Strategies Towards Scenarios of Forest Damage Due to Air Pollution

W O R K I N G P A P E R

STEUTEGIES TOWAEWS SCENARIOS OF FOREST DAMAGE DUE TO AIR POLLUTION

March 1985 WP-85-012

I n t e r n a t i o n a l I n s t i t u t e for Applied Systems Analysis

NOT FOR QUOTATION WITHOUT PERMISSION OF THE AUTHOR

STRATEGIES TOWARDS SCENARIOS OF FOREST DAMAGE DUE TO AIR POLLUTION

Annikki Make16

March 1985 WP-85-012

Working Papers are interim r e p o r t s on work of t h e International Institute f o r Applied Systems Analysis and have r e c e i v e d only lim- ited review. Views o r opinions e x p r e s s e d h e r e i n do not neces- sarily r e p r e s e n t those of t h e Institute o r of i t s National Member Organizations.

INTERNATIONAL INSTITUTE FOR APPLIED SYSTEMS ANALYSIS 2361 Laxenburg, Austria

PREFACE

From mid 19701s, symptoms of a new f o r e s t decline which is spreading

at

a n increasing speed h a v e been detected in Central Europe. Today, many scientists s h a r e t h e opinion t h a t t h e new symptoms are connected with a i r pollution, y e t no single pollutant o r damage mechanism h a s been proved as t h e single c a u s e of f o r e s t dieback.

A s a r e s p o n s e t o t h e need f o r g r e a t e r understanding of t h e problem, scientists are gathering empirical evidence and developing new methods f o r t r e a t i n g t h e r e s p o n s e of f o r e s t s t o e x t e r n a l stress f a c t o r s . P a r t of this development, mathematical modeling is becoming more important a l s o in this field.

This p a p e r reviews t h e c u r r e n t s t a t e of knowledge about t h e relation- ship between a i r pollution and f o r e s t damage in E u r o p e in a mathematical modeling framework. I t will s e r v e as a basis f o r a n o t h e r submodel of t h e RAINS model (Regional Acidification and INformation Simulation).

Leen Hordijk

Acid Rain P r o j e c t L e a d e r

ABSTRACT

This r e p o r t i s a review of t h e requirements and possibilities of developing a regional scale, long term model f o r t h e r e s p o n s e of f o r e s t s t o atmospheric pollution. A g e n e r a l input-output s t r u c t u r e of such models i s delineated, and t h e input, output and potential model s t r u c t u r e s are con- s i d e r e d . With t h e o b j e c t i v e of specifying t h e input v a r i a b l e s , potential c a u s e s of f o r e s t damage are classified according

to

t h e i r dynamic p r o p e r - ties. F o r e s t damage i s f u r t h e r specified, s o as t o define t h e output vari- ables. Theoretical and empirical models f o r t h e r e s p o n s e of f o r e s t s t o a i r pollution are reviewed, as w e l l

as

some empirical evidence applicable

to

t h e models. The possibilities of constructing s c e n a r i o s of damage o v e r E u r o p e are assessed, and t h e p r e s e n t t h e o r e t i c a l and p r a c t i c a l r e s t r i c t i o n s are evaluated.

STRATEGIES TOWARDS SCENARIOS OF FOREST DAMAGE DUE TO AIR POLLUTION

Annikki Makela

1. INTRODUCTION

1.1. Forests in changing environment

From t h e middle of t h e 19701s, symptoms of a new f o r e s t disease have been detected in Central Europe. A t least 1.5 million h e c t a r s of f o r e s t has s o far been totally destroyed (Wentzel 1982), and much more is classified as

"affected" (Walderkrankung , 1983). The disease f i r s t attacked silver f i r (Abies a l b a ) , 80% of which has now been lost, and i t is currently making p r o g r e s s in stands of Norway Spruce (Picea a b i e s ) (Walderkrankung, 1983). Recent observations indicate t h a t t h e disease is also spreading out t o Sweden (Kvist & Barklund, 1984).

Today. many scientists s h a r e t h e opinion t h a t t h e emergency of t h e new symptoms was connected with a thorough change in t h e pollutant load. In Wentzel's (1982) classification, a new phase of a i r pollution w a s reached in

t h e e a r l y 1960's. when local and regional emission s o u r c e s had gradually given way t o global pollution with remote t r a n s p o r t of chemicals and

an

ever-denser network of local sources such

as

industries and traffic. Simul- taneously, t h e variable load with a c u t e episodes had been replaced by

a

low but permanent pollution level.

1.2. Scenario analysis to assess alternative policies

The ratification of t h e Geneva Convention on Transboundary Air Pollu- tion in 1983 showed t h a t t h e European countries

are

deeply concerned about t h e problem and determined t o t a k e action t h a t will r e d u c e emissions.

Since t h e aim is not only t o r e a c h

^an

ecologically sustainable level of emis- sions, but also t o do t h i s at t h e lowest possible cost, r a t h e r detailed infor- mation about t h e expected consequences of alternative industrial policies should b e available t o t h e decision makers.

With t h e objective of increasing t h e background information, t h e IIASA Acid Rain P r o j e c t is developing

a

model system (Regional Acidification IMformation Simulation, RAINS) which produces scenarios about t h e conse- quences of alternative energy policies and emission control actions. The input t o t h e model system is t h e pollutant emissions in Europe

as a

function of s p a c e and time, and i t produces scenarios of (1) how t h e input s t r a t e g y is reflected in t h e physical environment, e.g. a i r and soil, and (2) how t h e change in t h e physical environment affects different ecosystems, especially lakes and forests.

