oslo, Norway, and Department of Environmental Sciences, University of Virginia, Charlottesville, Virginia 22904-4123

Received November 26, 2009. Accepted June 8, 2010.

Despite great progress made in the past 25 years, acid deposition continues to cause widespread damage to the environment in Europe and eastern North America. Legislation to limit emissions of sulfur and nitrogen compounds in Europe is now under revision. The most recent protocol was based in part on the critical loads concept. The new protocol may also take into consideration the time delays between dose and response inherent in natural ecosystems. Policy decisions to reduce adverse effects on ecosystems entail a trade-off: quick response will require deeper cuts in emissions and thus higher costs, whereas lower costs with lesser cuts in emissions will give slower response. Acidification of lakes and damage to fish populations in Norway is used as an example. Under current legislation for emission reductions, surface waters will continue to slowly recover, but for many decades lakes in about 18% of Norway will continue to have water quality insufficient to support healthy populations of brown trout and other indicator organisms. Additional emission reductions can speed up the rate and degree of recovery.

1. Introduction During the 1900s chronically high emissions of sulfur (S) and nitrogen (N) compounds to the atmosphere caused acid deposition, acidification of surface waters, and damage to fish populations over large regions of Europe (1-5). In response to the environmental damage of transboundary air pollution, in the 1970s governments in Europe began work on developing international policies to reduce emissions of S and N. These efforts have largely been organized under the auspices of the United Nations Economic Commission for Europe (UNECE) as the Convention on Long-Range Transboundary Air Pollution (LRTAP) (http://www.unece.org/env/lrtap) (5). The first protocols under the LRTAP Convention called for fixed percent reductions (so-called “flat rate”) in emissions by country within a specified year relative to a given base year. Thus the First Sulfur Protocol called for a 30% reduction in emissions by the year 1993 relative to the base year 1980. Because emissions from various sources are not transported and deposited uniformly, and because the natural environment has large variations in susceptibility to acidification, it * Corresponding author phone +47 95889769; fax: +47 22185200; e-mail: [email protected]. † Norwegian Institute for Water Research. ‡ University of Virginia. 10.1021/es100792m

 2010 American Chemical Society

Published on Web 06/22/2010

was clear that a strategy of flat-rate reductions was not the most cost-effective approach. Some of the reductions would probably give little or no environmental improvement. Work under the Convention thus progressed to “scientific-based” protocols, which with respect to acidification were largely based on critical loads to “semi-natural” ecosystems (pristine and managed forests, freshwaters, and heathlands). Nilsson and Grennfelt (6) defined the critical load as “the highest annual deposition level at which adverse effects on natural ecosystems are unlikely to result in the long term”. European countries produced national maps of critical loads for various types of ecosystems, and these were used together with a regional atmosphere emission-transport-deposition model (RAINS) (7) to optimize the environmental gains of reducing emissions most efficiently. The most recent protocol, the multipollutant, multieffect Gothenburg protocol signed in 1999, aimed at obtaining reductions in the total area of ecosystems to which acid deposition exceeded the critical load. Under the Gothenburg protocol countries were given individual targets for reducing of S and N emissions, and these are to be met by the year 2010 (http://www.unece.org/ env/lrtap) (5). The critical load approach is static; no consideration is given to the fact that ecosystems require time to recover in response to decreases in acid deposition. For surface waters such time lags are caused by chemical processes and slow changes to the large pools of base cations and sulfur in catchment soils (8) as well as biological lags such as those inherent in the re-establishment of fish and invertebrate populations (9). The level to which deposition must be reduced to restore a given ecosystem depends on how long one can wait for the responsesa lower deposition will give a quicker recovery. This time-dependent deposition goal is called the “target load” (10). Calculation of target loads necessitates the use of dynamic models, models that take into account the time lags in catchment and ecosystem response. In November 2007 the Coordination Centre for Effects (CCE) of the International Cooperative Programme for Modelling and Mapping (ICP M&M, part of the LRTAP Convention) issued a call for results from application of dynamic models to soils and surface waters. Several future scenarios of sulfur (S) and nitrogen (N) deposition were specified. In response to this call the Norwegians used the dynamic model MAGIC (Model of Acidification of Groundwater In Catchments) (11-13) to project future recovery of surface waters in Norway. Here we describe the procedures used, show estimates of future water quality and fish population status, and consider the implications for policy options for additional reductions in S and N emissions.

