Air pollution and European forests: Policy implications based on simulation models

S. Nilsson and O. Sallnäs

Sten Nilsson is principal investigator of the IIASA Forest Study and Professor al the College of Forestry, Swedish University of Agricultural Sciences, Garpenberg.
Ola Sallnäs is a researcher at the Swedish University of Agricultural Sciences.

Forest decline attributed to air pollution has become a major concern in European society. Although the phenomenon is by no means a new one (researchers have noted forest damage from air pollutants on a local scale for more than a century), the recent appearance of outwardly visible stress symptoms in trees in many areas of Europe and North America has raised the level of attention being devoted to this issue. In 1988, under the convention on Long-Range Transboundary Air Pollution established in 1985, an ad hoc assessment of defoliation in 25 European countries was undertaken (see Table 1). Although this assessment does not permit scientific conclusions to be drawn regarding the causes of observed symptoms, in many cases the existence and extent of forest decline cannot be explained without considering the influence of air pollution.

There is widespread concern within the scientific, industrial, labour, economic and public sectors of European society that a continuation of recent trends in forest decline may lead to a plethora of undesirable consequences. Actions taken over this decade will play a large part in determining the nature of the forest and forest economy several decades into the future. Therefore, those who would design and implement solutions to the problem of forest decline must face the challenge of coordinating short-term actions to achieve long-term objectives.

Within the Environment Programme of the International Institute for Applied Systems Analysis (IIASA), the project "Biosphere Dynamics" seeks to examine long-term, large-scale interactions between the world's economy and its environment. The project conducts its work through research efforts and applied case-studies. The IIASA Forest Study has been under way since 1986 and is addressing the consequences of forest decline attributed to air pollutants. The immediate focus of this study is on the future development of forest resources in Europe. Its objectives are the following:

· to gain an objective view of the future development of forest resources in Europe;

· to build a number of alternative and consistent scenarios about future developments and their effect on the forest sector, international trade and society in general;

· to illustrate the effects of forest decline caused by air pollutants, existing and changed silvicultural management, and expansion of the forest land base;

· to identify meaningful policy options, including institutional, technological and research/monitoring responses that could be pursued to deal with these effects.

Subregional country groupings: EEC-9: Belgium, Denmark, Federal Republic of Germany, France, Ireland, Italy Luxembourg, Netherlands, United Kingdom. CENTRAL: Austria, Switzerland. NORDIC: Finland, Norway Sweden. SOUTHERN: Greece, Portugal, Spain, Turkey, Yugoslavia. EASTERN: Bulgaria, Czechoslovakia, German Democratic Republic, Hungary, Poland, Romania.

This article focuses on efforts related to objective number 3. To forecast various scenarios of the development of forest resources and wood supply in Europe, a matrix-type simulation model has been developed, the Timber Assessment Model. A detailed data base developed by the Forest Study, on resources in Europe, is linked to the simulation model. The model generates the potential biological wood volume over time by country or even subregions of a country, species group, and age, leading to a general wood-supply assessment for Europe (see Figure for subregional country groupings and the separate box for more detail on the methodology). It should be underlined that the scenarios deal with the potential biological wood supplies and not the market supply. In reality there are factors that can and do restrict actual harvests, e.g. roundwood prices, the behaviour of forest owners and restrictions on timber utilization to ensure non-wood benefits. The results and conclusions presented in this article deal with Europe excluding the USSR. A separate article covering the forests of the USSR will be presented in a future issue of Unasylva.

Modelling forest decline attributed to air pollutants

In generating data for the model we have tried to be as quantitative as possible. Our first task was to calculate the distribution of the forests of each country over several sensitivity classes on the basis of the capabilities of forest soils to buffer against acidification. Specific "critical loads", i.e. quantitative estimates of an exposure to one or more pollutants, below which significant harmful effects do not occur according to our present knowledge, have been assigned to the individual sensitivity classes. Critical loads for sulphur and nitrogen have been defined by UN-ECE (1988).

