Critical Loads and Critical Limit Values edited by Hannu Raitio and Tuire Kilponen Proceedings of the Finnish-Swedish Environmental Conference October 27—28, 1994 Vaasa, Finland Metsäntutkimuslaitoksen tiedonantoja 513 The Finnish Forest Research Institute. Research Papers 513 Critical Loads and Critical Limit Values edited by Hannu Raitio and Tuire Kilponen Proceedings of the Finnish-Swedish Environmental Conference October 27-28, 1994 Vaasa, Finland Metsäntutkimuslaitoksen tiedonantoja 513 The Finnish Forest Research Institute. Research Papers 513 2 Contents Foreword 4 Sirkka Haut o järvi Opening of the Conference 6 Markku Luoma Opening of the Conference 9 Jean-Paul Hettelingh, Peter A.M. de Smet and Robert J. Downing The critical loads concept and critical loads in Europe 11 Michael Ashmore Critical levels for forestry in Europe 27 Maximilian Posch Allocation of critical loads to S and N deposition 36 Wim de Vries, Gert J. Reinds and Maximilian Posch Assessment of the forested area in Europe at risk under various acid deposition scenarios 48 Staffan Westerlund The concepts of critical loads and environmental quality standards 66 Kevin Bishop Critical load models for the regional and local scale: a cautionary scientific perspective 74 Heikki Niininen Critical load and critical limit values from the standpoint of the energy producer 77 Mats Nebaeus Critical load and critical limit values from the stand point of the local administration 81 Reino Lammi Critical loads and critical limit values from the standpoint of the wood and paper industry 87 3 Kalervo Mäkelä and Matti Perttilä Critical loads to the Gulf of Bothnia 90 Arne Henriksen, Juha Kämäri, Maximilian Posch, Martin Forsius, Anders Wilander and Tatyana Moiseenko Critical loads for surface waters in Scandinavia and the Kola region 97 Harald Sverdrup, Per Warfvinge and Kaj Rosen Critical loads of acidity and nitrogen for Swedish forest ecosystems, and the relationship to soil weathering 109 Alf Björklund Critical load and special characteristics of the Quark area of Finland 139 Hakan Staaf Critical loads/levels as a tool for implementing national strategies for pollution control 151 Lennart Mattson Critical load and critical limit values from the standpoint of the administration 157 Marketta Karhu Critical loads and critical limit values from the point of view of local government district environmental authorities 162 Hannu Ilvesniemi and Michael Starr The use of the Ca/Al ratio in the calculation of critical loads for forest soils 165 Remarks of the Moderators Kaj Rosen Some aspects on the concept of critical loads A moderators conclusions 177 Pirjo-Liisa Nurmela From the viewpoint of a government environmental officer attending a conference dealing with critical loads 181 Kari Kuusiniemi Critical loads and critical limit values The legal point of view 188 4 Foreword The Council of the Strait of the Gulf of Bothnia (CSGB) is a regional body of co operation under the jurisdiction of the Nordic Council of Ministers. In Finland, CSGB's functioning concentrates on the region covered by the County of Vaasa and in Sweden on the region covered by the County of Västerbotten and the municipality of Örnsköldsvik. One of the current areas of emphasis for CSGB are various environmental projects. One such project is concerned with the interaction between the state of health of spruce forests and the local conditions in the vicinity of the Strait of the Gulf of Bothnia. The idea of arranging a Finnish-Swedish environmental conference was the result of an initiative put forward by The Council of the Strait of the Gulf of Bothnia in 1991. In the beginning of 1993 the idea crystallised in connection with the work of "Granskogens hälsotillständ"-project, a forest project. One reason for choosing critical loads and critical limit values as the subject was that included among the goals of the forest project were to examine more closely the concept of critical loads in regard to the Strait region and to enhanceco-operation between researchers and between researchers, and environmental authorities and to promote the dissemination of research results. Thus, the Finnish-Swedish environmental conference partly fulfils the objectives of the forest project. The Organizing Committee of the conference was appointed in autumn of 1993. In addition having members from the Council of the Strait of the Gulf of Bothnia and the Finnish Forest Research Institute's Parkano Research Station, the Committee has representatives of the County of Vaasa (Finland), Vaasa Water and Environment Authority (Finland), Vaasa University (Finland), Agricultural University of Sweden and the University of Umeä (Sweden), and the County of Västerbotten (Sweden). Critical load is an internal property of the ecosystem under assessment. It is highly geographically tied, ecosystem-related property. Different regions and different ecosystems differ in sensitivity; i.e. are capable of withstanding different loads. Defining critical loads thus presupposes the availability of precise information on the effect that deposition has on different ecosystems. The precise determining of critical loads and critical limit values helps us to focus emission-reducing actions in places where this is most urgent. Critical loads and critical limit values are, therefore, matters of importance for all concerned researchers, authorities and pe ople involved in the formulation of legislative measures. 5 The Organising Committee has selected the following themes for the seminar: 1. Critical loads of the environment and critical limit values from the point of view of aquatic and forest ecosystems. 2. State of current knowledge and research needs. 3. Critical loads and critical limit values from the poits of view of the environmental authorities, the research, the legislation and the industry. 4. Environmental legislation: legislative possibilities for controlling environmental pollution and 5. Critical loads and special characteristics of the Strait of the Gulf of Bothnia. Altogether twenty presentations of invited speakers and nineteen posters were presented during the two conference days. The Finnish Forest Research Institute has assumed the responsibility for the publishing of the conference proceedings. The editors wish to thank the authors for the cooperation which has enabled us to publish this volume. The conference is funded by The Council of the Strait of the Gulf of Bothnia, the Academy of Finland, Swedish National Environment Protection Board, Imatran Voima Oy and Wisaforest Oy Ab. Thanks are due to the above-mentioned organizations for their support. The main responsibility for the practical conference arrangements has been borne by the Finnish Forest Research Institute's Parkano Research Station and the Western Finland Institute for Economic Research of the Vaasa University. Hannu Raitio Tuire Kilponen 6 Opening of the Conference Sirkka Hautojärvi Secretary General Ministry of the Environment P.O. Box 399 FIN-00121 Helsinki, Finland Ladies and Gentlemen, It is my pleasure to be able to address to you here today in this Finnish-Swedish Conference where views on critical loads and levels will be exchanged. I am pleased to notice many outstanding speakers on the programme. This gives a promise of a very fruitful Conference here in Vaasa. Unfortunately Ms. Pietikäinen, our minister of the Environment was unable to be present here herself due to her other obliging commitments. She would have liked to be here today. She asked me to extend her very best wishes to the Conference. Being aware that there are far better experts in this field present here than I am, I will like to stress in my opening statement four points. Firstly the importance of international cooperation in environmental issues, because only through common efforts we can make progress in the path towards sustainable development. Secondly, we should be ambitious enough to put our good intentions into action. And thirdly, we have to recognize the need for further research and information on issues affecting sustainable development. The level of knowledge is far for being high enough. However, more and more knowledge is being gathered which gives us a reliable basis for action. My last point is that good exchange of knowledge and information is needed which gives me another opportunity to stress the importance of the occasions like the one here in Vaasa. Transboundary air pollution has long been the most serious environmental problem in Europe. In Finland, acid rain is affecting our lakes and forests. Finland together with 26 other countries signed the second sulphur protocol in Oslo in June this year. The new sulphur protocol marked the start of the second stage of the European agreements on emission reduction. It meant a continuation of the 1985 protocol on a 30 per cent reduction in sulphur emissions, adopted in Helsinki. 