A t

present, t h e input comprises a n

energy pathway

which depends

upon selected energy s o u r c e s and pollutant reduction policies in 27 Euro-

pean countries. From this input, t h e system f i r s t calculates t h e annual

emissions of sulphur o v e r Europe. Then t h e atmospheric t r a n s p o r t of sul- p h u r and t h e deposition are calculated on a n annual basis (Alcamo

et

al.

1984). The consequent soil acidification is determined using a model based on a buffer r a n g e t h e o r y (Kauppi

et

^al. 1984) (See Figure 1 ) . F u r t h e r , a model of lake acidification i s under p r e p a r a t i o n (Kamari, 1984).

CONTROL ALTERNATIVES 1. Flue ga, control

2. Fuel cleaning

3. Low sulfur power plants 4. Low sulfur fuel

Figure 1. The p r e s e n t s t r u c t u r e of t h e RAINS model.

Sulfur m i ~ O n O b

Atmospheric P r o w

t

The above c o n c e r n s changes in t h e physical environment, termed as task (1) above. S o f a r , t h e response of f o r e s t s t o t h i s change h a s been con- sidered assuming t h a t soil acidity has a n u p p e r limit which cannot b e exceeded without critically disturbing t h e ecosystem (Kauppi

et

al. 1984).

The p r e s e n t objective i s t o delineate a s t r a t e g y of how t o f u r t h e r promote task (Z), i.e. modelling t h e r e s p o n s e of f o r e s t s

to

t h e change in t h e i r physi-

E n W Y

phwy

Sulfur dwositionb Energy

-

Emissions

-

Forest Soil p~

Soil pH b

c a l environment.

1.3. Developing forest impact scenarios

Using a n a c c u r a t e language, t o develop scenarios of f o r e s t damage o v e r Europe means, t o produce a spatially and temporally distributed esti- mate f o r t h e d e g r e e of damage as a function of a i r pollution, which is simi- larly distributed o v e r s p a c e and time. In space, o r geographically, t h e out- put should extend o v e r Europe, and in time it should r e a c h about 50 y e a r s forward. The time resolution of t h e existing submodels is from one month t o one y e a r , and t h e spatial grid net consists of rectangles with 50-150 km sides.

Since t h e existing atmosperic submodels of t h e RAINS model deal with t h e distribution and deposition of sulphur, a natural continuation would b e t o base t h e f o r e s t damage scenarios on sulphur derivatives. Furthermore, as t h e injurious consequences of soil acidification have already been con- sidered (Kauppi

et

al. 1984), a n obvious i t e m f o r f u t u r e work seems t h e impacts of a i r b o r n e sulphur dioxide. However, as many o t h e r potential causes of damage have been gaining emphasis in t h e scientific discussion since t h e s t a r t of t h e p r o j e c t , i t w a s r e g a r d e d necessary t o review t h e problem from a wider perspective. Section 2 t r e a t s t h e most frequently r e f e r r e d potential causes of damage, with t h e objective of choosing a suit- a b l e input f o r t h e s c e n a r i o model.

The desired output of t h e model, t h e d e g r e e of f o r e s t damage, is prob- lematic because i t does not have a standard meaning, but a variety of p e r - ceptions exist. F u r t h e r , if a qualitative meaning is specified, a selection of potential quantifications will remain. In o r d e r t o specify a suitable output

variable, Section 3 considers different ways of defining and measuring dam- age. Section 4 aims

at

connecting output t o input, and i s devoted t o both t h e o r e t i c a l and empirical models f o r t h e response of f o r e s t s t o a i r pollu- tion. Some empirical evidence are also reviewed. Section 5 deals with t h e possibilities of constructing scenarios of f o r e s t damage in Europe and evaluates t h e p r e s e n t t h e o r e t i c a l and p r a c t i c a l r e s t r i c t i o n s .

2. POTENTLAL CAUSES

OF

DAMAGE

F o r e s t damage caused by a i r pollution i s not a new phenomenon. In t h e e a r l y days of industrialization, i t w a s not unusual t h a t t h e neighbourhood of a smelter, f o r instance, suffered frequent exposures t o high concentrations of pollutants, and a c u t e symptoms of damage o c c u r r e d . The f i r s t systematic studies of such e v e n t s were written in t h e middle of t h e 1 9 t h c e n t u r y (Stockhardt, 1850, 1871; S c h r o e d e r and Reuss, 1883). L a t e r on, s t a c k s became h i g h e r and t h e neighbourhood exposed t o impurities enlarged. This began t h e period of regional a i r pollution and c h r o n i c damage. Both of t h e s e periods, classified as t h e f i r s t and t h e second p h a s e of a i r pollution by Wentzel (1982), s h a r e t h e p r o p e r t y t h a t t h e c a u s e of damage w a s rela- tively easy t o t r a c k , t h e symptoms o c c u r r i n g o v e r a limited area as w e l l

as

during a r e s t r i c t e d time. F u r t h e r , t h e removal of t h e pollution s o u r c e nor- mally a l s o removed t h e damage (Wentzel, 1902). a r e a s o n why t h e r e w a s a p p a r e n t l y no need t o consider t h e e f f e c t s as accumulating o r delayed.

Until v e r y r e c e n t l y , t h e most f r e q u e n t explanation of f o r e s t damage w a s sul- p h u r dioxide originating in e n e r g y combustion.

A s mentioned above, a new p h a s e of a i r pollution w a s r e a c h e d in t h e e a r l y 1960's (Wentzel, 1982). Although s e v e r e concern of widespread dam-

a g e w a s expressed a l r e a d y in t h e 1960's (Knabe, 1966), t h e problem only gained wide publicity in t h e late 1970's when f o r e s t damage s t a r t e d t o proceed rapidly. The explanation based on sulphur dioxide no more seemed t o function, since (1) damage o c c u r r e d in areas where concentrations of sulphur dioxide were f a r below e x p e r t recommendations, such as t h e 25 l . l ~

/ m 3

recommended by IUFRO (Materna, 1983), and (2) t h e emissions of sulphur reduced in many European countries since t h e 1970's (Figure 2).