2. Materials and Methods 2.1. The MAGIC Model. The dynamic model MAGIC (version 777) was used to calculate future acidification status (as measured by acid neutralizing capacity, ANC) for Norwegian surface waters. ANC is defined as the equivalent sum of base cations (Ca, Mg, Na, K) minus the equivalent sum of strong acid anions (SO4, NO3, Cl). MAGIC (Model of Acidification of Groundwater In Catchments) is a lumped-parameter model of intermediate complexity, developed to predict the long-term effects of acidic deposition on soils and surface water chemistry (11-13). MAGIC consists of (1) a section in which the concentrations of major ions are assumed to be governed by simultaneous reactions involving sulfate adsorption, cation exchange, dissolution-precipitation-speciation of aluminum, and dissolution-speciation of inorganic and organic VOL. 44, NO. 14, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5345

carbon, and (2) a mass balance section in which the flux of major ions to and from the soil is assumed to be controlled by atmospheric inputs, chemical weathering inputs, net uptake in biomass, and losses to runoff. At the heart of MAGIC is the size of the pool of exchangeable base cations in the soil. As the fluxes to and from this pool change over time owing to changes in atmospheric deposition, the chemical equilibria between soil and soil solution shift to give changes in surface water chemistry. The degree and rate of change in surface water acidity thus depend both on flux factors and the inherent characteristics of the affected soils. The data required to run MAGIC must be spatially and temporally aggregated (“lumped”) to represent the whole catchment area and the time step of the model. Soil physiochemical parameters are aggregated vertically by the soil profile and spatially to account for the areas of the different soil types within a catchment. The model requires input data for various parameters of lake chemistry, soil chemistry, deposition chemistry, and hydrology. Cosby et al. (11) give details. The MAGIC model has been extensively applied and tested over the past 20 years at many sites and in many regions around the world (11). Overall, the model has proven to be robust, reliable, and useful in a variety of scientific and environmental management activities. Recently MAGIC was used to assess the recovery of European surface waters from acidification given a S and N deposition scenario of full implementation of current legislation (CLE scenario), i.e., the Gothenburg protocol and other agreed legislation (14). 2.2. Input Data. The lake chemistry and site data were taken from 1007 statistically selected lakes of the 1995 Norwegian national survey (15). The lakes were sampled in autumn 1995, and samples were analyzed for major ions. Lake and catchment characteristics were obtained from topographic maps and other national databases. Of these 1007 lakes, 1003 had all the necessary input data for chemistry and lake/catchment characteristics. Soil data for forested areas came from the NIJOS (Norwegian Institute for Soil and Forest Inventory, now part of The Norwegian Forest and Landscape Institute) national forest inventory on a 9 × 9 km grid basis. These were aggregated (arithmetic averages weighted by soil mass) to the 12 × 12 km critical load grid. Of the 1003 lakes, 345 were located in a grid cell where forest data were available (i.e., forest present); 658 lakes in grid cells had no forest data available (i.e., no forest present). Soil data for nonforested areas (mainly upland, mountains, and heathlands) came from various research projects (details given by ref 16. Annual net accumulation of base cations in forest biomass was assumed to be Ca 21 meq/m2/yr, Mg 4 meq/m2/yr, and K 4.5 meq/ m2/yr for forests and zero for nonforested areas. Forest management was assumed to not change over time. These values come from estimated annual average removal of cations in harvested timber (data from the national forest inventory) and average base cation concentrations in tree boles for the two major species Norway spruce (Picea abies L.) and Scots pine (Pinus sylvestris L.) (17). Deposition in the calibration year 1995 was estimated for each lake catchment from the measured lake chemistry and the deposition estimates for grid-average land cover provided by the CCE (18). The deposition estimates could not be used directly because these are not specific to each lake catchment. Factors such as orographic effects cause deposition fluxes to vary locally. First SO4 deposition was calculated using a massbalance approach. The sources for SO4 in lakewater were assumed to be comprised of (1) weathering of soil minerals, (2) deposition of seasalts, (3) natural background (preindustrial) deposition of nonmarine SO4, and (4) anthropogenic SO4, termed “excess” SO4. Seasalt SO4 was assumed to equal 0.103 Cl (ratio of these ions in seawater). The natural 5346

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 14, 2010

FIGURE 1. Relationship between the probability of a healthy population of brown trout (S. trutta) and ANCooa in 546 Norwegian lakes sampled in 1986 (modified from ref 24). background of excess SO4 of approximately 1 meq/m2/yr was included in the CCE data (18). SO4 from weathering was assumed zero in all cases where excess SO4 deposition calculated from the water chemistry and discharge was below 100 meq/m2/yr (in 1995). The threshold of 100 meq/m2/yr was chosen as this is the maximum excess SO4 deposition in Norway in 1995 (19). In cases where calculated excess SO4 deposition was above 100 meq/m2/yr, the weathering component was assumed to account for the remainder. Deposition of NOx and NHy in 1995 was then calculated from the anthropogenic SO4 (above) and the ionic ratios in the deposition data (NOx/SO4, NHy/SO4) as provided by the CCE. Deposition of Cl was assumed equal to the output flux at each lake (based on concentration in lakewater and discharge). The deposition of Na, Mg, Ca, K, and marine SO4 was then calculated from the deposition of Cl and the equivalent ratios of these ions in seawater (Ca/Cl ) 0.037, Mg/Cl ) 0.196, Na/Cl ) 0.856, K/Cl ) 0.018). Deposition sequences for the “historical” period 1880-2010 were specified by the CCE for the three components S, NOx, and NHy (18). The historical deposition values were normalized to the deposition values calculated for the year 1995. 2.3. Calibration. MAGIC was calibrated to each lake using an automatic optimizer that iteratively adjusted the values of a series of parameters such that the simulated lake chemistry for the calibration year 1995 matched the observed to within (2 µeq/L and the simulated percent soil base saturation matched the observed for each base cation to within (0.2 percentage points. For lake chemistry acid neutralizing capacity (ANC) was used to quantify lake acidification status. A total of 990 lakes were successfully calibrated. The calibrated files were then run with MAGIC to produce the data for the four evaluation years, 1980, 1990, 2000, and 2010. Most of the 17 lakes that did not calibrate had Na/Cl ratios lower than that of seawater, yet no discernible other source of Cl. Nitrogen dynamics in MAGIC 777 were turned off. Instead the net % retention of nitrogen input in the catchment plus lake was set to the 1995 value and assumed to be constant through the entire historical and future periods. This assumption was justified on the basis of 30-year records of deposition and lake chemistry in Norway, which show no systematic trend over time in net % retention of N (20). 2.4. Scenarios. Two future scenarios of S and N deposition were used: the CLE scenario (current legislation, the Gothenburg protocol plus other agreed national legislation) and the MFR scenario (maximum technically feasible reduction). Both were assumed to be fully implemented by the year 2020 (linear change from 2010 to 2020). The resulting lake ANC was then estimated