7 It is no doubt we can say that air pollution control has made a progress during the decade after the Executive Body Meeting in Helsinki. This is apparent in both practical efforts to reduce emissions, as well as, in planning. Cooperation in research carried out since the Helsinki sulphur protocol, has made it possible to base the recent protocol on the concept of critical loads. In fact, planning for emission reductions is now based on estimates of nature's tolerance in various parts of Europe, rather than equal reduction percentages for each country. In general, the more you have knowledge, the more specific you can be. According to the newly signed protocol, the long-term goal of the signatories is to prevent sulphur depositions from exceeding the critical loads calculated for each area by the experts. At present, critical loads are exceeded by a substantial margin. The negotiations operated on the premise that the first step towards this goal would be to reduce the excess in question by at least 60 per cent by the year 2000. The result of the negotiations binds the parties to reduce their emissions in accordance with the reduction percentages recorded in the protocol. The target mentioned above will not be fully reached. Finland, for its part, will attempt to achieve a major reduction in sulphur depositions, and in fact the emission reduction announced by Finland will be 80 per cent of the 1980 amount by the year 2000. With respect to Finnish ecosystems, the target reduction percentages in the new protocol are not yet satisfactory. Finland expressed some disappointment with the low emission reduction targets during the negotiations, and appealed to many governments on behalf of better achievements. Still we had good reasons to sign the protocol. In many ways the protocol can be regarded as progressive. First of all, basing the protocol on critical loads marks a significant progress. Likewise, inclusion of individual emission targets in a legally binding document must be regarded as a step forward. Thirdly, the fact that an Implementation Committee was established to review the implementation of this protocol and compliance by the parties is an important achievement. As the last point I would like to pick up an important element of the protocol. It is the possibility to implement emission reduction through joint efforts between countries. The Finnish Ministry of the Environment has granted FIM 185 million (USD 35 million) in support for investment in environmental protection and energy savings since 1991. Support has also been provided for 190 technical aid projects. The fact that the protocol makes it possible to equate such actions with emission reductions at home is important for countries like Finland, who have long used the least expensive means of reducing their own emissions. Future acidification in Finland depends largely on efforts to reduce emissions in areas near Finland. The Baltic states did not attend the european protocol negotiations. Moreover, only a 38 per cent reduction target has been entered in the 8 annex to the protocol for Russian Federation. At the Geneva negotiations Russia did not want separate emission reduction figures for its western regions, close to the Finnish boarder, although it did express its readiness to make bilateral agreements with neighbouring countries. Finland and Russia have in fact undertaken negotiations on a bilateral agreement and the experts on critical loads have already met several times. The first treaty negotiations took place in Helsinki at the beginning of May and the second meeting is expected to be in early December. The negotiators have set a bilateral framework agreement for air pollution control as their goal by the beginning of 1996 at the latest. Finland and Estonia have signed an agreement concerning cooperation in the area of air pollution control already in July 1993. The sulphur and nitrogen emission of Estonia have reduced by 50 per cent (because they have stopped to export electricity produced in the oil shale fired power plants of Narva). There is a further need for emission reductions due to acid deposition, which exceeds the critical loads in South Eastern Finland and possibly Southern parts of Estonia. Nordic Working Group on Air and Sea Environment has desided to continue its efforts in catalyzing the scientific and technical work needed to support the further negotiations in Geneve and also the work inside the European Union. We need sufficient information on issues affecting sustainable development, that is, on the interdependence of nature, human well-being and economic prosperity. By producing this information, science and research can provide a sound foundation for new solutions. In the future, the aims and demands of environmental protection will increasingly be greared towards production based on the principles of sustainable development. The focal point will shift from cleaning up discharges to saving raw materials and energy resources and to preventing environmental damage. Natural scientists now face a new, more difficult task of preparing the ground for nitrogen protocol of the second stage. Apart from the critical loads of acidic deposition, they should investigate eutrophication and the direct impact of nitrogen dioxide and ozone on vegetation and human health. This is a very complicated multi-component and multi-effect exercice, where also cost-efficiency aspect must be taken into account. Still we must be optimistic that through common efforts and firm determination we can make progress in the path towards sustainable development. With these words, I wish you all succesfull seminar on critical loads and levels. 9 Opening of the Conference Markku Luoma Department Secretary Vaasa Provincial State Office P.O. Box 200 FIN-65101 Vaasa, Finland Ladies and gentlemen, Cooperation between the province of Vaasa and Västerbotten goes back a long way. "The Council of the Quark" was founded in 1972. (To those unfamiliar with this concept, the Quark is the narrow part of the Gulf of Bothnia). Initially our cooperation primarily focused on trade, culture and advancement of communication between our two provinces. Our cooperation in the area of environmental conserva tion began in earnest in 1989 with the delegation of duties between the various regional authorities and officials involved in this work. This gave rise to an efficient network of collaboration in the sphere of environmental protection. Our first official project was a survey of the current state of the aquatic environ ment in the region of the Quark. Our cooperation then progressed from individual projects to official collaboration between our two provinces and their local authorities. A steering committee was set up to coordinate this work, chaired alternately by the Heads of the Environmental Departments at the Provincial State Offices of Vaasa and Västerbotten for a term of two years at a time. The steering committee formulated a strategy on environmental cooperation, which was then ratified by the Quark Council. Its chief aim is to provide general guideli nes for environmental cooperation and to ensure that new projects are launched on a regular basis. The strategy also outlines the need and potential for future funding. Roughly 1 —2 million Finnish marks are channelled into cooperative projects every year. A project set up to establish critical loads on eutrophied lakes in the Quark region was completed in 1993. A three-year project on the impact of airborne loads on forest environments is currently in its final year. 10 We have considered it useful to hold seminars in conjunction with these projects. These seminars have usually been open to interested parties in the regions of Vaasa and Västerbotten. The seminar now to be opened. "Critical loads and limits" is based on findings from earlier projects. We have nevertheless decided to broaden the scope of this seminar to enable participants from other regions to attend. 11 The critical loads concept and critical loads in Europe Jean-Paul Hettelingh, Peter A.M. de Smet and Robert J. Downing Coordination Center for Effects National Institute of Public Health and Environmental Protection (RIVM) P.O. Box 1 3720 BA Biltlioven, The Netherlands Abstract The sensitivity of ecosystems has been included as a basis for the revision of the second sulfur protocol of the Convention on Long Range Transboundary Air Pollution of the United Nations Economic Commission of Europe. Critical loads, ie, the maximum allowable acid deposition which does not increase the probability of damage to forest soils and surface waters in Europe, have been computed and mapped. The critical load is an indicator of sustainability of ecosystems against acidification in the sense that risk of damage is reduced whenever acid deposition does not exceed the critical load. The reduction of the excess of sulfur deposition over critical loads by 60 % (60 % gap closure) has emerged as a starting point for policy negotiations leading to the second sulfur protocol. The critical load concept has thus lead to a sulfur protocol that protects about 81 %, 86 % and 90 % of terrestrial and surface water systems in 2000, 2005 and 2010 respectively. In addition, the total European area in which sulfur deposition exceeds critical loads by more than 500 eq ha"' yr" 1 will be reduced from about 19 % in 1980 to about 1 % in 2010. Introduction The critical load is defined as "the highest deposition of a compound that will not cause chemical changes leading to harmful effects on ecosystem structure and function" (Nilsson and Grennfelt 1988). In this paper critical loads for acidity and sulfur are emphasized, because of their importance for the recent UN/ECE sulfur reduction agreement. Critical loads for acidity and sulfur have been computed and mapped 1 in Europe (Hettelingh et al. 1991, 1992 a, Downing et al. 1993) for forest soils and surface waters. The aim of the critical load approach is to achieve ' Critical loads have been corrected for base cation deposition, which buffers the acidifying effect of sulfur, and for base cation uptake, which increases the acidifying effect. The term critical sulfur deposition, denoted CD(S), is then used to characterize corrected critical loads. 