Therefore, new hypotheses about t h e causes and mechanisms of t h e disease have e n t e r e d t h e discussion. Ozone and soil acidification, in p a r t i c u l a r , have been brought on t h e list of potential agents of damage, but a l s o o t h e r elements such as heavy metals, secondary photooxidants o t h e r than ozone.

and even t h e radioactive r e l e a s e from nuclear power plants, as w e l l as com- binations of all t h e s e have been suggested.

From t h e point of view of modelling, i t i s illustrative t o group t h e hypotheses on t h e basis of t h e dynamic response mechanisms they presume.

Such a classification h a s a l r e a d y been presented by Kohlrnaier

et

al. (1984), who distinguish between d i r e c t and delayed impacts. The delayed impacts c a n b e f u r t h e r divided into delays in (1) physical environment and (2) bio- logical response. Thus, t h e hypotheses concerning t h e potential causes of damage can b e classified as follows:

[I] Direct (non-delayed) impacts.

The e f f e c t closely follows t h e exposure, i.e. t h e time development of damage follows t h a t of t h e a i r pollutant concentration. What i s actually meant by a delay

-

a second, a month, o r a y e a r

-

depends upon t h e time s c a l e of t h e p r o c e s s e s considered. A s r e g a r d s f o r e s t s

,

t h e impact c a n

hardly b e considered delayed if i t o c c u r s within one y e a r after t h e exposure. The e a r l i e r , a c u t e damage a t t r i b u t e d

to

s u l p h u r dioxide w a s consistent with t h i s impact mechanism, but because of r e a s o n s men- tioned above, t h i s i s n o more t h e case. The hypothesis t h a t ozone i s t h e c a u s e of damage, instead, gains i t s v e r y s u p p o r t from t h e f a c t t h a t in Western E u r o p e , t h e concentrations of nitrogen oxides and t h e i r a i r b o r n e d e r i v a t i v e s , such as ozone, have i n c r e a s e d in p a c e with t h e new f o r e s t d i s e a s e (Guderian et al., 1984, P r i n z

et

al., 1982, P r i n z et al., 1984). Development of sulphur a n d NO, emissions in some c o u n t r i e s

are

depicted in Figure 2 (Figures Za, 2b, 2c).

lo6

tons sulfurlyear

Figure 2a.

1900 1920 1940 1960 1980 Year

Development of sulphur emissions in Europe (Forsurning 1982).

lo3

tons sulfurlyear

500

r

Sulfur emissions in Sweden 1950-80 and an estimate for 1990

1950 1960 1970 1978 1990 Year

Figure 2b. Development of sulphur emissions in Sweden (Forsurning

. . .

, 1982).

Sulfur dioxide

t

consumption ^Energy

t

Nitrogen oxides

Power plants Industrial plants

Household and commercial Traffic

Iron and steel production Chemical industries

Figure 2c. Annual emissions of SO2 and NOx in t h e Federal Republic of Germany (UBA, 1977).

[2] Delay in t h e physical environment

The effect only o c c u r s a f t e r a long time of e x p o s u r e , because t h e tox- ification of t h e environment i s a slow process. However, as soon as t h e environment h a s become toxified, t h e trees

start

t o show damage. The soil acidification hypothesis p r e s e n t e d by Ulrich (e.g. Ulrich, 1983) belongs t o t h i s group. Hence, s u l p h u r c a n c a u s e damage through a delayed accumulation process. However, since o t h e r s u b s t a n c e s have

use

not y e t been investigated in t h i s r e s p e c t , t h e y cannot b e excluded e i t h e r .

[3] Delay in t h e r e s p o n s e

The damage i s delayed, b u t t h e delay i s connected with t h e r e s p o n s e of t h e tree r a t h e r t h a n its environment. The r e c e n t

stress

hypothesis (e.g. Wentzel, 1982, S c h u t t et al., 1983, Matzner and Ulrich, 1984) assumes a n impact mechanism of this type. In i t s g e n e r a l form, t h e hypothesis claims t h a t d i f f e r e n t stress f a c t o r s c a n c a u s e a similar chain of events t h a t lead t o t h e observed symptoms, including a n interaction with n a t u r a l stress f a c t o r s . T h e r e f o r e , i t i s important

to

g e t insight into t h e eco-physiological p r o c e s s e s t h a t are common t o all t h e f a c t o r s t h a t lead t o t h e disease. Certain substances h a v e been pointed out

as

important stress f a c t o r s , and t h e corresponding chains of injury have been sketched. Wentzel (1982) assumes t h a t i t is sulphur dioxide t h a t initially launches t h e p r o c e s s , Schiitt e t al. (1983) d e s c r i b e a chain initiated by a n air-pollutant imposed d e c r e a s e in pho- tosynthesis, and Matzner a n d Ulrich (1984) suggest t h a t i t i s t h e acidi- fied soil t h a t imposes t h e fatal pollutant stress.

3.

DAMAGE

3.1. Concept of damage

When t h e effect of pollutants on plants i s studied from a scientific point of view, all r e a c t i o n s of plants t o t h e abnormal f a c t o r s are of i n t e r e s t . In decision making, however, only t h o s e changes t h a t somehow r e d u c e t h e value of t h e plant are important. To make a distinction between t h e s e two

t y p e s of impact, Guderian ( e t al. 1960; 1977) suggests t h a t t h e

terms

"injury" and "damage" should b e used in specific meanings. Though not necessarily harmful, all pollutant-imposed r e a c t i o n s should b e called

"injury", while "damage" should include t h o s e changes t h a t r e d u c e t h e intended value o r use of t h e plant. In a c c o r d a n c e with t h i s definition, w e shall u s e t h e

t e r m

damage t o indicate a harmful change in t h e plant's normal behaviour.