FIGURE 2. Sulfur and nitrogen deposition in southernmost Norway (EMEP grid ij5157). Shown are estimated deposition normalized to 1995 given historical nonmarine S and N emissions in Europe modeled with the EMEP model (54) and projected to 2020 given full implementation of the CLE or MFR protocols (lines) and measured 1977-2007 at Birkenes (triangles) (data from EMEP/NILU) (19). Birkenes lies within the EMEP grid ij5157.

FIGURE 3. Maps of S deposition in Norway for 1980 (mean 1978-82), 2000 (mean 1997-2002), and 2020 (CLE scenario). Data from NILU (19). by running the calibrated MAGIC files for each lake with the of Ca, Cl, Mg, and SO4 (µeq/L). The chemical parameters given future deposition scenarios. The results were evaluated were chosen because they reflect the major sources of solutes for the years 2020, 2030, 2050, and 2100. in freshwaters not impacted by local pollution (chemical 2.5. Extrapolation to Critical Load Grids. Norway has weathering in soil Ca, Mg; seasalt inputs Cl, Mg, SO4; acid deposition SO4). Each of the 2304 grid cells was assigned a earlier submitted critical load (CL) estimates to the CCE for MAGIC modeled lake. each of 2304 squares in a grid covering the entire country 2.6. Criterion for Biological Damage. Brown trout (21). The grid was defined as 0.125° latitude and 0.25° population status was selected as the biological indicator of longitude. Each square is approximately 12 × 12 km. adverse effects of acidification. A 1986 survey of lake water Henriksen and Buan (21) compiled representative water chemistry and brown trout populations in 827 lakes in Norway chemistry data for each of the grid squares from existing showed a clear relationship between ANC and population data sources and used these to calculate CL for surface waters. status (23). Damaged populations were defined in the survey The data reflect water chemistry of 1995. as those described as “sparse” or “extinct”. At ANC < 20 µeq/L Our results for dynamic modeling of the 990 lakes were the probability of damaged population exceeded 5%. This extrapolated to the 2304 squares in the CL grid. This was defined as the ANClimit which then was used to calculate extrapolation was done by means of an analogue matching the CL (21). procedure in the “MAGIC library” held at IVL Swedish Lydersen et al. (24) modified the ANC-brown trout status Environment Institute (22). This procedure takes the water relationship to take into account the acidifying influence of chemistry data for each of the 2304 grid squares and finds natural organic acids (Figure 1). They defined a modified the closest match in the 990 lakes calibrated by MAGIC. The ANC that incorporates strong organic acids, ANCoaa ) ANC match was the lake with the minimum Euclidian distance of - (10.2/3) × DOC, where DOC is the concentration of the following parameters: geographical distance (latitude, dissolved organic carbon (mgC/L), and ANC is in µeq/L. longitude), annual runoff (m/yr), and concentrations in 1995 VOL. 44, NO. 14, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5347

FIGURE 4. Observed and simulated concentrations of SO4 and ANC at Lake Lille Hovvatn, southernmost Norway. Left-hand panels: MAGIC simulations driven by observed annual variations in deposition amount and chemistry (from ref 29); right-hand panels: MAGIC simulations driven by the regional historical and future scenarios of S and N deposition as specified by the CCE (this study).

FIGURE 5. Maps of ANCooa as simulated by MAGIC for the years 1980, 2000, and 2010. The classes represent ANCooa levels for which water quality is suitable for brown trout populations. Red: >75% probability of damaged population (ANCooa < -36.7 µeq/L); orange: 25-75% probability of damaged population (-36.9 to -19.8 µeq/L); light green: 2.5-25% probability of damaged population (-19.8 to 0 µeq/L); light blue: 0.5-2.5% probability of damaged population (0 to 12 µeq/L); blue: < 0.5% probability of damaged population (>12 µeq/L). Larssen et al. (25) calculated a logistic regression relationship between ANCoaa and fish health, and obtained an ANCooa limit of 15 µeq/L for brown trout. For ANCooa above this level the probability of a healthy population is >95% (Figure 1).