Modelling methodology A matrix-type simulation model was built to generate various scenarios of the development of forest resources and potential wood supply in Europe under different assumptions about future forest decline rates and characteristics, silvicultural practices, and forest land expansion policies in Europe. Since there are different forest structures among the countries of Europe, we had to employ three different model concepts founded on different assumptions. AREA-BASED APPROACH In the area-based approach, specific forest types in our study are described by age and standing volume. The different forest types are characterized by country; region; owner; forest structure (high forest, coppice, etc.); and species. Site class was used as an additional separating variable when supporting data were available. The model is based on a structure developed by Sallnäs (1990). DIAMETER DISTRIBUTION APPROACH In the diameter distribution model, the basic entity on which the description of the forest is based is the individual tree instead of the forest area. The state of the forests belonging to a forest type is described by the distribution of stems over a set of diameter classes. In turn, each diameter class is associated with a mean volume per stem. Dynamics are introduced via transition of stems between the diameter classes. Forest types were characterized with the same criteria as in the area-based approach except that site class could not be used owing to insufficient data. A more detailed description is found in Houllier (1989). SIMPLIFIED APPROACH Owing to problems of data quantity and quality, use of the two approaches described above was not possible for Greece, Turkey and some parts of Yugoslavia. In these cases, potential harvests were estimated as a percentage of standing volume and growth rates as initial growth percentages multiplied by a factor that depends on the relations between actual volume and initial volume. MODELLING FOREST DECLINE ATTRIBUTED TO AIR POLLUTANTS To incorporate decline effects in our forest simulators, the description scheme or matrix of the forest was expanded by two variables: decline class, and sensitivity class. In addition, the transition rates were made changeable over time. Changed silvicultural regimes were also incorporated into the models, in terms of shorter rotations, increased intensity in thinnings and enlarged regeneration.

By combining the Forest Study data base discussed above and the IIASA Regional Acidification Information and Simulation Model (RAINS), (Alcamo et al., 1988), it was possible to estimate the extent of forest area with depositions exceeding target loads today and in the future (see Table 2). Deposition estimates generated by the RAINS model for the year 2000 have been used in our Timber Assessment Model. These deposition estimates are based on current (end-1988) plans to reduce emissions of SO₂ and NO_x as announced officially by individual governments. It should be noted, however, that although most countries predict nearly total elimination of SO₂ and NO_x emissions by the year 2000, this is probably unrealistic.

How excess pollution deposition will damage trees of different types and ages has been estimated by the Forest Study using a model developed in the German Democratic Republic (Bellman et al., 1988). The Prognosis and Decision Support Model for Environment Conservation (PEMU) translates the effect of exceeding target loads into damage and decline patterns by calculating the life expectancy of affected trees and the time that affected trees spend in each of four recognized damage classes (indicated by the level of leaf or needle loss).

TABLE 1. Extent of standing volume of exploitable closed forest In different decline classes in 1987-88 (in million m³)

	Coniferous			Non-coniferous
	Light	Moderate	Severe	Light	Moderate	Severe
Austria	163.8	t 8.2	4.0	49.6	8.2	1.6
Belgium	12.0a	3.2	1.0	12.5	2.9	0.6
Bulgaria	39.4	8.8	1.1	59,2	14.0	3.6
Czechoslovakia	301.8	148.2	37.0	93.4	55.7	13.3
Denmark	3.6	0	0	11.2	2.6	0.2
Finland	301.7	192.2	28.2	71.1	21.5	1.5
France	110.1	49.8	6.9	139.8	41.5	9.6
German Dem. Rep	112.2	...52.6...		24.3	...9.1...
Germany, Fed. Rep.	263.4	98.2	6.0	133.6	49.9	2.5
Greece	33.3	5.9	0.8	32.6	17.1	0.9
Hungary	4.9	2.8	0.8	30.1	8.2	7.3
Ireland	6.4	1.1	0.1	--	--	-
Italy	36.4b	18.1	2.6	79.4	37.0	4.3
Luxembourg	0.5	0.2	--	4.0	1.1	0.2
Netherlands	3.4	2.1	0.4	3.0	1.6	0.4
Norway	135.4	77.1	18.4	--	--	-
Poland	305.8	187.2	36.0	38.2	14.8	5.4
Portugal	4.0	1.8	0.2	4.5	--	0.7
Spain	61.3	20.2	3.6	58.6	12.6	2.2
Sweden	601.4	228.8	29.4	64.9	18.6	0.7
Switzerland	88.4	32.2	8.0	25.2	4.9	1.9
United Kingdom	43.3	43.5	10.9	175.5	55.2	2.2

^a Average for Flanders and Wallonia
^b Average for Bolzano and Sardinia

TABLE 2. Proportion of total forest area where critical loads for sulphur and nitrogen are exceeded (percentage)

	Sulphur				Nitrogen
	Coniferous		Deciduous		Coniferous	Deciduous
	1985	2000	1985	2000	1985-2000	1985-2000
Nordic	59	48	19	7	75	52
EEC-9	88	76	34	24	83	55
Central	98	93	50	46	100	86
Southern	62	84	18	40	34	21
Eastern	98	98	84	76	76	47

The calculations show the impact of sulphur deposition to be dramatic. So long as the level of sulphur deposition remained below two grams per square metre per year (g/m²/year), a pine tree in an area of low soil sensitivity would easily live more than 100 years and spend at least 60 of these in a healthy state, i.e. with more than 90 percent of its foliage intact. That same tree would live only 65 years if it experienced sulphur depositions of 2-4 g/m²/year. Moreover, it would spend only its first 30 years in a fully healthy state. At 42 years of age, it would have lost 25 percent of its needles, and by the time it reached 55 years, it would have less than 40 percent of the needle coverage of a healthy tree. In an area with highly sensitive soils with sulphur depositions of 24g/m²/year, the same tree would live only 26 years and would never be fully healthy.