A more a c c u r a t e definition of damage r e q u i r e s t h a t t h e limit between a c c e p t e d and harmful behaviour i s specified. In t h e following, t h e area of a c c e p t e d behaviour will b e called t h e norm.

The c h a r a c t e r i z a t i o n of t h e norm i s a value judgement. What people r e g a r d a c c e p t e d depends upon t h e i r point of view, e x p e r i e n c e and objec- tives. F o r instance, biochemists would look

at

d i s t u r b a n c e s in t h e vital r e a c t i o n chains, eco-physiologists would

measure

d e c r e a s e in photosyn- t h e s i s and i n c r e a s e in leaf n e c r o s i s , f o r e s t e r s would p e r h a p s point out dec- lining timber yield, while f o r e s t h i k e r s would b e most i n t e r e s t e d in healthy looking canopies.

From t h e point of view of t h e p r e s e n t objective, t h e concepts of visi- ble injury and g r o w t h r e d u c t i o n seem t h e most important. They con- c e r n trees and f o r e s t s r a t h e r t h a n o r g a n s and p r o c e s s e s , and a g r e a t d e g r e e of measurement a n d monitoring h a s b e e n based on t h e s e concepts.

They are also in t h e i n t e r e s t of t h e g e n e r a l public and of f o r e s t industry.

3.2. Visible injury

Visible injury i s a n a t u r a l definition of damage if t h e f o r e s t basically s e r v e s a social function, especially providing r e c r e a t i o n areas and contri- buting t o t h e s c e n e r y . Also, in a r e a s where t h e pollutant situation has a l r e a d y become s e r i o u s i t may b e t h e only sensible way of viewing t h e prob- lem. Hence, most of t h e r e c e n t surveys and r e p o r t s of f o r e s t decline in Central Europe are based on this view of f o r e s t damage (Walderkrankung, 1983; Materna, 1983).

Visible injury o c c u r s as leaf necrosis, immature leaf and needle death, and

at

a f u r t h e r s t a g e , t h e death of branches and t h e fall-off of t h e top.

When a n individual tree i s concerned, t h e r e f e r e n c e norm i s a tree t h a t looks healthy, i.e. does not show any of t h e above symptoms. Since t h e o c c u r r e n c e of visible injury v a r i e s among individual t r e e s , t h e definition of f o r e s t damage r e q u i r e s t h a t one specifies how l a r g e a proportion of trees i s accepted t o show t h e symptoms b e f o r e damage i s r e p o r t e d .

A measure of t h e d e g r e e of visible damage of individual trees t h a t has already become s t a n d a r d i s t h e number of needle a g e classes in t h e crown.

In t h e Federal Republic of Germany, f o r instance, t h r e e damage classes are s e p a r a t e d , and s t a n d a r d examples of t h e classes have been published t o unify t h e quantification p r o c e d u r e (Walderkrankung, 1983). The damage of deciduous

trees

i s more difficult t o quantify because t h e i r a p p e a r a n c e changes o v e r t h e growing season. The percentage of necrotic leaves h a s been used as a n indicator.

When quantifying visible injury

at

f o r e s t level,

it

is a question of find- ing a measure t h a t a g g r e g a t e s any distribution of individual t r e e damage t o a n index t h a t e x p r e s s e s t h e seriousness of t h e damage. A s this i s by no means a straightforward t a s k , i t is understandable t h a t a variety of

different usages exist. The official r e p o r t s of t h e Federal Republic of G e r - many simply distinguish between affected and healthy a r e a s , though a n e x a c t definition of a n affected area is not always provided. Materna (1983) h a s developed some f a i r l y e x a c t definitions of t h e t y p e "a stand is affected t o a d e g r e e I if x % of t h e trees a r e damaged t o t h e d e g r e e J", which have been used in studies of f o r e s t damage in Erztgebirge. Mixed stands impose some f u r t h e r problems if one of t h e species shows symptoms while t h e oth- ers look healthy. The German r e p o r t s then t a k e t h e p e r c e n t a g e of t r e e s belonging t o t h e damaged species and a g r e e t h a t t h e same p e r c e n t a g e of t h e a r e a shall b e r e p o r t e d as damaged.

Owing t o t h e different usages in quantifying visible injury, t h e figures t h a t are r e p o r t e d

as

r e g a r d s t h e e x t e n t of damage a r e not p r o p e r l y com- parable. Especially, a distinction should b e made between t h e e x t e n t of tree and f o r e s t damage. The former

measures

t h e p e r c e n t a g e of trees dam- aged in t h e whole growing stock, while t h e l a t t e r gives a figure f o r t h e

"affected area", which a l s o contains healthy t r e e s . Hence, t h e figures given

at

t h e f o r e s t level give higher p e r c e n t a g e s than those concerning individual

trees

only. To solve these problems, i t i s important t o develop consistent ways of defining t h e d e g r e e of damage.

3.3. Growth reduction

Since damage as reduction

in

growth relates t o t h e economic value of t h e f o r e s t , i t i s of i n t e r e s t in p a r t i c u l a r f o r countries t h a t p r a c t i s e fores- t r y . I t may b e suitable f o r predictive purposes also,

as a

s e v e r e reduction in growth h a s been observed t o proceed t h e a p p e a r a n c e of visible symptoms (Figure 3).

mm mm

3.5

r ...