3. Results and Discussion As a result of the implementation of S emission reductions in Europe, deposition of anthropogenically derived S in Norway has decreased by nearly 80% from peak levels around 1980 to levels that are very close to those expected for the year 2010 with full implementation of the Gothenburg protocol (19) (Figure 2). Emissions and deposition of N, on the other hand, have declined by only about 20%. Although the role of N in freshwater acidification has increased relative to that of S, the long-term data show no systematic change in the fraction of incoming N retained in catchment soils and the lakes themselves (20). The highest levels of S deposition are in southernmost Norway (Figure 3), and it is here that the most severe acidification of freshwaters has occurred. Nationwide surveys in 1986 of water chemistry and fish populations in lakes showed that in the 1980s acid deposition exceeded the critical load in surface waters in about 30% of the country (26, 27). The decline in S deposition since 1980 led to gradual improvements in lake water chemistry (28). 5348

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 14, 2010

The MAGIC simulations illustrate this long-term rise and fall of acidification. One of the 990 lakes in the 1995 data set is Lake Lille Hovvatn, a now-barren former brown trout lake in southernmost Norway. Lake Lille Hovvatn has been the subject of research and monitoring since 1974, and serves as the untreated reference lake for several research projects (29). Hindar and Wright (29)calibrated MAGIC to Lake Lille Hovvatn using the observed annual deposition chemistry to drive the model, and they obtained good agreement between measured and simulated water chemistry (Figure 4). The long-term simulations here for Lake Lille Hovvatn show the acidification and partial recovery of ANC (Figure 4) in response to long-term changes in S deposition (Figure 2). The MAGIC simulations suggest that there will remain substantial acidification of lakes in southernmost Norway in the year 2010 despite full implementation of the CLE scenario (Figure 5). Water quality in about 20% of the country will still be insufficient to support self-reproducing brown trout populations. Brown trout is the most widely used biological indicator of acidification status in Norwegian waters as this species is acidsensitive, is the major inland sports fish species, and is widely distributed throughout Norway. Other fish species and organism groups are also acid sensitive and show adverse effects. Atlantic salmon (Salmo trutta L.) populations have been severely

FIGURE 6. Maps of ANCooa as simulated by MAGIC for the years 2020, 2050, and 2100 given the CLE and MRF scenarios of future S and N deposition. See legend to Figure 5 for details. impacted by acidification in southern Norway; salmon is extinct in 7 major salmon rivers (1). The governments of Norway and Sweden conduct large-scale programs of liming to restore salmon and brown trout populations (30). Acidification also affects most other organism groups, and among indicator species are Daphnia species in the zooplankton (31) and the mayfly Bætis rhodani in the zoobenthos (32). These species all have similar requirements for water quality with respect to acidification parameters such as ANC, pH, and dissolved inorganic aluminum. Recovery of brown trout populations will thus generally mean that conditions should be satisfactory for other indicator organisms. A case study of recovery from acidification at Lake Saudlandsvatn, southernmost Norway, showed that brown trout, Daphnia, and Bætis all have responded to increased pH and ANC over the last 10 years (33). Biological recovery, however, is fraught with various “bottlenecks” such as migration rates and vulnerability to short-term acid episodes (9). The MAGIC simulations indicate that there will be further improvements in lake water chemistry in the future. Under the CLE scenario ANCooa is expected to increase modestly to the year 2020 and then very slowly throughout the rest of the century, with about 18% of the country still affected in the year 2100 (Figure 6). Under the more aggressive MFR scenario, greater improvements in ANCooa are expected, and in the year 2100 about 12% will still be affected. If the policy goal is to achieve healthy fish populations as a restoration target, then the greater and earlier the reduction in acid deposition, the faster this goal can be met. The situation projected for the CLE scenario for the year 2100 is quite similar to that projected for the MFR scenario for the year 2020. This is because replenishment of base cations to the soil exchange complex by chemical weathering is a slow process; the lower the sulfur deposition the greater the rate