12 protection of ecosystems by sufficiently reducing emissions of acidifying compounds such as sulphur dioxide (SO 2 ), nitrogen oxide (NO x ) and ammonia (NH3 ). Emission reductions are considered sufficient when the resulting acidic deposition does not exceed critical loads. Critical loads have been used together with integrated models such as the Regional Acidification Information and Simulation model (RAINS) to assess a variety of emission reduction strategies in relation to abatement costs and environmental protection. Such assessments are used in negotiations to develop protocols on the reduction of sulfur dioxide and nitrogen oxides within the framework of the Convention on the Long-Range Transboundary Air Pollution (LRTAP) of the United Nations Economic Commission for Europe (UN/ECE). In June 1994, a second protocol for reducing sulfur emissions was signed by 30 countries. The starting point for negotiating national emission reductions was to diminish by 60 % the excess of 1990 sulfur deposition over critical loads for sulfur. This scenario, which was assessed by the RAINS-model, became better known as the "60 % gap closure" scenario. It was the first time that negotiations included environmental effects, by means of the critical load concept, in addition to technical and economic considerations of emission abatements. This paper presents an overview of the critical load concept, the assessment of critical loads in Europe and their usage in the negotiation of the second sulfur protocol. The application of the critical load concept to a multitude of pollutants constitutes the final section of this paper. Critical loads: calculation and mapping The critical loads concept A critical load is an indicator for sustainability of an ecosystem in the sense that it provides a value for maximum allowable load of a pollutant at which risk of damage to an ecosystem is reduced. When critical loads are exceeded, certain physical and chemical properties of an ecosystem may change to an extent where the risk of damage is increased. The chemical changes and effects induced by pollutants vary with pollutants, and the kind of exposure. For example, sulfur dioxide deposition cause one effect, ie, soil and surface water acidification. However, the deposition of nitrogen oxides and ammonia causes acidification as well as eutrophication. The critical load concept has currently been restricted to sulfur, which is the pollutant for which a UN/ECE protocol was first to be revised. The critical load concept is a methodology according to which critical loads in Europe are used as a criterion to assess whether emission reduction strategies are sufficient, taking national cost of emission reduction into account (Hettelingh et al. 13 1992 a). Figure 1 (dotted line) shows the objective of the iterative procedure is to avoid that sulfur deposition exceeds critical loads at the lowest possible national costs. Since reducing emissions to reach critical loads throughout Europe was found to be infeasible the objective of meeting critical loads was altered during policy negotiations. First an objective by which excess should be reduced (gap closure) is formulated. Then, the required emission reductions at minimum European costs are computed. Finally, the actual excess in different regions in Europe is expressed in terms of the extent to which the national area of ecosystems is protected. This procedure is repeated by changing the objective finally leading to the identification by groups under UN/ECE of the 60 % gap closure being an appropriate starting point of emission reduction negotiations. The 60 % gap closure strategy aims to reduce the 1990 exceedance by 60 % by 2000, which means that parts of Europe will continue to have depositions that exceed critical loads. Since critical load exceedance may lead to "harmful effects", the question of when these harmful effects will occur becomes relevant. The answer to this question can be investigated by using dynamic soil models, which simulate the temporal changes in the balance of geochemical processes in a soil receiving acid deposition. The procedure of using dynamic models is also part of the critical load concept as illustrated in Figure 1 (solid line). The application of dynamic models is also useful to assess recovery time of an ecosystem once emissions have been sufficiently reduced. The dynamic part of the critical load concept has been initiated (Hettelingh et al. 1992b, Hettelingh and Posch 1993, Bleeker et al. 1994) but has not yet been part of the integrated assessment of emission reduction strategies. However it is expected that the dynamic part of the critical load concept will become increasingly important as indications of target years for establishing required emission reduction vary over Europe. This is especially the case as UN/ECE LRTAP negotiations move towards reductions in nitrogen oxide emissions, which have to be viewed in terms of the total acidifying effect, ie, including the agreed sulfur emission reductions. Critical load computation Critical loads have been mapped (Hettelingh et al. 1991, Downing et al. 1993) by the Coordination Center for Effects (CCE). Critical loads were computed by national experts using local data or by the CCE on the basis of existing European databases when national contributions were not available. The computation method is based on the Steady State Mass Balance Method (SSMB) which is extensively described elsewhere (Sverdrup et al. 1990, Hettelingh and de Vries 1991, Hettelingh et al. 1991). SSMB assumes a steady state of chemical compounds in soils and surface waters. The critical load thus becomes an indicator of sustainability of ecosystems against acidification. The ratio of base cations to aluminum, and the 14 Figure 1. The critical load concept consisting of: (1) iterating required emission reductions to meet critical loads and costs (short circle; dotted line), and (2) computing time horizons before ecosystem damage (continued critical load exceedance) or recovery (whenever exceedance ceases to exist). The computation of time horizons is represented by the solid line (long circle). (Source: Downing et al. 1994, p. 2) 15 aluminum concentration, are used as indicators for steady state of geochemical processes. By assigning established critical values to these indicators (eg, the concentration of aluminum in soil solution should not exceed 0.2 meq/1 and the base cation to aluminum ratio should not be less than one), it is possible to compute the allowable acidification for each ecosystem. The critical load for sulfur was mapped similarly. An extensive overview of critical values for the base cation to aluminum ratio for a large variety of plants and trees can be found in Sverdrup and Warfvinge (1993). The European map of critical loads for sulfur used in designing the second sulfur protocol is shown in Figure 2. Figure 2 shows the distribution of critical load ranges over EMEP grid cells. Each range reflects a computed critical load which, when not exceeded by sulfur deposition, results in a protection of about 95 % of the natural forests and/or surface waters. It can be seen that most sensitive ecosystems are located in Northern Europe and scattered areas in Western and Eastern Europe, where a sulfur deposition exceeding the range of 0 to 200 eq ha" 1 yr" 1 leads to an increased risk of damage. In Central-West and Eastern Europe large areas have natural systems exposed to risk of damage when sulfur deposition exceeds a range of 200-500 eq ha" 1 yr"'. Critical loads exceeding 1000 eq ha" 1 yr" 1 are found in southern European natural systems that are least sensitive to sulfur deposition in comparison to other parts of Europe. Other protection levels may be chosen instead of the so-called 5-percentile critical load that decreases the risk of damage to 95 % of the terrestrial and surface water systems in each grid cell. Since a 150x150 km 2 EMEP grid cell may contain many ecosystems of varying sizes, it is possible to compute cumulative distributions of critical loads. The 5-percentile critical load value has been chosen for the policy analysis. Calculating critical loads of acidity did not provide European policy makers with sufficient tools for the evaluation of required sulfur emission reductions. The reason is that UN/ECE protocols concentrate on distinctive acidifying compounds (sulphur or nitrogen), rather than on acidity as a whole. Therefore it was necessary to develop methods to apportion the critical load of acidity between the acidifying share of sulfur and the acidifying share of nitrogen. From a scientific point it is doubtful whether discrimination between sulfur and nitrogen is justified since an ecosystem is indifferent to whether acid stress is caused by sulfur or by nitrogen. The critical load for sulfur has been derived from the critical load of acidity by assigning the share of sulfur to the ratio of sulphur to net nitrogen deposition 16 Figure 2. Ranges of sulfur critical loads in Europe after correction for base cation deposition and base cation uptake. (Source: Coordination Center for Effects, RIVM). (Hettelingh et al. 1991, p. 40). This subdivision was obtained using the following assumptions: (a) The share of the present sulfur deposition in total acid deposition is used as proxy of the part of the critical load of actual acidity which can be attributed to sulfur. 