Although the cause-effect relations in PEMU have been fully quantified only for sulphur depositions on pine stands, preliminary results from a new spruce-decline model indicate the same basic results.

Scenarios on future developments

For our analysis of potential forest-sector futures under different assumptions about pollution-induced decline, silviculture and forest land area, we have implemented a number of scenarios. The base year for the simulations is 1985 and all have a time horizon of 100 years, i.e. up to 2085.

In the Handbook Basic Scenario, it is assumed that the forests of each country are treated strictly in accordance with "ideal" silviculture programmes for the stands in the forests of each country. Results from using this approach show the degree to which existing forest policies have been implemented. In the Forest Study Basic Scenario, the objective is to strive for consistently high levels of both growing stock and harvests over the total simulation period.

A second pair of scenarios, parallel to the two described above but assuming forest decline based on the output of the RAINS and PEMU models, have also been developed. They are labelled the Handbook Decline Scenario and the Forest Study Decline Scenario.

Pollution damage to European conifers has been calculated by the Forest Study

The four scenarios discussed above do not account for any change of the forest land base over time in Europe. However, there is strong potential for overall increases in the amount of forest land in Europe, a reaction to proposals for the redirection of agricultural subsidization from food production (already at a surplus) to tree production, as well as increased attention to reforestation of degraded and unused lands for a variety of end uses. Therefore, we have also simulated a series of forest land expansions. The results of these simulations are beyond the scope of this article, but interested readers may wish to consult the full report of our study (Nilsson et al., 1990).

The kind of quantitative scenario-building in which we have engaged does not have the purpose of sketching the most likely forest sector futures; rather the aim is to determine what policies make sense under a wide range of possible futures, and what specific information should be generated to improve our abilities to manage forest resources properly.

Results

Our results show that, as expected intuitively, the handbook scenarios give the highest harvest potentials (Table 3). By comparing the 100-year average harvests from the Forest Study Basic Scenario with actual harvests in 1987 it can be seen that there is a potential to increase long-term sustainable harvests by about 100 million m³/year.

TABLE 3. Potential harvest under study scenarios (in million m³/year)

	Basic handbook	Basic forest study	Decline handbook	Decline forest study	Actual 1987 removals
Nordic	153.4	155.3	155.2	144.2	120.7
EEC-9	159.3	150.1	146.4	126.2	109.3
Central	26.6	24.8	25.4	18.9	22.3
Southerna	-	78.4	-	71.0b	72.1
Eastern	130.7	126.0	123.2	91.7	99.5
Total Europe	-	534.5	-	452.1	423.9

^a Lack of data precludes calculation for Southern region and therefore for total Europe
^b Decline effects for Spain not calculated, owing to lack of data

TABLE 4. Potential harvest losses caused by air pollutants (percentage)

	(Basic - Decline)/Basic x 100	(Basic - Decline)/1987 x 100
Nordic	7	9
EEC-9	16	22
Central	23	26
Southern	13	14
Eastern	16	20
Total Europe	16	20

Notes: Scenario data used to calculate the percentages are 100-year averages. Basic = Forest Study Basic scenario. Decline = Forest Study Decline Scenario. 1987 = FAO statistics of actual harvests. In the estimate for the Southern subregion no decline effects have been calculated for Spain.

Under the decline scenarios, a 16 percent total loss of potential harvest is forecast owing to damage caused by air pollutants expected to be emitted in Europe up to 2000-2005 (see Table 4). This represents a loss of about 85 million m³/year, averaged over 100 years. On a subregional basis, the lost potential would be Nordic, 11.0; EEC-9, 23.9; Central, 5.8; Southern, 10.2; and Eastern, 34.3. It is noteworthy that actual 1987 harvests were substantially lower than those forecast by the Timber Assessment Model. Therefore, the decline effects could be even greater than those forecast in our simulations.