Symptoms of damage

1

^3.5

Figure 3. Growth development of healthy and diseased

trees

(Walder- krankung

...,

1983).

3.0

If damage i s understood

as

reduction in radial growth, i t i s not as sim- ple as in t h e c a s e of visible injury t o find a suitable norm. This i s because t h e normal annual growth increment of trees v a r i e s

as

a function of s e v e r a l f a c t o r s , even if t h e trees are growing under similar climatic and edaphic conditions. First, owing t o t h e g r e a t within-stand variation, i t i s unjustified t o r e f e r t o t h e normal growth of a n individual t r e e . Instead, a whole stand should b e chosen as a basic unit. Secondly, t h e r e i s variation from y e a r t o y e a r due t o t h e stochastic variation in weather. This can b e a v e r a g e d out by looking

at

t h e growth increments o v e r a period of time. Finally, since growth

rate

i s a g e dependent, t h e norm becomes a function of t r e e age. The norm i s hence a growth c u r v e f o r a whole stand showing t h e a v e r a g e time

- _-.-.-.-.

^Healthy

^-

Severely damaged 3.0

2.5

- -

^2.5

-

^2.0

.\

0.5

.,/.'.

,-.-.-

.-.

-

^I

-

0.5

development of volume p e r unit area. This norm is s p e c i e s and s i t e specific.

Figure 4 depicts such c u r v e s f o r some s p e c i e s and growing s i t e s in Finland (Koivisto, 1959).

STRND VOLUME DEVELOPMENT -.

^-r

0 . 4 ^L I

1. 2s. Y. 7s. am. a=. 1m.

AGE [ YEARS1

Figure 4. Stand volume development f o r some

tree

s p e c i e s and growing s i t e s in S o u t h e r n Finland according t o Koivisto (1959).

1 . S c o t s pine ( P i n u s s y l v e s t r i s ), MyrtiLLus t y p e 2 . Norway s p r u c e (Picea a b i e s ) , &rtiLLus t y p e 3 . Silver b i r c h (BetuLa p e n d u l a ) , &rtiLLus t y p e 4 . S c o t s pine ( P i n u s s y l v e s t r i s ^), W a d o n i a -type.

Detecting growth reductions may b e difficult because s m a l l changes easily remain embedded. T h e r e are two kinds of methods f o r detecting injury from empirical data. T r e e ring analysis provides a powerful method f o r detecting t r e n d s in diameter growth, and i t h a s a l r e a d y been applied t o

t h e s e a r c h of a i r pollutant injury (e.g. McLaughlin

et

al., 1984; Hari e t al., 1984). Data provided by f o r e s t inventories can also help t o assess possible injury (e.g. Arovaara

et

al., 1984).

Measuring growth damage r e q u i r e s comparing and ordering different growth curves. A s in t h e c a s e of measuring visible injury a t f o r e s t level, t h e r e is no unique way of defining a measure. One possibility is t o compare t h e average annual yields o v e r t h e rotation period in even-aged stands. In t h e National Acid Precipitation Assessment Program (NAPAP) c a r r i e d out in t h e United States, species specific growth rates are being determined

as

a function of t h e a i r pollution load (Rosental.1984).

4. RESPONSE OF FOREST TO AIR POLLUTION

4.1. Conceptual model: stress theory

Air pollutants a f f e c t trees simultaneously with o t h e r environmental stresses. Trees r e a c t t o this combination of

stresses

in a dynamic manner (e.g. Huttunen, 1975; Keller, 1978 ;Schiitt

et

al., 1983; Matzner and Ulrich, 1984). A conceptual framework of how plants respond t o environmental stress is provided by t h e concepts of s t r a i n and resistance, developed by Levitt (1972). S t r e s s is defined

as an

environmental f a c t o r capable of inducing

a

potentially injurious s t r a i n in living organisms. S t r a i n is any change in t h e normal behaviour of t h e organism, e i t h e r physical o r chemi- cal. If t h e s t r a i n can b e removed by removing t h e s t r e s s , i t is elastic.

Injury is equivalent with plastic o r irreversible strain. The s t r a i n imposed by a specific

stress

depends upon t h e resistance of t h e organism. A

meas- ure

f o r resistance is given by t h e magnitude of t h e lowest

stress

capable of

imposing plastic s t r a i n .

S t r a i n development and emergency of injury

are

dynamic processes.

For instance, both s t r a i n and resistance can change slowly under constant

stress,

and t h e outbreak of injury may b e delayed. Vice v e r s a , if t h e

stress is

removed, t h e s t r a i n need not disappear immediately, but i t may e i t h e r s t a y t h e same o r start decaying gradually o v e r time. To clarify t h e dynamics, Kauppi

⁽¹⁹⁸⁴⁾

distinguishes s t r a i n rate from accumulated strain.

In t h e s e t e r m s , a i r pollutant in t h e a i r could b e identified with stress, t h e pollutant gas entering plant tissue would b e s t r a i n , and t h e accumulated effects of this extraneous compound would become accumulated strain.

Hence, constant

stress

imposes

a

constant rate of accumulation of s t r a i n , and it is t h e accumulated s t r a i n t h a t is responsible f o r t h e potential injury.

Another dynamic response

is

t h a t resistance changes under strain.

Sometimes resistance increases, and then i t is

a

question of acclimatization t o a n unfavourable environment. This kind of behaviour seems likely if t h e plant -

as a

species - has gone through

a

similar environmental change repeatedly in t h e lapse of evolutionary time;

a

mechanism capable of minim- izing t h e disturbance c a n b e expected t o have developed.

Manion

(1981)

h a s classified different

stresses

on t h e basis of t h e i r

dynamic nature. If t h e

stress

immediately brings about injury, it is called

inciting. For instance, a c u t e necrosis of needles caused by

a

sudden expo-

s u r e t o

a

high sulphur dioxide concentration

is a

response t o a n inciting

stress.