17 (b) The share of the present nitrogen deposition in total acid deposition contributes to acidification only when it is not taken up or immobilized by the ecosystem. In other words, nitrogen is acidifying when the ecosystem is unable to use nitrogen as a nutrient. This fraction holds when the total present load of nitrogen exceeds the nitrogen uptake capacity. The sulfur fraction is assumed to be equal to unity when nitrogen uptake exceeds the total nitrogen load. In that particular case the critical load of acidity becomes equal to the critical load of sulfur. Currently, the use of the sulfur fraction has been improved by a methodology in which the excess of the critical load for acidity and nutrient nitrogen is assessed simultaneously as described in Posch et al. (1993) and elsewhere in these proceedings (Posch 1994). The RAINS model The Regional Acidification, INformation and Simulation model (RAINS; see also Alcamo et al. 1989) is used to assess cost-effective emission reduction alternatives. RAINS consists of: (a) an energy/emission module that computes national emissions as function of energy combustion (SO 2, NO x ) and agricultural practices (NH 3 ); (b) an atmospheric module computing acidic depositions in 150x150 km 2 grid cells using the EMEP transport model2 (Tuovinen et al. 1994), and (c) an effects module comparing acidic depositions to critical loads produced by CCE. RAINS allows the computation of deposition levels given an energy/emission pattern or for optimization (Amann 1989, 1991). In the optimization mode, minimum cost emission reductions over Europe as a whole can be computed subject to any particular set of deposition levels prescribed over 150x150 km 2 cells covering Europe (EMEP). For the preparation of the second sulfur protocol the RAINS model was used to assess sulfur reduction strategies for the years 2000, 2005 and 2010 in comparison to sulfur emissions in 1990, the base year. The optimization mode was used to find cost-optimal sulfur emission reductions for example with respect to the 60 % gap closure scenario. Results of the critical load concept This section summarizes the results of the second sulfur protocol. The sulfur dioxide emission reductions were agreed on a country by country basis, resulting 2 The European Monitoring and Evaluation Programme of the UN/ECE provides governments with information and data on depositions, concentrations, long-range transport and transboundary fluxes of air pollutants. 18 in a decrease of the European total from about 41 million tonnes in 1990 to 26 million tonnes in 2010. Figure 3 shows sulfur dioxide emission levels from 1980 to 2010. Emissions in 2010 are projected to be about 53 % less than 1980, the base year for the first sulfur protocol. The result of the emission reductions is a decrease of the areas where sulfur deposition exceeds critical loads. T his can be seen by comparing Figures 4, 5, and 6 which show the excess of critical loads in 1980, 1990, and 2010 respectively. Figure 4 shows an excess of more 2000 eq ha" 1 yr" 1 covering an area from the east of the United Kingdom, northern France, Germany, Poland, the Czech and Slovak Republics to Hungary and northern Italy. The excess of critical loads is lower in areas in the periphery of Europe. Areas of non-excess are found in the south and east of Europe. In 1990 (Figure 5) the area of highest excess is reduced to the center of Europe. The area where no excess of sulfur deposition occurs covers larger parts of southern Europe including most of France. The result of the second sulfur protocol is displayed in Figure 6, which shows the projected excess of sulfur deposition over critical loads for sulfur in 2010. An excess above 2000 eq ha" 1 yr" 1 does not occur anywhere. However, still a rather large area of Europe is subject to risk of damage with areas suffering from an excess ranging from 200 to 2000 eq ha" 1 yr" 1 . The levels of ecosystem protection from 1980 to 2010 are shown in Figure 7 and in Table 1. In 1980, slightly more than half (54.0 %) of the terrestrial and surface waters systems in Europe received levels of sulfur deposition below critical loads. By 1990, the results of the first sulfur protocol are clear: now 69 percent of European ecosystems are not exceeded, and the percentage of ecosystems with the greatest exceedance (more than 2000 acid equivalents per hectare per year) has been reduced from 4.7 to 2.2 percent. Analysis of future deposition patterns using the RAINS model shows that 90 percent of ecosystems will be protected by 2010, when the second sulphur protocol is fully implemented. Although Figures 3 and 4 show that progress has been made with respect to increasing the quality of the environment, it is necessary to remain cautious with the interpretation of the results for a number of reasons. First, these results do not include the effect of nitrogen deposition. The area subject to excess remains extensive when nitrogen is included in the analysis. Secondly, because the magnitude of an excess, however small, does not give any indication about the damage to be expected. Depending on the ability of natural systems to cope with acidity, even a small excess can still cause considerable risk of damage rather soon. The actual risk and the pace by which damage may occur can be forecasted using 19 dynamic models, which are increasingly being used. Finally, multiple stresses due to the interaction of different pollutants indirectly (through deposition), or directly (through air concentrations) complicates the evaluation of policies geared towards the reduction of one isolated pollutant, such as sulfur. Figure 3. European emissions of sulfur dioxide as reported to UN/ECE for 1980, and 1990 and as projected according to the agreements of the second sulfur protocol. 20 Figure 4. The excess of critical loads by sulfur deposition in 1980. (Source: Coordination Center for Effects, RIVM.) 21 Figure 5. The excess of critical loads by sulfur deposition in 1990. (Source Coordination Center for Effects, RIVM.) 22 Figure 6. The excess of critical loads by sulfur deposition in 2010 according to the emission reductions agreement of the second sulfur protocol. (Source: Coordination Center for Effects, RIVM.) 23 Figure 7. Ranges of the excess of sulfur deposition over critical loads for sulfur. In 1980 about 46 % of terrestrial and surface water ecosystems is subject to risk compared to about 10 % in 2010. The increase in 2000, 2005 and 2010 of protected ecosystems is according to the sulfur emission reductions as agreed in the sulfur protocol. (Source: Coordination Center for Effects, RIVM). Table 1. Percentage of ecosystem area exceeded by critical load of sulfur. E@S^HHHi^H@9BHHflEHHHEflHHHD3HHHI^9 24 Final remarks and outlook The critical load approach has successfully been applied to support European and national policies to reduce sulfur emissions. The method is now being extended to allow for trade-off between sulfur and nitrogen deposition without exceeding the critical load for acidity and eutrophication (Posch et al. 1993). It can be shown that without additional reduction ofNO x and NH 3 emissions, most ecosystems in Europe are still at risk. The reduction of sulfur emissions alone is not sufficient, as demonstrated using dynamic modeling approaches that forecast the state of acidification in case of continued excess of the critical load of acidity by both sulfur and nitrogen deposition. However, deposition alone is not necessarily the only cause of increased risk for damage. Concentrations of sulfur dioxide, nitrogen oxides and ozone have shown to cause direct damage such as yellowing of tree leaves and decrease of crop yield. Critical levels, ie, concentrations below which no damage is expected, have been formulated for ozone (Fuhrer et al. 1993), sulfur dioxide and nitrogen oxides (UN ECE, 1993). The result is that the assessment of environmental benefits of emission reductions should include both levels and loads simultaneously. As more pollutants, each causing specific effects, are addressed in developing environmental policies, it is likely that the critical load approach as applied to sulfur will have to be extended. It becomes more appropriate to use the term critical threshold approach, which is an iterative procedure of emission reductions of various pollutants, with the objective of not exceeding thresholds which risk ecosystem sustainability. In trying to establish targets to ensure sustainability of natural environmental systems, the complication of interacting pollutants and various exposure routes does not provide policy makers with simple solutions. For example, nitrogen oxide in the air may lead to an excess of critical levels for ozone through interaction with volatile organic compounds (VOCs). In addition, the reduction of nitrogen oxide emissions with the objective to meet critical loads for acidity and nutrient nitrogen will also have to be paired to policies of reducing VOCs in some parts of Europe if ozone levels are to be kept to acceptable limits. The meaning of a critical threshold approach in this particular case could be that policies are developed with the aim to avoid (a) the excess of the critical load of acidity by sulfur and nitrogen simultaneously, (b) the excess of the critical load of nutrient nitrogen by nitrogen oxides and ammonia, (c) the excess of the critical level of ozone by the interaction of nitrogen oxides with volatile organic compounds, (d) the excess of critical levels of sulfur dioxide and nitrogen oxides. Interactions with other pollutants are likely (eg, acidity with heavy metals) and complexity may increase even more as regional environmental quality is affected by global change (eg, alteration of land cover) in addition to stress caused by pollutants. 25 To ensure the derivation of mutually consistent critical thresholds, it is important to investigate the synergistic relationships between direct and indirect effects of pollutant concentrations and depositions respectively. Therefore, a reductionist approach towards the establishment of critical thresholds for each pollutant under consideration should be avoided as much as possible. Continued scientific research is required to extend the methodology currently developed in treating sulfur and nitrogen simultaneously with respect to critical loads of acidity and nutrient nitrogen. This extension has to include ozone and critical levels on the short term to accommodate UN/ECE requirements. Ultimately, science may also consider relationships between acidification levels and leaching of toxics as well as interactions with climate change. Policy making frameworks, such as the LRTAP Convention are increasingly recognizing the importance of not isolating one pollutant as policy target. The preparation of the next UN/ECE protocol on reducing nitrogen, includes investigations of combined effect with ozone and ammonia. The RIVM Coordination Center for Effects can contribute to preparing further modeling and mapping methodologies to accommodate the scientific support of multi-pollutant multi-effect protocols. References Alcamo, J., Shaw, R. & Hordijk, L. (eds.) 1990. The RAINS model of acidification: science and strategies in Europe. Kluwer Academic Publishers, Dordrecht, the Netherlands. Amann, M. 1989. Energy use, emissions, and abatement costs. In: Alcamo et al. 1989. - 1991. The efficient multinational allocation of emissions reduction measures for reduction of acid deposition: Application Example for Austria. Academic Dissertation, University of Karlsruhe (in German). Bleeker, A., Posch, M., Forsius, M. & Kämäri, J. 1994. Calibration of the SMART acidification model to selected IM-catchments. In: International Cooperative Programme on Integrated Monitoring of Air Pollution Effects on Ecosystems, Annual Synoptic Report 1994. National Board of Waters and the Environment, Helsinki. Downing, R.J., Hettelingh, J.