of replenishment. The policy choice is thus to incur the additional costs of further cuts in emissions now to achieve rapid restoration of acidified lakes, or to do nothing beyond current legislation and wait many decades for recovery of water chemistry. This illustrates the dynamic aspect of the acidification and recovery process. Modeling studies on groups of acid-sensitive lakes elsewhere in Europe indicate that in most areas implementation of the CLE scenario will greatly diminish the acidification problem, but acidification with adverse biological effects will continue to occur for some lakes in southern Norway, southern Sweden, the Tatra Mountains of Slovakia, the Italian Alps, and parts of the UK (14, 34). MAGIC and other biogeochemical models of acidification have been extensively applied also in North America, and used to project the future recovery of acid-sensitive salmon rivers in Nova Scotia (35), lakes in eastern Canada (36), trout streams in the Appalachian Mountains (37), and lakes in the Adirondack Mountains (38) given various future scenarios for S and N emissions. All these studies worked with groups of streams or lakes; our study here from Norway projects the results to waters over an entire contiguous area, in our case to the 12 × 12 km grid covering the entire country. The UNECE has completed the review of the Gothenburg protocol, and the LRTAP Convention is now beginning work with possible revision of the protocol. For acidification effects the Gothenburg protocol was based on the static critical load concept. For the revision it is now possible to include dynamic aspects as well, such as these MAGIC results for acidification of surface waters in Norway. Any such modeling entails a degree of uncertainty, which can arise from many sources. There is uncertainty in (1) the measured data used (water chemistry, soil chemistry, catchment characteristics), (2) estimated data such as historical VOL. 44, NO. 14, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5349

S and N deposition, (3) the methods used to aggregate up to lake catchments or grid squares, (3) assumptions behind the model, and not least, (4) confounding factors which might operate in the future, such as climate change. Confidence in MAGIC simulations is enhanced by applications at ecosystemscale manipulation experiments such as the Norwegian RAIN project (39) and sites with long-term data sets such as calibrated catchments (40, 41) and lakes (29). Perhaps the greatest source of uncertainty in the dynamic modeling results here is associated with future confounding factors such as the long-term fate of the large store of N in catchment soils, climate change, and changes in concntrations of dissoved organic carbon (DOC). The retention and loss of N is a factor likely to be important for the recovery of acidified waters in the future. The lakecatchments in this study retain most of the N deposited from the atmosphere. Clearly this cannot go on forever; at some point the ecosystem must become “nitrogen saturated” (42) and begin to leak a greater and greater fraction of N deposition. Synoptic studies indicate that N-rich forests and heathland ecosystems leach a larger fraction of incoming N relative to N-poor ecosystems (43, 44). Yet long-term whole ecosystem experiments with N addition at N-poor sites such as that conducted at Gårdsjo¨n, Sweden, indicate that N saturation is a very slow process (45), and long-term monitoring data from surface waters in Europe and North America do not suggest that widespread increased N leaching has occurred over the past 20 years (46, 47). The modeling results presented here assume no longterm change in climate, yet over the coming decades it is probable that temperature, precipitation, and frequency of extreme weather events will change. These factors can affect surface water chemistry, directly as well as indirectly through processes in catchment vegetation and soil. Climate change can exacerbate or ameliorate the effects of acid deposition (48). For example, increased temperature may cause increased decomposition of soil organic matter with release of nitrate (NO3) to surface waters thus decreasing ANC. On the other hand, increased temperature might cause faster forest growth with increased uptake of N from soil solution and thus decreased leaching of NO3 to surface water. Wright et al. (49)evaluated the relative sensitivity of several of these possible climate-induced effects on MAGIC simulated water chemistry at 14 sites in Europe and North America. Quantification of these various effects and in particular their interactions is difficult at best, as there are few experimental data at appropriate temporal and spatial scales available. Several studies have shown links between climatic fluctuations and decadal trends in surface water chemistry (50, 51), and these relationships have been used with MAGIC to simulate future water chemistry given joint scenarios of acid deposition and climate change (52). The widespread recovery of acidified lakes and streams in the 1990s has in many cases coincided with increased concentrations of dissolved organic carbon (DOC), and the cause is believed to be decreased concentrations of strongacid anions such as SO4 in soil solution (53). In Norway the upward trend in DOC concentrations in lakes has flattened out since the early 2000s as has the downward trend in SO4 concentrations (20). Changes in DOC concentrations potentially will affect the projections of future fish population status, as the parameter ANCooa is a function in part of the DOC concentration. The effect is relatively minor, however, compared to the large overall changes in ANCooa. A change in DOC of 3 mgC/L will cause only a 1 µeq/L change in ANCooa. The 1995 survey showed that lakes in Norway have median DOC concentration of about 2 mgC/l and only 4% have TOC > 10 mgC/l (15). In 78 of the lakes measured annually the increase since the late 1980s has been about 15-35% (20). Thus for lake Lille Hovvatn the increase in DOC over the 5350

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 14, 2010

period 1980-2004 has been from about 3 to 5 mgC/L which will cause a decrease in ANCooa of about 0.6 µeq/L, while the overall increase in ANC during this period has been from about -60 to 0 µeq/L (Figure 4) (29). There have been great achievements in reducing European emissions of S (and to a lesser extent N), such that acidified freshwaters in Norway and elsewhere in Europe have shown major recovery during the past 20 years. But as this dynamic modeling work shows, even with full implementation of the CLE scenario, freshwaters in large areas of southernmost Norway will continue to suffer from acidification during the 21st century. Water quality will not be sufficient to allow viable populations of fish and other organisms. The implications for policymakers are two-fold: further reductions in S and N emissions are required, and in the meantime freshwaters or their terrestrial catchments will have to be limed to create water quality conditions suitable for fish and other organisms.