A

s t r e s s t h a t does not cause immediate injury but

still

harms t h e

organism by e i t h e r accumulating s t r a i n o r decreasing resistance

is

called

predisposing.

A

predisposing

stress

can gradually give way t o

a

contribut-

ing

stress

factor: while t h e t r e e

is

exposed t o one stress f a c t o r ,

its

resis-

tance t o another f a c t o r c a n decrease. The final cause of t h e injury

is

hence not t h e primary stress f a c t o r , but t h e contributing f a c t o r instead. A s t a n d a r d example i s t h e invasion of insects t o a f o r e s t weakened by, say, drought o r f r o s t .

Resistance t o additional stress f a c t o r s such as a i r pollution a l s o v a r i e s geographically, owing t o variation in climate and edaphic f a c t o r s . Under ultimate conditions such as high elevations and n o r t h e r n areas r e s i s t a n c e t o additional stress f a c t o r s i s lower t h a n in more favourable conditions (e.g. Huttunen, 1975; Materna, 1979; Materna, 1983). In terms of Manion's dynamic stress concepts, t h e h a r s h n a t u r a l conditions could b e understood as predisposing stress t h a t d e c r e a s e s r e s i s t a n c e t o inciting a i r pollutant stresses.

4.2. Quantitative models

4.2.1. Quantification of stress theory

Although t h e stress t h e o r y seems t o provide a good conceptual model f o r understanding how

trees

respond t o a i r pollution, t h e quantification of t h e t h e o r y f o r t h e purpose of prediction i s r e s t r i c t e d by t h e f a c t t h a t t h e concepts involved d o not h a v e a s t r i c t physical i n t e r p r e t a t i o n . F i r s t , s t r a i n and injury are defined as changes in "normal" behaviour which, owing t o variation in both inherited p r o p e r t i e s and t h e environment, generally con- s i s t s of a l a r g e a n d r a t h e r a fuzzy

set

of individual behaviours. To decide what i s meant by a normal, healthy

tree

t h e r e f o r e always involves a n a r b i - trary choice.

Secondly, since t h e concepts are v e r y aggregate, i t may b e difficult t o isolate a c e r t a i n physical entity as a basis f o r a quantitative i n t e r p r e t a t i o n . In a n attempt t o d o this, Taylor (1978) introduces a n i n t e r p r e t a t i o n of t h e

concepts in t h e case of gaseous

air

pollutant

stress. Stress

i s understood

as

t h e presence of

a

pollutant gas in t h e immediate neighbourhood of plant tissue, and strain i s a result of t h e penetration of t h e extraneous substance into t h e tissue. However, this strain may launch a whole chain of events, where t h e strain may be counteracted by a resistance mechanism, this coun- teraction imposing a secondary strain that

w i l l

be counteracted by a secon- dary resistance t o yield a t e r t i a r y strain, and s o on. For instance, t h e resistance mechanism against an extraneous substance in t h e plant tissue may involve some e x t r a energy consumption which, in turn, imposes a secondary strain through t h e deficiency of metabolic products. If all these stages a r e isolated, s o many strain and resistance variables a r e obtained t h a t it hardly

seems

reasonable t o build

a

predictive model upon them.

Further, t h e dynamic interaction of strain and resistance processes makes it difficult t o derive more aggregate variables from t h e basic processes

in

an adequate way.

The approach presented by Kauppi

(1984)

f o r t h e analysis of tempera-

t u r e s t r e s s on pine seedlings starts off with aggregate variables t h a t

are

given a n operational definition through measurements. If t h e

stress

theory

is

t o be utilized, such an approach

seems

applicable f o r pollutant

stress

also. However, t h e

stress

theory still

seems

t o have

its

major contribution

in t h e conceptual understanding of pollutant impacts, while t h e existing

models intended f o r prediction

are

based on different concepts and metho-

dology.

4.2.2. D o s e and response

P e r h a p s t h e simplest attempt t o quantify t h e complicated impact of a long-lasting pollutant

stress

on a biological object i s t h e so-called dose- response model. f i r s t presented already in 1922 by O'Gara, and f u r t h e r developed by, e.g.,Thomas and Hill (1935), Guderian

et

al. (1960), Zahn (1963) and Larsen and Heck (1976). This model h a s been based on t h e idea t h a t damage f i r s t o c c u r s a f t e r exposure t o a threshold dose:

when

D > D o

damage o c c u r s (1)

The dose depends both on pollutant concentration and exposure time.

In t h e s t a n d a r d form of t h e model, relationship (1) h a s a v e r y special form: i t i s t h e product of t h e "effective" concentration and t h e "effective"

exposure time, "effective" r e f e r r i n g t o t h e e x c e s s o v e r a threshold time o r concentration under which damage never occurs. Hence, t h e dose ( D ) i s

where c i s concentration,

t

i s exposure time and c o and t o are t h e respec- tive threshold values. According t o t h e model, damage can b e brought about by all t h e combinations of c and

t

t h a t make t h e product

D

exceed t h e threshold

Do.

An n-fold i n c r e a s e in c implies t h a t t h e time r e q u i r e d f o r developing damage d e c r e a s e s by a f a c t o r of

-.

1

n

A criticism t h a t h a s been laid on t h e model i s due t o t h e observation t h a t t h e threshold dose i s not constant but depends on t h e exposure time.

Therefore, attempts have been made t o define t h e dose in a n o t h e r way such t h a t a constant threshold value could b e observed. F o r instance, Guderian

et

al. (1960) have suggested t h e following model:

which h a s fitted d a t a b e t t e r t h a n t h e f o r m e r one. According t o t h i s model, a n n-fold i n c r e a s e in c will imply t h a t t h e time r e q u i r e d f o r developing dam- a g e c a n d e c r e a s e more than by a f a c t o r of

-.