-P., de Smet, P.A.M. (eds.) 1993. Calculation and mapping of critical loads in Europe: Status Report 1993. Coordination Center for Effects, National Institute of Public Health and Environmental Protection, Bilthoven. ISBN 90-6960-047-1. Fuhrer, J. & Achermann, B. (eds.), 1993. Critical Levels for Ozone: a UN-ECE workshop report. Swiss Federal Research Station for Agricultural Chemistry and Environmental Hygiene, No. 16, Liebefeld-Bern, Switzerland. 26 Hettelingh, J.-P. & Posch, M. 1993. Critical loads and a dynamic assessment of ecosystem recovery, in: J.Grasman and G.van Straten (eds.), Predictability and nonlinear modelling in natural sciences and economics, Kluwer Academic Publishers, Dordrecht, The Netherlands. , Downing, R.J. & de Smet, P.A.M. 1992 a. The critical load concept for the control of acidification. In: T. Schneider (ed.). Acidification research: evaluation and policy applications. Elsevier Studies in Environmental Science 50, Amsterdam, pp. 161-174. , Gardner, R.H. & Hordijk, L. 1992b. A Statistical Approach to the Regional Use of Critical Loads. Environ. Pollut. 77:177-183. - & de Vries, W. 1991. Mapping Vademecum. Coordination Center for Effects, National Institute of Public Health and Environmental Protection, RIVM Rep. No. 259101002, Bilthoven, The Netherlands. - , Downing, R.J. and de Smet, P.A.M. (eds.). 1991. Mapping Critical Loads for Europe. Coordination Center for Effects, National Institute of Public Health and Environmental Protection, Rept. 259101001, Bilthoven, The Netherlands. Nilsson, J. & Grennfelt, P. (eds.). 1988. Critical Loads for Sulphur and Nitrogen, Report from a workshop held at Skokloster, 19-24 March 1988. Nordic Council of Ministers, Miljorapport 1988:15, Copenhagen. Posch, M., Hettelingh, J.-P., Sverdrup, H.U., Bull, K. & de Vries, W. 1993. Guidelines for the computation and mapping of critical loads and exceedances of sulphur and nitrogen in Europe. In: Downing et al., 1993. Posch, M. 1994. Allocation of critical acidification load to S and/or N deposition. Proceedings Finnish-Swedish environmental conference: critical load and critical limit values, Vaasa, Finland. Sverdrup, H., de Vries, W. & Henriksen, A. 1990. Mapping Critical Loads: A Guidance Manual to Criteria, Calculation, Data Collection and Mapping, Nordic Council of Ministers, Miljorapport 1990:14, Copenhagen. & Warfvinge, P. 1993. Effect of soil acidification on growth of trees, grasses and herbs as expressed by the (Ca+ Mg + K)/Al ratio. University of Lund Report, Lund, Sweden. Tuovinen, J.-P., Barrett, K. & Styve, H. 1994. Transboundary Acidifying Pollution in Europe: Calculated fields and budgets 1985-93. EMEP MSC-W Report 1/94. Norwegian Meteorological Institute, Oslo. UN ECE, 1993. Manual, on Methodologies and Criteria for Mapping Critical Levels/ Loads and Geographic Areas where they are exceeded. Convention on Long-Range Transboundary Air Pollution, Task Force on Mapping, Geneva. Federal Environmental Agency, Texte 25/93, Berlin. 27 Critical levels for forestry in Europe Michael Ashmore Centre for Environmental Technology Imperial College of Science Technology and Medicine 48 Prince's Gardens London SW7 2PE, U.K. Introduction Over the past decade, there has been increased interest in the development of critical loads for forest soils to prevent long-term acidification or eutrophication. These critical loads were central to the recent negotiation of a new protocol on the reduction of sulphur emissions within the framework of the UNECE Convention on Long Range Transboundary Air Pollutuion. Critical loads refer to the total deposition of sulphur, nitrogen or acidity; in contrast, critical levels refer to the direct effects of atmospheric pollutants on vegetation. Critical levels for vegetation were first set at a workshop held at Bad Harzburg in 1988 (UNECE 1988), where the following definition was adopted:- "CRITICAL LEVELS means the concentrations of pollutants in the atmosphere above which direct adverse effects on receptors such as plants, ecosystems or materials, may occur according to present knowledge." The direct adverse effects of air pollutants, especially sulphur dioxide, on forestry have long been recognised, and historically these effects have been substantial in many parts of Europe. In addition, there is increasing evidence of direct adverse effects of ozone on forests in several areas of Europe. Indeed, air quality standards to protect forestry were being promulgated, for example for sulphur dioxide by lUFRO, before the concept of critical levels was developed. Although critical loads and levels have been developed within the context of long range transboundary air pollutants, several of the pollutants for which critical levels have been formulated are strongly influenced by local sources, and thus their concentrations can be influenced by local pollution control measures. Thus the critical level concept and the values currently set may be of value in the management of local and regional air quality. However, care must be taken to ensure that the critical level values are relevent to the specific local situation. 28 This paper provides a brief overview of critical levels in the context of forestry. In considers some of the practical difficulties in setting critical level values for forests, and also the issue of what precisely is meant by an adverse effect. It then describes the development of the current critical levels values for forestry and the basis on which they have been set. Finally it briefly considers future research needs, emphasising the need to integrate the assessment of critical levels and loads. How are critical levels set for forestry ? The definition of the critical level given above raises a number of interesting questions which are central to defining meaningful values. Firstly, what is meant by the term 'adverse effect'? In managed arable agriculture, there is a clear economic meaning to this term. However, forestry serves a number of important functions, besides its value as an economic crop. This raises questions about how to assess adverse effects, for example, on the conservation value of forests. Often this is related to plant species of the woodland floor, to insect diversity, or to the presence of epiphytic lichens, as much as to the vitality of the trees themselves. Currently, critical levels are based on experiments or observations on the trees, but these other conservation criteria may need further consideration; indeed a lower critical level of S0 2 than that for trees has now been set to protect sensitive epiphytic lichens. Thus the values of critical levels need to be related to a specific criterion used to define the 'adverse effect'. There are also important scientific issues in defining the term adverse effect. It is well known that air pollutants can have a range of effects on plants, ranging from subtle biochemical changes to visible injury and growth reductions. Although it is generally agreed that effects on growth offer the most robust basis for defining an adverse effect, two particular problems should be identified. Firstly, in the case of nitrogen deposition, for example, the initial effect is often a growth stimulation, i.e. a beneficial rather than an adverse effect. However, continued deposition can lead to nutrient imbalances, or an enhanced sensitivity to secondary stresses. Here, clearly, growth is not an adequate basis on which to define an adverse effect. Secondly, subtle changes in leaf structure, in translocation patterns, in enzyme activities etc. may in many circumstances be insufficient in themselves to lead to adverse effects on growth or tree vitality. However, it is now realised that pollutants may alter the sensitivity of plants to a range of biotic and abiotic stresses, often as a result of precisely these subtle changes in structure or function. Here, whether a particular physiological change is adverse or not depends on the presence of other stress factors. Hence, the definition of 'adverse effect' requires careful scientific judgement and depends on the particlar pollutant and receptor under consideration. 