Air pollution emissions in Europe have risen steadily during this century (except for specific recent initiatives to control sulphur dioxide). Even optimistic scenarios for future patterns of air pollutant emissions in Europe give little cause for optimism about air pollutant stress on forests. So-called lag effects of air pollutants on forest ecosystems may last for several decades, so that even if pollutant emissions are adequately controlled soon, forests will not respond fully to the clean atmosphere immediately. Therefore, all other things being equal, real decline is likely to be more serious than our projections.

Policy implications of continued forest decline in Europe

There are two major potential entry points into the relationship between air pollutant emissions and forest resources: pollution control; and raising the stress resistance of forests. With respect to air pollution control, our results show that current policies for emission reductions of sulphur and nitrogen oxides will make relatively small reductions in the areas of forest at risk to detrimental effects from these pollutants. Blanket application of fixed-proportion reductions of air pollutant emissions across Europe clearly ignores the facts that some forest ecosystems are more sensitive than others; that countries are emitting very different quantities of air pollutants, on absolute as well as on per caput, per unit land area, or per gross domestic product bases; and that pollutants are usually not deposited in the locale of their emission. Under the auspices of the UN-ECE Convention on Long-Range Transport of Air Pollutants, policies should be devised for pollution controls strongly targeted at specific pollutants and specific polluters, with the objective of reducing pollutant depositions to below critical loads for all European forests.

National governments are clearly responsible for developing and implementing new targeted policies for air pollution control. However, because of the international nature of the problem, international organizations such as the European Community and the UN-ECE have a responsibility to bring national governments together to undertake international impact assessments and develop strong pollution control policies acceptable to all parties.

TABLE 5. Outlook for European wood supply/demand balances

	Balances at year
	mid-1980s	2000	2010
	(million m3 roundwood equivalent)
Nordic
Domestic demand a	31.2	41.2	43.4
Surplus/deficit
1987 actual harvest	+89.5
Decline b		+88.9	+91.5
EEC-9
Domestic demand	202.5	279.0	326.5
Surplus/deficit
1987 actual harvest	-93.2
Decline		-160.1	-200.0
Central
Domestic demand	13.8	18.3	22.8
Surplus/deficit
1987 actual harvest	+8.5
Decline		·0.2	-3.9
Southern
Domestic demand	49.3	67.2	77.2
Surplus/deficit
1987 actual harvest	+22.8
Decline c		-1.9	-9.3
Eastern
Domestic demand	69.4	91.3	105.6
Surplus/deficit
1987 actual harvest	+30.1
Decline		-5.4	-17.6
Total Europe
Domestic demand	352.4	497.0	575.5
Surplus/deficit
1987 actual harvest	+71.5
Decline		-75.4	-139.3

^a Roundwood demand to meet domestic consumption of final industrial products
^b Potential wood supply according to the Forest Study Decline Scenario
^c No decline effects calculated for Spain, owing to lack of data

A key constraint in implementing strong targeted policies for air pollution control in Europe is the availability of funds, particularly in the countries where controls most need to be installed. Significant difficulties can be expected in international negotiations, first in determining which countries ought to shoulder the blame for pollutant depositions exceeding critical loads, and second in determining who should take on the financial burden of controlling that pollution.

Policy implications with respect to silviculture practices

There are three basic silvicultural elements that are critical in determining the level of vitality of forest stands: timing and intensity of thinnings that control density of trees in forest stands; age at final felling that controls the degree to which stands become overmature; and matching regenerated species with site potentials, controlling the vigour with which species grow on specific classes of sites. In the Forest Study, we have concentrated our analytical efforts on thinnings and final felling as tools to reduce the risk to forest stands of air pollution stress.

We conclude that the problem of silviculturally induced stress in European forests is widespread and serious. Many European forests have a low resistance to stress because they are too dense (thus in a strongly competitive environment), too old (physiologically overmature), or inappropriately matched to site conditions. If these obstacles are not overcome soon, we may expect that continuation of current silvicultural practices will further contribute to forest decline. However, forest stand resistance to stresses such as air pollution could be raised significantly by implementing known silvicultural practices.

Forest research institutions have the responsibility to reaffirm good silviculture practices and to adjust them as required in response to the stress of air pollution. Landowners - both governments, commercial interests and private individuals - need to find means to implement resilience-building silvicultural practices, and to be made aware of the consequences of not implementing them. In particular, they need to be informed that old and dense forest stands have low resistance to air pollution stress, and that judicious cutting is the most powerful tool to restore this resistance. The forest products industry will need to be receptive to the increased wood supply that would result from strong implementation of proper silviculture.