Acknowledgments We thank Filip Moldan at the Swedish Environmental Research Institute IVL for conducting the lake matching, and Maximilian Posch at the CCE for expert assistance. This work was supported in part by the Norwegian State Pollution Control Authority and the Norwegian Institute for Water Research.

Literature Cited (1) Overrein, L.; Seip, H. M.; Tollan, A. Acid precipitation - Effects on forest and fish. Final report of the SNSF project 1972-1980; SNSF Project, Norwegian Forest Research Institute: Ås, Norway, 1980. (2) Hesthagen, T.; Sevaldrud, I. H.; Berger, H. M. Assessment of damage to fish populations in Norwegian lakes due to acidification. Ambio 1999, 28, 112–117. (3) Tammi, J.; Appelberg, M.; Beier, U.; Hesthagen, T.; Lappalainen, A.; Rask, M. Fish status survey of Nordic lakes: Effects of acidification, eutrophication and stocking activity on present fish species composition. Ambio 2003, 32, 98–105. (4) Pleijel, H., Ed. Transboundary Air Pollution. Scientific Understanding and Environmental Policy in Europe; Studentlitteratur: Lund, Sweden, 2007. (5) Sliggers, J., Kakebeeke, W., Eds. Clearing the air: 25 years of the Convention on Long-range Transboundary Air Pollution; Economic Commission for Europe, United Nations, Sales No. E.04.11.E.20: Geneva, 2004. (6) Nilsson, J.; Grennfelt, P. Critical loads for Sulphur and Nitrogen; Nordic Council of Ministers: Copenhagen, 1988. (7) Amann, M.; Cofala, J.; Heyes, C.; Klimont, Z.; Scho¨pp, W. The RAINS model: a tool for assessing regional emission control strategies in Europe. In Pollution Atmospherique Spe´cial 20e Anniversaire de la Convention sur la pollution atmosphe´rique transfrontalie`re a` longue distance; Paris, 1999; pp 41-63. (8) Reuss, J. O.; Johnson, D. W. Acid Deposition and the Acidification of Soils and Waters; Springer Verlag: New York, 1986; Vol. 59. (9) Yan, N. D.; Leung, B.; Keller, W.; Arnott, S. E.; Gunn, J. M.; Raddum, G. G. Developing conceptual frameworks for the recovery of aquatic biota from acidification. Ambio 2003, 32, 165–169. (10) Jenkins, A.; Cosby, B. J.; Ferrier, R. C.; Larssen, T.; Posch, M. Assessing emission reduction targets with dynamic models: deriving target load functions for use in integrated assessment. Hydrol. Earth Syst. Sci. 2003, 7, 609–617. (11) Cosby, B. J.; Ferrier, R. C.; Jenkins, A.; Wright, R. F. Modelling the effects of acid deposition: refinements, adjustments and inclusion of nitrogen dynamics in the MAGIC model. Hydrol. Earth Syst. Sci. 2001, 5, 499–518. (12) Cosby, B. J.; Hornberger, G. M.; Galloway, J. N.; Wright, R. F. Modelling the effects of acid deposition: assessment of a lumped parameter model of soil water and streamwater chemistry. Water Resour. Res. 1985, 21, 51–63. (13) Cosby, B. J.; Wright, R. F.; Hornberger, G. M.; Galloway, J. N. Modelling the effects of acid deposition: estimation of long term water quality responses in a small forested catchment. Water Resour. Res. 1985, 21, 1591–1601. (14) Wright, R. F.; Larssen, T.; Camarero, L.; Cosby, B. J.; Ferrier, R. C.; Helliwell, R. C.; Forsius, M.; Jenkins, A.; Kopa´cek, J.; Majer, V.; Moldan, F.; Posch, M.; Rogora, M.; Scho¨pp, W. Recovery of

(15)

(16)

(17) (18) (19)

(20)

(21)

(22) (23)

(24)

(25)

(26) (27) (28)

(29) (30) (31) (32) (33)

(34)

(35)

(36)