29 While the definition of an adverse effect raises conceptual difficulties, there are often much more difficult practical problems in defining critical level values. Ideally, one would have well-defined dose-response relationships over a range of concentrations, including those causing no measurable effect (Bull 1991), obtained under realistic conditions for a variety of tree species. In practice, such relationships come from one of two sources. Firstly, dose-response relationships can be derived from experimental studies; for practical reasons, however, these can only be carried out over relatively short periods on seedlings or young trees, and there are grave difficulties in transposing their results to long-term effects on forest ecosystems. Secondly, dose-response relationships can be derived from field observations; these can allow long-term effects on mature forests to be evaluated, usually along a transect away from a large pollution source. However, in such situations, there is frequently spatial variation in the levels of not just one, but of several pollutants, and this makes it difficult to use the data to define critical levels for individual pollutants. Furthermore, close to large point sources, the frequency of episodes of very high concentrations is much greater than at other locations with the same long term mean concentration, and this can modify the dose-response relationships. Confounding of such field effects with climatic and other factors is also possible, while for secondary pollutants such as ozone it is difficult to find appropriate spatial gradients in concentration. In the case of critical loads for forest soils, the values are derived on the basis of models, or mass-balance equations, which describe the chemical behaviour of the soil system. In contrast, the reliance in setting critical levels on either experimental data or field observations is a serious limitation, since it makes it difficult to extrapolate beyond those species and situations for which actual data are available. Nevertheless, progress has been made in developing mechanistic models of forest response to air pollutants, and these could play an important future role, for example in predicting mature forest responses from data on young trees, or in assessing how air pollutants may interact with other stress factors to infuence forest response (Ashmore et al. 1990). Besides these problems, there is a lack of data on the response to air pollutants for many species. This is particularly noticeable for Mediterranean species, about which there is a dearth of information (Ashmore et al. 1990), but it is also true of many species of central and northern Europe which are of conservation or aesthetic value, rather than of economic significance. Hence any assessment of critical levels is usually based only on data for a very limited number of species, of which the most sensitive is normally used, since this will be the species showing an adverse effect at the lowest concentration. Ideally, different critical levels might be set for 30 different species, just as critical loads are set for different soils, but this is difficult on the basis of our limited current knowledge. Similarly, different critical levels might be set for different soil conditions and different climates since these are known to modify dose-response relationships. In practice this is also difficult, since these factors have not been adequately investigated. It is vital that all these uncertainties are taken into account when applying the critical levels described below in any policy context. The values have been set 'according to current knowledge' and this knowledge is very limited. It is also important to consider, when applying a critical level in a regional or local context, whether the critical level is appropriate for the particular situation and the particular policy concerns. The critical levels described below should only be taken as general guidelines and not as precise 'no effect' standards. This discussion clearly indicates the need for further research to provide a sounder foundation for the critical levels set for forestry, and to allow the critical level concept to be developed to allow distinctions to be made between the sensitivity of different forests, based on species composition, soils, climate and other factors. More detailed reviews of these research needs can be found in the background papers for the Egham and Bern workshops at which the current critical levels were set (Ashmore and Wilson 1994, Fuhrer and Achermann 1994). Values of critical levels The values of critical levels for forestry first recommended at the Bad Harzburg meeting (UNECE 1988) are summarised in Table 1. These critical levels were reassessed at a further workshop at Egham in 1992 (Ashmore and Wilson 1994); for ozone, an additional workshop was held in Berne in November 1993 (Fuhrer and Achermann 1994) to provide a more substantive basis to the ideas developed at the Egham workshop. The reports of these workshops provide a detailed scientific justification for the critical levels proposed, which are summarised here. Although short-term critical levels were set at the Bad Harzburg meeting, many of these were removed at the Egham workshop, since it was felt that effects of air pollutants on the growth and vitality of forestry, and indeed other types of vegetation, were largely determined by long-term exposure, rather than isolated incidents of moderate concentrations. Thus this discussion will focus on these long term critical levels. 31 Table 1. Comparison of long-term critical levels for forestry proposed at the Bad Harzburg, Egham and Bern workshops. * This critical level applies where annual accumulated temperature above 5 °C is less than 1000 degree days. In the case of sulphur dioxide, the critical levels set at Bad Harzburg were based on a considerable body of field observation and experimental data from Europe. The modifications made at Egham were largely based on the known interaction between sulphur dioxide and low temperature. The precise modification of the critical level was derived from a model developed by Makela et al. (1987) based on field observations of Materna in Czecholovakia, showing greater sensitivity to sulphur dioxide at higher elevations. This is also consistent with the observation of effects of sulphur dioxide at low concentrations in Finland. The critical levels were also referenced to winter, as well as annual means, for evergreen species, in view of the known increase in sensitivity over winter. The Egham workshop also considered effects on lichens and bryophytes, and set a new critical level of 10 fig m 3 to protect sensitive epiphytic lichens. Although this critical level is not relevent to forest production, it is significant in terms of protecting the conservation value of forest ecosystems. Pollutant Bad Harzburg Egham/Bern so 2 20 (ig m" 3 (annual mean) 20 fig m" 3 (annual or winter mean) 15 |ig m" 3 (annual or winter mean)* NO x 30 jig m" 3 (annual mean as NO,) 30 jag m" 3 (annual mean as NO x ) NH 3 100 |ig m 3 (monthly mean) 23 \xg m" 3 (monthly mean) 8 (j.g m" 3 (annual mean) Acid Mist - 1 p.g S m 3 (sulphate particulate) where cloud cover 10 % Oxone 25 nl l" 1 (seasonal 7-hour mean) 10000 nl r'.h cumulative exposure above 40 nl 1"' over 6 months 32 The critical levels for sulphur dioxide can be compared with other international guidelines. The WHO have set a value of an annual mean of 30 |ig m' 3 (WHO 1987), which is higher than any current critical level for forestry, but this guideline is likely to revised shortly. lUFRO in the early 1980s set two annual mean values: 50 p,g m" 3 to allow full production at most sites, and 25 p.g m" 3 to allow full production at extreme sites in the mountains or in boreal zones (Wentzel 1983). These values now seem a little high given current knowledge. For nitrogen oxides and ammonia, the alterations to critical levels made at Egham were not primarily associated with effects on forestry. The critical level for ammonia is derived from data on sensitive heathland species, and it is possible that the appropriate critical level for forest species might be somewhat higher. A new critical level for direct effects of wet deposition was introduced at the Egham workshop. This was largely based on a review of experimental data showing direct effects of acid mist on a range of physiological parameters, and on growth and visible injury, in tree species, especially conifers (Cape 1993). This important effect on forests, which are frequently exposed to high levels of deposition of polluted cloud and mist, had previously been ignored both in the critical levels and the critical loads context. The critical level, which was based on experimental data demonstrating effects on leaf surfaces, was set at 150 |i,M l" 1 sulphate as an annual mean. However, since long-term data on sulphate concentrations in mist or cloud are not readily available, a procedure based on defining an equivalent particulate sulphate concentration of 1 (ig S m" 3 was developed, with the critical level only applying in areas with frequencies of cloud cover above 10 %. It should be noted that this critical level is based on mists with low base cation contents, and it is not applicable in those parts of Europe (e.g. the Mediterranean region and much of the Alps), where base cation levels in wet deposition are relatively high. In the case of ozone, the critical levels defined at Bad Harzburg were largely based on American data, since there was very little relevent data for Europe. However, there has been a substanital accumulataion of data on ozone effects on European species over the last decade, and this has provided the basis for the new critical level now set for ozone. The first important change, at the Egham workshop, was to develop the concept of cumulative exposure above a threshold concentration as the basis for defining the critical level; this is based on the observation that exposure patterns giving the same mean, but a higher frequency of episodes of high concentrations, caused greater impacts on vegetation. The exposure index is calculated by summing the differences between hourly mean concentrations and a threshold concnetration, only for those hours in which the threshold concentration is exceeded. 