Implications for supply/demand balances for industrial roundwood

To judge from 1987 actual harvest data and 1985 demand date (both derived from FAO sources), it appears that in the late 1980s Europe as a whole was in a surplus supply situation (see Table 5). The surplus of overall wood supply amounted to 71.5 million m³/year. When we look to the future at years 2000 and 2010 the picture changes significantly. Even in the case of the scenarios that assume no forest decline, Europe may face an annual roundwood deficit of more than 47 million m³ by 2010. In the decline scenarios, that deficit rises to about 140 million m³/year.

Immediate targeted reductions of air pollutants in Europe are required

The strongest influence of pollution-induced forest decline in making potential deficits worse is in the EEC-9, Eastern and Southern subregions; the most serious potential deficit is in EEC-9, largely because roundwood demand is expected to grow so strongly in the countries it comprises.

Industrial capacity

At the end of the 1980s, most countries in Europe had industrial overcapacities compared with actual levels of timber harvest. Exceptions include Denmark, Ireland and the United Kingdom. Under the Forest Study Basic Scenario, most countries would be short of industrial capacity if harvesting of total sustainable potential were assumed. However, a comparison of current capacities with potential wood supplies under forest decline conditions, as exemplified by the Forest Study Decline Scenario, suggests that many European countries would continue to experience overcapacity if decline were to continue as we expect it might.

Under the Forest Study Basic Scenario, capacities would need to be expanded in the Nordic and EEC-9 subregions regardless of whether forest decline were to occur, although the degree of capacity expansion required would be significantly less in the decline cases. In the case of the Central subregion, all scenarios predict significant overcapacity. Policies relating to softening the impact of closures and to restructuring the industry will be needed. Finally, in the Southern and Eastern subregions, if forest decline persists as we assume in our decline scenarios, current overcapacity will be exacerbated.

Non-timber forest values

Due to serious paucities of appropriate data and functional relationships for building forecasting models, we have been unable to prepare Europe-wide scenarios of future non-timber forest benefits (recreation, hunting, forest products other than timber, etc.) and their responses to changing forest structure and to pollution-induced forest decline. However, it is clear that forest decline no discrimination as regards forest values and will therefore negatively affect non-timber benefits as well. In fact, what scant literature there is on effects of air pollution on non-timber forest values (Stoklasa and Duinker, 1988) suggests that the economic losses associated with impacts on non-timber benefits may significantly overshadow those associated with timber.

Conclusions

Forests are strategic resources for human well-being, and their importance will without doubt increase in the future. Europe has a strong platform from this strategic viewpoint. The forest land base is likely to remain stable or even to increase, since many governments are making signs in this direction. With more careful and insightful forest resource management and policy formulation across Europe, the forest resources can surely provide more wood and mote non-wood benefits, but only if the major threat of air pollution can be avoided.

We conclude that immediate targeted reductions of air pollutants in Europe are required. However, even full success in controlling the major damaging air pollutants, which is not likely in the short teen, would not be enough to halt the decline in European forests. We believe that inadequate implementation of good silvicultural practices compounds pollution-related decline, and therefore that improvements in implementing basic forest management are also required.

Design and implementation of policies to effectively address this need will requite strong partnerships among forest owners and industrial, governmental and environmental interests. Building the required trust, raising the critical awareness, generating the needed understanding, and finding sufficient financial resources for promptly and effectively tackling the problem of the pollution-induced decline of European forests will call for economic and environmental cooperation on a scale similar to that being mobilized on behalf of the world's tropical forests.

Bibliography

Alcamo, J. et al. 1988. Acidification in Europe: a simulation model for evaluating control strategies. RR-88-2. Laxenburg, Austria, IIASA.

Bellman, K. et al. 1988. The PEMU forest-interpact-model. FORST K. A pine stand decline and wood supply model. In Systems and simulations, vol. II. Ser. Math. Res., vol. 47. Berlin, Akad. Verlag.

Houllier, F. 1989. Data and models used for France in the context of the Forest Study, IIASA. Laxenburg, Austria, IIASA. (Unpublished manuscript)

Nilsson, S., Sallnäs O. & Duinker, P. 1990. Forest decline in Europe -forest potentials and policy implications. Laxenburg, Austria, IIASA.

Sallnäs, O. 1990. A matrix growth model of the Swedish forest. Uppsala, Sweden, Studia Forestalia Suedica.

Stoklasa, J. & Duinker, P.N. 1988. Social and economic consequences of forest decline in Czechoslovakia. Laxenburg, Austria, International Institute for Applied Systems Analysis.

UN. 1986. European timber trends and prospects to the year 2000 and beyond. New York, ECE/FAO United Nations.

UN-ECE. 1988. ECE critical levels. Workshop report. Bad Hartzberg, Federal Republic of Germany.