1

n

Another version of t h e model with a more quantitative objective p r e d i c t s t h e d e g r e e of damage as a l i n e a r function of t h e dose received (e.g. Thomas & Hill, 1935). Denoting t h e d e g r e e of damage by y , this c a n b e e x p r e s s e d as follows:

A s opposed t o Model (1) where damage i s a binary v a r i a b l e with values

"damage o c c u r s " and "damage does not occur", t h i s model provides t h e d e g r e e of damage as a continuous variable.

L a r s e n and Heck (1976) have f u r t h e r extended t h e idea of dose and response in t h e i r lognormal model, which includes t h e conventional dose- response model as a d e g e n e r a t e case. In t h i s model dose i s defined as

where t h e exponent p h a s been observed t o g e t values within t h e interval 0 <p U (in t h e dose-response model p =I). Degree of damage i s e x p r e s s e d

continuously as a function of dose, such t h a t t h e following conditions

are

fulfilled:

[I] Degree of damage i s proportional t o t h e logarithm of dose, r a t h e r t h a n t h e arithmetic value of t h e d o s e itself.

[ 2 ] T h e r e is a r a n g e of doses s o low t h a t no response i s measured.

[3] T h e r e is a maximum response t h a t will not b e exceeded even if t h e dose is increased.

The function h a s been formulated with t h e aid of t h e s t a t i s t i c a l lognor- mal distribution. A two-variable function l i n e a r in lnc and lnt h a s been obtained, t h e p a r a m e t e r s of which have been estimated on empirical basis utilizing t h e definitions of t h e lognormal distribution. Larsen and Heck (1976) have identified t h e model f o r 1 4 mainly agricultural s p e c i e s under ozone exposure and f o r 4 s p e c i e s under sulphur dioxide exposure. Degree of damage h a s been defined as percentage of leaf necrosis. The r e s u l t s

are

promising.

A s t h e dose-response model does not consider s t r a i n and r e s i s t a n c e as explicit variables, potential differences in r e s i s t a n c e between species and growing s i t e s should b e incorporated in t h e p a r a m e t e r s of t h e model. The applicability of t h e model hence largely depends upon t h e possibilities of identifying t h e p a r a m e t e r s as functions of f a c t o r s affecting stress resis- tance.

4.2.3. Growth damage

While t h e method of t h e above models i s t o p r e d i c t d a m a g e directly, t h e following a p p r o a c h will yield as output g r o w t h u n d e r air p o l l u t i o n

.

The method i s built upon a simple dynamic growth model, t h e p a r a m e t e r s of which are functions of t h e a i r pollutant level. G r o w t h d a m a g e in t h e s e n s e discussed in Section 3 i s obtained when t h e model output is compared with a selected norm.

The model describes t h e time development of t h e d r y weight of foliage, W (kg p e r h e c t a r ) , by means of a differential equation:

where J" (W) is a shading function with t h e p r o p e r t y

i.e. as the foliage grows, i t gradually c r e a t e s a n increasing shade t h a t decreases productivity. Further,

X

i s t h e fraction of growth increment allocated t o foliage, cp is t h e maximum net production of d r y weight p e r d r y weight of foliage p e r y e a r , and @ i s t h e foliage death r a t e .

This is a general form of a class of models t h e representatives of which can be distinguished on t h e basis of how they actually define t h e shading function

J"

(W) (e-g. dlgren, 1983; Kohlmaier

et

al., 1984; M&kela and Hari, 1984). The applicability of t h e model t o t h e analysis of a i r pollutant impacts on f o r e s t s is based on t h e f a c t t h a t i t includes as parameters t h e annual specific photosynthetic r a t e , q (kg carbon p e r foliage a r e a p e r y e a r ) and t h e senescence

rate

of t h e foliage, .)(I (kg senescent foliage p e r kg total foli- a g e p e r year), which are among t h e most important f a c t o r s a i r pollution directly affects. By modelling t h e impact of a i r pollutants on t h e s e parame-

ters

t h e effects on canopy growth can be analyzed. Such a n approach has been suggested f o r instance by MZkelZ

et

al. (1981), K e r c h e r & Axelrod (1981) and Kohlmaier

et

al. (1984). Agren and Kauppi (1983) applied a simi- l a r method f o r predicting possible changes in growth due t o

an

anthropo- genic nitrogen input t o t h e soil.

Kohlmaier e t al. (1984) discussed t h e behaviour of t h e equilibrium of t h e model under a pollutant load. The relative growth rate of t h e canopy is

The system has an equilibrium, say

W o ,

when R=O. P r o p e r t y (7) implies t h a t t h e equilibrium is stable, but its actual form largely depends upon t h e shading function j'(W).

I t is easy t o show on t h e basis of (7) t h a t

Hence, if e i t h e r specific n e t photosynthesis d e c r e a s e s o r specific senes- cence rate increases, a s is assumed t o happen under a permanent pollutant load, t h e maximum sustainable biomass decreases. This d e c r e a s e can b e quantified with t h e aid of t h e model, provided t h a t t h e response of rp and $ t o a i r pollutants is known.

In general, i t is not only t h e equilibrium, but also t h e o t h e r dynamic p r o p e r t i e s of t h e system t h a t t h e model applies to. Hence i t provides a means f o r analyzing t h e change in productivity as a response t o pollutant impacts on photosynthesis and senescence. The use of t h e model f o r predic- tion involves, however, s e v e r a l problems. First,

an

extensive modelling r e q u i r e s t h a t t h e values of t h e parameters a r e known o v e r a range of species and environmental conditions. Even though c e r t a i n d a t a are avail- able f o r t h e estimation, i t

seems

r a t h e r unlikely t h a t a satisfactory cover- a g e can b e achieved.