33 Analysis of experimental data on tree seedlings, as well as on agricultural crops, at the subsequent Bern workshop confirmed that a threshold concentration of 40 nl 1"' was appropriate. Calculating exposure over 6 months, including all 24 hours in the day, the critical level for forests was set at 10000 nl l'.h above 40 nl.l" 1 , to prevent a 10 % reduction in growth. The figure was based on experiments on beech, birch and oak seedlings in experimental facilities providing conditions close to those in the field. However, it was emphasised that this was very much a provisional value, given the uncertainty in extrapolating from these short-term experiments to long term effects on mature forests. Critical levels and loads for forestry Since the problem of widespread forest decline began to be identified in Europe 15 years ago, many different hypotheses have been developed to explain its causes. While these causes remain uncertain in many areas, it is now recognised that forest health is often determined by a combination of factors; these may include climatic stresses, the influence of pests and pathogens, management practices and deposition of air pollutants. Even when considering only the effects of air pollutants, it is clear that the effects may be complex. Thus where pine is growing in the vicinity of large sources of sulphur emissions, there may be high concentrations of sulphur dioxide directly affecting the vitality of needles, episodes of acid mist directly damaging needle surfaces, and high levels of acid deposition to the soil, causing soil acidification and nutrient leaching. These sulphur emissions are often also accompanied by heavy metals, whose impact may depend in turn on soil acidity. Where high regional levels of ozone also occur, the direct effects on needles may be influenced by synergistic effects with sulphur dioxide; ozone is also known to affect root development through reductions in assimilate transport, and these effects may interact with those of acid deposition or heavy metals on the root system and its mycorrhizae. Despite the increasing sophistication of our understanding of the impact of multiple stresses on forest ecosystems, there has been little attempt to incorporate this understanding into the critical levels/loads approach (Ashmore et al. 1990). In the case of critical levels, the potential importance of interactions between air pollutants was clearly recognised at the Egham and Bern workshops, but it was felt that the available experimental data was inadequate for these interactions to be formally incorporated into the critical levels actually set. Similarly the importance of pollutant interactions with other biotic and abiotic stresses was recognised as important, but only in the case of the interaction between sulphur dioxide and low temperatures was a modified critical level set, based on field observations. 34 This is clearly an important future issue in developing and refining the critical level concept. However, there has been less recognition of the importance of integrating the critical level and critical load concepts, despite the recognition of the interactions between soil acidification and direct pollutant impacts above ground. This needs to be considered not only in terms of setting critical levels and loads, in a way which incorporates their interaction, but also in terms of mapping the exceedence of existing, separately determined, critical level and load values. For example, a new sulphur protocol has been negotiated based largely on exceedence of critical loads for soils. In much of western Europe the levels of sulphur dioxide found nowadays are below those likely to damage forests but this is not the case in many areas of eastern Europe, where sulphur dioxide concentrations are much higher. By only considering critical loads and ignoring exceedence of critical levels of sulphur dixoide, the procedure used is likely to have underestimated the impact of sulphur deposition on forest health in these regions, even assuming the effects of direct sulphur dioxide damage and of soil acidification were only additive. Ideally then, the way forward for forests would be to develop a more integrated concept of critical pollutant deposition, incorporating critical loads for sulphur and nitrogen, critical levels for relevent pollutants, and perhaps also the critical loads recently proposed for heavy metals (van den Haut 1994). Initially, a simple mapping approach might assist in identifying those forests in Europe for which the combined impact of multiple pollutant stresses, as indicated by exceedence of their critical level/load, may be important. However, such an approach has clear limitations because it would not be based on any mechanistic understanding of forest responses to particular pollutant combinations. Thus there is a need to develop scientific tools which would allow the implications of these combined impacts for critical levels and loads to be assessed. One approach might be based around modelling of those parameters (e.g. nutrient cycling) which are central to the response of forest ecosystems to multiple stresses. While such a development will be intellectually challenging, it reflects the essential need to base the combined assessment of critical levels and loads around a proper mechanistic understanding of the response of forest ecosystems, rather than of individual trees. Acknowledgements Our work on critical levels is supported by the UK Department of the Environment. I thank Prof. Nigel Bell for his comments on this manuscript. 35 References Ashmore, M.R., Bell, J.N.B. & Brown, I.J. 1990. Air pollution and forest ecosystems in the European Community. Air Pollution Research Report 29. Commission of the European Communities, Brussels. Ashmore, M.R. & Wilson, R.B. (eds.) 1994. Critical Levels of Air Pollutants for Europe. Department of the Environment, London. Bull, K.R. 1991. The critical loads/levels approach to gaseous pollutant control. Environ. Pollut. 69: 105-123. Cape, J.N. 1993. Direct damage to vegetation casued by acid rain and polluted cloud: definition of critical levels for forest trees. Environ. Pollut. 82: 167-180. Fuhrer, J. & Achermann, B. (eds.) 1994. Critical levels for ozone; a UN-ECE Workshop Report. FAC Report 16, Swiss Federal Research Institute for Agricultural Chemistry and Environmental Hygiene, Liebefeld-Bern. Makela, A., Materna, J. & Schopp, W. 1987. Direct effects of sulfur on forests in Europe - a regional model of risk. Working Paper 87-57, International Institute for Applied Systems Analysis, Laxenburg, Austria. UNECE 1988. UNECE Critical Levels Workshop Report. Bad Harzburg, FRG, March 1988. van den Hout, K.D. 1994. The impact of atmospheric non-acidifying pollutants on the quality of European forest soils and the North Sea. VROM, Den Haag. Wentzel, K.F. 1983. lUFRO studies on maximal S0 2 imisssions standards to protect forests. In: Ulrich, B. & Pankrath, J. (eds.) Effects of Accumulation of Air Pollutants in Forest Ecosystems, pp. 295-302. D. Reidel, Dordrecht. WHO 1987. Air Quality Guidelines for Europe. World Health Organisation, Copenhagen. 36 Allocation of critical loads to S and N deposition Maximilian Posch Water and Environment Research Institute P.0.80x 250, FIN-00101 Helsinki, Finland Abstract In this paper the historical development of the critical load concept and its use in emission reduction assessments is presented. Special emphasis is put on the methodology for aggregating large numbers of individual critical load values to arrive at a single value for a selected grid. While this has been rather straight-forward for the current method for calculating critical loads - and has been successfully used in negotiating the Second Sulphur Protocol - the newly proposed method for treating sulfur and nitrogen in conjunction also requires new procedures. The unique critical load values are replaced by a critical load function for each ecosystem, and this formulation gives considerable freedom in the allocation of S and N deposition reductions. Although several details still need clarification, the new methodology seems well suited for assisting future negotiations on a multiple pollutant multiple effect protocol. Introduction The signing of the Second Sulphur Protocol in June 1994 constitutes an important step in the history of emission reduction agreements. For the first time ecological concepts have been taken into account in formulating the goals of emission reductions in quantitative terms in a binding international agreement. In this new Protocol the sensitivity of an ecosystem (forest soil or lake) is measured by the so called 'critical load', i.e. the amount of (sulfur) deposition it can stand without detrimental effects in the long run. The concept of a critical load came up in the early 1980s, and - sponsored by the Nordic Council of Ministers - a series of workshops was organized to clarify the concept, to devise methods for calculating critical loads and to discuss procedures to map them (Nilsson 1986, Nilsson and Grennfelt 1988). In 1988 a 'Task Force on Mapping' was initiated under the UN/ECE Convention on Long Range Transboundary Air Pollution. In a workshop held in Bad Harzburg (Germany) in 1989 the methods for mapping critical levels and loads were documented in a manual (Sverdrup et al. 1990, ÜBA 1993). In 1990 a 'Coordination Center for Effects' (CCE) was established in Bilthoven (The Netherlands) with the task to synthesize the different national critical load 37 calculations and to produce standardized European critical load maps which are used within the ECE in the negotiations on emission reduction protocols. Originally, most of the effort has been directed to produce critical load maps of sulfur in order to assist negotiations for the Second Sulphur Protocol (Hettelingh et al. 1991), but in the near future N critical load maps will be required for the forthcoming negotiations of a new nitrogen protocol. In anticipation of that need workshops were held in Lökeberg (Sweden) in 1992 (Grennfelt and Thörnelöf 1992) and recently in Grange-over-Sands (England) which dealt with both the acidifying and nutrient aspects of nitrogen. The workshop reports and new guidelines published by the CCE (Downing et al. 1993) will serve as background documents for future activities in the field. In this paper we first shortly present how the critical load of sulfur is calculated for an ecosystem with the present method and how a single critical load