acidified European surface waters. Environ. Sci. Technol. 2005, 39, 64A–72A. Skjelkvåle, B. L.; Henriksen, A.; Faafeng, B.; Fjeld, E.; Traaen, T. S.; Lien, L.; Lydersen, E.; Buan, A. K. Regional innsjøundersøkelse 1995. En vannkjemisk undersøkelse av 1500 norske innsjøer; Statens forurensningstilsyn, 1996. Larssen, T.; Cosby, B. J.; Høgåsen, T.; Lund, E.; Wright, R. F. Dynamic modelling of acidification of Norwegian surface waters; Report 5705-2008; Norwegian Institute for Water Research: Oslo, Norway, 2008. Sverdrup, H.; de Vries, W.; Henriksen, A. Mapping Critical Loads; Miljørapport 1990:14; Nordic Council of Ministers: Copenhagen, 1990. Hettelingh, J.-P.; Posch, M.; Slootweg, J. CCE Status Report 2008: Critical Load, Dynamic Modelling and Impact Assessment in Europe; Coordination Centre for Effects: Bilthoven, Netherlands, 2008. Aas, W.; Solberg, S.; Berg, T.; Manø, S.; Yttri, K. E. Monitoring of Long-Range Transported Air and Precipitation-Atmospheric deposition, 2007; Report 1033/2008; State Pollution Control Authority: Oslo, Norway, 2008. SFT. The Norwegian monitoring programme for long-range transported air pollutants. Annual report - Effects 2008; TA-2546/ 2009; The Norwegian Pollution Control Authority (SFT): Oslo, Norway, 2009. Henriksen, A.; Buan, A. K. Critical Loads and exceedances of Critical Loads for Surface Water, Soils and Vegetation in Norway (In Norwegian); Norwegian Institute for Water Research: Oslo, Norway, 2000. IVL. Description of the MAGIC library (In Swedish), 2007 (http:// www.ivl.se/affar/grundl_miljos/proj/magic/bibliotek.asp). Lien, L.; Raddum, G. G.; Fjellheim, A.; Henriksen, A. A critical limit for acid neutralizing capacity in Norwegian surface waters, based on new analyses of fish and invertebrate responses. Sci. Total Environ. 1996, 177, 173–193. Lydersen, E.; Larssen, T.; Fjeld, E. The influence of total organic carbon (TOC) on the relationship between acid neutralizing capacity (ANC) and fish status in Norwegian lakes. Sci. Total Environ. 2004, 362, 63–69. Larssen, T.; Huseby, R. B.; Cosby, B. J.; Høst, G.; Høgåsen, T.; Aldrin, M. Forecasting acidification effects using a Bayesian calibration and uncertainty propagation approach. Environ. Sci. Technol. 2006, 40, 7841–7847. Henriksen, A.; Lien, L.; Rosseland, B. O.; Traaen, T. S.; Sevaldrud, I. Lake acidification in Norway - present and predicted fish status. Ambio 1989, 18, 314–321. Henriksen, A.; Lien, L.; Traaen, T. S.; Sevaldrud, I.; Brakke, D. F. Lake acidification in Norway - present and predicted chemical status. Ambio 1988, 17, 259–266. Skjelkvåle, B. L.; Wright, R. F.; Henriksen, A. Norwegian lakes show widespread recovery from acidification: results of national surveys of lakewater chemistry 1986-1997. Hydrol. Earth Syst. Sci. 1998, 2, 555–562. Hindar, A.; Wright, R. F. Long-term records and modelling of acidification, recovery, and liming at Lake Hovvatn, Norway. Can. J. Fish. Aquat. Sci. 2005, 62, 2620–2631. Henrikson, L.; Hindar, A.; Thornelof, E. Freshwater liming. Water, Air Soil Pollut. 1995, 85, 131–142. Havens, K. E.; Yan, N. D.; Keller, W. Lake Acidification - Effects On Crustacean Zooplankton Populations. Environ. Sci. Technol. 1993, 27, 1621–1624. Raddum, G. G.; Skjelkvåle, B. L. Critical limits of acidification to invertebrates in different regions of Europe. Water, Air Soil Pollut. 1995, 85, 475–480. Hesthagen, T.; Fjellheim, A.; Schartau, A. K.; Wright, R. F.; Saksgård, R.; Rosseland, B. O. Chemical and biological recovery of Lake Saudlandsvatn, a highly acidified lake in southernmost Norway, in response to decreased acid deposition. Can. J. Fish. Aquat. Sci. In review. Jenkins, A.; Camarero, L.; Cosby, B. J.; Ferrier, R. C.; Forsius, M.; Helliwell, R. C.; Kopa´cek, J.; Majer, V.; Moldan, F.; Posch, M.; Rogora, M.; Scho¨pp, W.; Wright, R. F. A modelling assessment of acidification and recovery of European surface waters. Hydrol. Earth Syst. Sci. 2003, 7, 447–455. Clair, T. A.; Dennis, I. F.; Amiro, P. G.; Cosby, B. J. Past and future chemistry changes in acidified Nova Scotian Atlantic salmon (Salmo salar) rivers: a dynamic modeling approach. Can. J. Fish. Aquat. Sci. 2004, 61, 1965–1975. Clair, T. A.; Aherne, J.; Dennis, I. F.; Gilliss, M.; Couture, S.; McNicol, D.; Weeber, R.; Dillon, P. J.; Keller, W.; Jeffries, D. S.; Page, S.; Timoffee, K.; Cosby, B. J. Past and future changes to acidified eastern Canadian lakes: A geochemical modeling approach. Appl. Geochem. 2007, 22, 1189–1195.