value is obtained for a grid cell containing a (large) number of ecosystems. Secondly, we describe the new method of calculating critical loads for both S and N (the critical load function) and ways for using these functions to assess emission reduction requirements. We conclude the paper with a discussion of the methods and an outlook on future developments. Critical load of sulfur (current method) Critical loads for ecosystems can be derived in various ways, either by empirical methods or by models with different degrees of complexity. In the case of forest ecosystems not only direct effects play a (poorly understood) role, but also soil mediated effects, and to date mostly soil chemical criteria (e.g. a critical Al concentration or Al/Ca ratio) have been used to derive critical loads with simple steady-state models. The largest uncertainty in these calculations is the relationship between the critical chemical values and the harmful effects (De Vries 1993). Steady-state soil models calculate deposition levels which avoid the violation of soil chemical criteria in a steady-state situation. Therefore effects with a finite time scale, such as cation exchange and sulfate adsorption, are not included. The standard model is the so-called steady-state mass balance model, which has been used to produce maps of critical loads of acidity, S (and N) on a European scale. The critical deposition of sulfur, CD(S), has been defined as: 1 ' We here use the term critical deposition of S, since the term critical load of S is - for historical reasons - reserved for the quantity CL(S)=Sf • CL(A). 38 where BC d and BC U are the deposition and uptake of base cations, resp., and CL(A) is the critical load of (actual) acidity, the 'ancestor' of all critical loads. It is defined as the difference of the acid neutralizing capacity (ANC) produced by weathering and the ANC consumed by the maximum acceptable alkalinity leaching. The factor Sf in Eq.l is the so-called sulfur-fraction, defined by where Sdep and NJep are the deposition of S and N, and N„ and Ni are the uptake and immobilization of N, resp. The above formulation has been used by many European countries to compute critical loads of sulfur. In Finland, input data for 1450 lakes and about 3100 forest soil sites have been collected and critical loads calculated. This large number of critical load values calls for appropriate methods of presenting them; and since not only the values as such are of interest, but also their (approximate) geographical location, maps seem to be the obvious way to display the critical loads. However, it is impossible to print the critical load values for thousands of locations on a map. Consequently, the data have to be grouped into a small number of classes (intervals) and these classes are then plotted as different symbols and/or colors on a map. The advantage of such a display is that the geographical location of the sampling sites can be read from the map; the disadvantages are that data points are overlapping and information is lost by the small number of classes (grey shades or color codes), especially when limited to black and white displays (see, e.g., Posch et al. 1993 a). For presenting spatially varying information given by a large number of point data cumulative distributions have proven useful. The region under consideration is divided into regularly shaped subregions (grids) or irregularly shaped polygons, following some administrative or natural boundaries. All critical load values falling into a subregion (grid) are used to define a cumulative distribution function (cdf). For n values xl <...N +N. (2) s f - ■ ■ 1 otherwise 39 where w, is the relative weight (size) of ecosystem i (w,+...+w n =l). The size of a subregion has to be chosen large enough to make the cdf 'representative' for the critical loads in that subregion, i.e. adding more (randomly sampled) values of the subregion should not significantly change the cdf. Using cdfs, the only information lost is the location of the ecosystem within the subregion. Another advantage of displaying data points as cdfs is that the number of data points used is virtually unlimited. It is also easy to combine data sets for the same subregion, e.g. critical loads for lakes and forest soils, by properly re-scaling the weights. In Fig. 1 the critical loads of sulfur for lakes are displayed as cdfs in the so-called EMEP-grid, a coordinate system with grid cells of 150kmxl50km used by EMEP for modeling the long-range transport of sulfur and nitrogen in Europe. The same grid is also used by the CCE for producing harmonized critical load maps for Europe. Figure 1. Critical loads of sulfur for lakes, displayed as cumulative distribution functions in each of the EMEP grid cells covering Finland. 0 for xx n 40 Cumulative distribution functions, however, are not easily interpreted by the non technical user, and therefore a simpler form of display has to be selected. Normally a single statistical descriptor of the distribution is chosen (percentile, mean, etc.) and presented as shaded (colored) grids in different grid systems (small grids for national applications and larger grids in the European context). The mean would be easy to compute, however, it is not a good descriptor for skewed distributions. In addition, the notion of a critical load demands to characterize the lower end of the cdf, i.e. the most sensitive ecosystems. This can be done by selecting a (low) percentile of the cdf. The q-th quantile of a cdf F, denoted by x q , is the value for which F(xJ=q, i.e. x q , as a function of q, is the inverse of F; and thep-th percentile is defined as the p/\OO-th quantile. There is no unique way to calculate the percentile of a cdf defined for a finite number of values. We have selected the simplest method, illustrated in Fig. 2; other methods first smooth the cdf before calculating a quantile. In the UN/ECE mapping exercise the sth percentile has been chosen to represent the critical load for a grid, i.e. the aim is to protect 95 % of the ecosystems. In Fig. 3 the sth percentile of the critical loads of sulfur for lakes is displayed in three grid systems: The grid system in Fig. 3a ('/2° longitude by has been chosen for displaying critical loads in Finland, Fig. 3b shows the sth percentile critical loads in the EMEP-grid used in the European context, and Fig. 3c shows the same critical loads in the so-called NILU-grid, a 3x3-subdivision of the EMEP-grid - a grid which might be used in the future both by EMEP for the deposition modeling and the CCE for mapping critical loads. Figure 2. Example of a cdf from 5 values with non-equal weights (thin horizontal lines). The thick vertical lines define the quantile function, and the quantile for q=0.35 is indi cated by the dotted line. 41 Figure 3. The sth percentile of the critical loads of sulfur for lakes displayed in three different grid systems covering Finland. The exceedance of the critical load of sulfur, i.e. the amount by which deposition has to be reduced in order to avoid 'harmful effects' for the ecosystem, is given by and it is this quantity which enters the models for determining (cost) optimal sulfur emission reduction strategies. As mentioned in the Introduction, the critical loads (critical depositions) computed by the above model and mapped on a European scale have been successfully used in negotiating the new sulfur protocol. However, progress in the field itself and the need for updating the Sofia Protocol on nitrogen emissions makes it necessary to incorporate N (and possibly pollutants connected with it, such as VOCs) into the critical load formulation, and this will be outlined in the next section. Ex(S) = S Jep - CD(S) (4) 42 Critical loads of sulfur and nitrogen (proposed method) The concept of a critical load rests on the assumption that it is possible to derive a unique number, depending on ecosystem properties alone. However, the sulfur fraction (see Eqs. 1, 2) makes the critical load of sulfur dependent on the deposition of sulfur and nitrogen; and consequently CD(S) should be re-calculated, whenever the deposition changes (see Posch et al. 1993b and Kämäri et ai. 1993 for a discussion). Furthermore, future negotiations will concentrate on nitrogen emission reductions and - although N has been considered in the current method - critical loads of N will have to be derived which take into account not only the acidifying but also the eutrophying effects of N-deposition (and eventually the role of N as a precursor of tropospheric ozone). One starts from the acidity balance, which can be written as (see Posch et al. 1993 c): where now also denitrification, NJe, is included. A new aspect arises with NJe as denitrification is not a fixed quantity for an ecosystem, but depends on the net input of N. The simplest formulation is given by: where fde (oCL max(N), in this case the nutrient critical load is not binding, or CL„ ul (N) Water and Environment Research Institute P.O. Box 250, FIN-00101 Helsinki, Finland Abstract Critical loads for N, S and total acidity, and amounts by which they are exceeded by present atmospheric loads, were derived for forests in Europe using a steady state soil model. Furthermore, the chemical response of European forest soils to three emission-deposition scenarios for the years 1960-2050, i.e. official energy pathways (OEP), current reduction plans (CRP) and maximum feasible reductions (MFR), was evaluated with a dynamic soil model. Calculations were made for coniferous and deciduous forests on 80 soil types occurring on the FAO soil map of Europe, using a grid of I.o° longitude x 0.5° latitude. Results indicated that present acid loads exceed critical values in approximately 45 % of the forested area. The area exceeding critical loads was nearly equal for N (50 %) and S (52 %). Howe ver, the maximum exceedances were much higher for S than for N. The area with nitrogen satura