(37) Sullivan, T. J.; Cosby, B. J.; Webb, J. R.; Dennis, R. L.; Bulger, A. J.; Deviney, F. A. Streamwater acid-base chemistry and critical loads of atmospheric sulfur deposition in Shenandoah National Park, Virginia. Environ. Monit. Assess. 2008, 137, 85–99. (38) Sullivan, T. J.; Cosby, B. J.; Herlihy, A. T.; Driscoll, C. T.; Fernandez, I. J.; McDonnell, T. C.; Boylen, C. W.; NierzwickiBauer, S. A.; Snyder, K. U. Assessment of the extent to which intensively-studied lakes are representative of the Adirondack region and response to future changes in acidic deposition. Water, Air Soil Pollut. 2007, 185, 279–291. (39) Wright, R. F.; Cosby, B. J.; Flaten, M. B.; Reuss, J. O. Evaluation of an acidification model with data from manipulated catchments in Norway. Nature 1990, 343, 53–55. (40) Larssen, T. Model prognoses for future acidification recovery of surface waters in Norway using long-term monitoring data. Environ. Sci. Technol. 2005, 39, 7970–7979. (41) Larssen, T.; Cosby, B. J.; Høgåsen, T. Uncertainties in predictions of surface water acidity using the MAGIC model. Water, Air Soil Pollut.: Focus 2004, 4, 137. (42) Aber, J. D.; Nadelhoffer, K. J.; Steudler, P.; Melillo, J. Nitrogen saturation in northern forest ecosystems. Bioscience 1989, 39, 378–386. (43) MacDonald, J. A.; Dise, N. B.; Matzner, E.; Armbruster, M.; Gundersen, P.; Forsius, M. Nitrogen input together with ecosystem nitrogen enrichment predict nitrate leaching from European forests. Global Change Biol. 2002, 8, 1028–1033. (44) Evans, C. D.; Reynolds, B.; Jenkins, A.; Helliwell, R.; Curtis, C. J.; Goodale, C. L. Evidence that soil carbon pool determines susceptibility of semi-natural ecosystems to elevated nitrogen leaching. Ecosystems 2006, 9, 453–462. (45) Moldan, F.; Kjønaas, O. J.; Stuanes, A. O.; Wright, R. F. Increased nitrogen in runoff and soil following thirteen years of experimentally-increased nitrogen deposition to a coniferous-forested catchment at Gårdsjo¨n, Sweden. Environ. Pollut. 2006, 144, 610– 620. (46) Wright, R. F.; Alewell, C.; Cullen, J.; Evans, C. D.; Marchetto, A.; Moldan, F.; Prechtel, A.; Rogora, M. Trends in nitrogen deposition and leaching in acid-sensitive streams in Europe. Hydrol. Earth Syst. Sci. 2001, 5, 299–310. (47) Skjelkvåle, B. L.; Olendrzynski, K.; Stoddard, J. L.; Tarrason, L.; Traaen, T. S.; Tørseth, K.; Windjusveen, S.; Wright, R. F. Assessment of trends and leaching of nitrogen at ICP Waters sites (Europe and North America); Norwegian Institute for Water Research: Oslo, Norway, 2001. (48) Wright, R. F.; Bishop, K.; Dillon, P. J.; Erlandsson, M.; Evans, C. D.; Hardekopf, D. W.; Helliwell, R.; Hrusˇka, J.; Hutchins, M.; Kaste, Ø.; Kopa´cˇek, J.; Kra´m, P.; Laudon, H.; Moldan, F.; Rogora, M.; Sjøeng, A. M. S.; de Wit, H. A. Interaction of climate change and acid deposition. In Climate Change impacts on Freshwater Ecosystems - direct effects and interactions with other stresses; Kernan, M., Moss, B., Battarbee, R. W., Eds.; Wiley-Blackwell, In press. (49) Wright, R. F.; Aherne, J.; Bishop, K.; Camarero, L.; Cosby, B. J.; Erlandsson, M.; Evans, C. D.; Forsius, M.; Hardekopf, D. W.; Helliwell, R.; Hrusˇka, J.; Jenkins, A.; Kopa´cˇek, J.; Moldan, F.; Posch, M.; Rogora, M. Modelling the effect of climate change on recovery of acidified freshwaters: relative sensitivity of individual processes in the MAGIC model. Sci. Total Environ. 2006, 365, 154–166. (50) Evans, C. D.; Monteith, D. T.; Cooper, D. M. Long-term increases in surface water dissolved organic carbon: Observations, possible causes and environmental impacts. Environ. Pollut. 2005, 137, 55–71. (51) 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, 2617–2630. (52) Evans, C. D. Modeling the effects of climate change on an acidic upland stream. Biogeochemistry 2005, 74, 21–46. (53) Monteith, D. T.; Stoddard, J. L.; Evans, C. D.; de Wit, H. A.; Forsius, M.; Høgåsen, T.; Wilander, A.; Skjelkvåle, B. L.; Jeffries, D. S.; Vuorenmaa, J.; Keller, B.; Kopa´cek, J.; Vesely, J. Dissolved organic carbon trends resulting from changes in atmospheric deposition chemistry. Nature 2007, 450, 537–540. (54) Scho¨pp, W.; Posch, M.; Mylona, S.; Johansson, M. Long-term development of acid deposition (1880-2030) in sensitive freshwater regions in Europe. Hydrol. Earth Syst. Sci. 2003, 7, 436–446.

ES100792M

VOL. 44, NO. 14, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5351