Secondly, t h e model does not provide any information on stemwood.

The problem condenses in t h e partitioning of growth. If t h e partitioning coefficients were time constants, then stem growth would b e simply a con- s t a n t times t o t a l growth, i.e. A, ^'j( W) W, A, denoting t h e partitioning coeff i- c i e n t of stemwood. T h e r e i s s t r o n g evidence, however, t h a t A, i s not a con- s t a n t but depends on s e v e r a l f a c t o r s , including t h e size of t h e foliage, W.

A s t h e partitioning coefficients are interconnected, a l s o A, i s a function of W. This feedback, when incorporated in t h e model, c a n totally change i t s dynamic behaviour.

A more comprehensive model of growth under a i r pollution h a s been developed by Bossel

et

al. (1984a) (system BAUM). Though f a r more detailed, t h e model i s similar t o t h e above-described one in t h a t i t d e r i v e s growth from physiological p r o c e s s e s , such as photosynthesis and leaf senes- c e n c e r a t e , and considers t h e impact of a i r pollutants on t h e s e processes.

The model i n c o r p o r a t e s leaf, r o o t , wood and assimilate biomass as s t a t e variables, and accounts f o r s e v e r a l physiological p r o c e s s e s involved in

tree

growth. The "normal" driving v a r i a b l e s comprise light conditions, day length, a v e r a g e t e m p e r a t u r e and soil nutrient status. Five important a i r pollutants are included, using empirical evidence on t h e i r e f f e c t s on t h e physiological processes. The

tree

m o d e l is f u r t h e r linked with m o d e l s describing t h e

state

of t h e soil, which allows soil acidification t o b e included in t h e dynamics. The a u t h o r s emphasize, however, t h a t t h e model should b e r e g a r d e d as a n attempt t o qualitatively understand t h e p r o c e s s e s of pollutant impacts, r a t h e r t h a n as a quantitative, predictive tool.

4.3. Empirical evidence

Empirical r e s u l t s have been published as r e g a r d s t h e relationship between a i r pollution and f o r e s t damage. Some of them are reviewed h e r e , with t h e objective of giving a n idea of what kind of v a r i a b l e s and r e s u l t s are available f o r model identification.

Wentzel (1983) h a s collected t h e r e s u l t s concerning t h e relationships between SO2 concentrations and t h e damage observed t h a t h a v e been pub- lished

in

Europe. These studies are p a r t of t h e IUFRO program Air Pollu- tion. Figure 5 shows a summary of t h e s e results. The c u r v e s show t h e S O Z immission concentrations of various c e n t r a l European forest-damage areas as connecting c u r v e s between a c t u a l measured annual a v e r a g e values and peak values. The numbers on t h e c u r v e s denote t h e r e s p o n s e s of t h e f o r e s t s , as t h e y are r e p o r t e d in t h e quoted l i t e r a t u r e .

Materna (1984) h a s made some observations on t h e threshold times and concentrations as functions of t h e t o l e r a n c e of t h e f o r e s t . He notes t h a t a good measure f o r t h e t o l e r a n c e i s altitude from sea level. These r e s u l t s a r e shown in Table 1.

Guderian

et

al. (1984) h a v e c a r r i e d o u t a profound study in o r d e r t o establish some dose-response relationships f o r t h e response of f o r e s t s t o ozone. Tolerance i s i n c o r p o r a t e d by assuming t h a t t h e threshold v a r i e s as a function of s e v e r a l v a r i a b l e s affecting tolerance, e.g. t h e latitude of t h e f o r e s t , edaphic and climatic f a c t o r s , e t c .

SO, Field Measurements in European lmmission Areas

Klo

w

²⁰⁰ 300 500 1OOOyg

pg SO2 l m 3 (LOGARITHMIC CLASSIFICATION) Wbntzel I980

Figure 5. Relationships between SO2 concentrations and t h e damage ob- s e r v e d according t o various s o u r c e s (Wentzel, 1983).

1. Materna 1 9 7 3 : High r a n g e s of t h e Erzgebirge

-

^resis-

t a n c e t o f r o s t and o t h e r secondary injuries diminished.

2. Materna 1972 : 20% loss of wood increment in t h e E n g e - birge.

3. Materna

et

al., 1969 : Dieback of Norway s p r u c e stands in t h e Bohemian p a r t of t h e Erzgebirge.

4. Wentzel 1979 : Good s i t e s in Rhein-Main area sufficiently p r o t e c t e d .

5. Stein a. Dassler 1968 : Moderate injury in t h e Saxonian p a r t of t h e Erzgebirge.

Lux 1976 : Situation comparable with t h a t in Neiderlau- sitz.

6 . Knabe 1970 : Wentzel 1 9 7 1 : S e v e r e growth losses in t h e Ruhr area.

7. Knabe 1972 : Economical f o r e s t r y with Norway s p r u c e and S c o t s pine made impossible in t h e Ruhr

area.

8. Guderian a. Stratmann 1968 : Immission t y p e Bierstorf

=

single inmission s o u r c e in mountain valley. S e v e r e growth damage t o n e a r l y all species (young plants).

A relationship between t h e number of living needle a g e classes and sul- p h u r content of t h e needles, found out f o r instance by Knabe (1981), pro-

Table 1. The time between t h e beginning of a i r pollution influence and t h e desintegration of Norway s p r u c e stands.

h e i g h t above

vides some empirical basis f o r determining t h e parameter values in Equa- tion (6). (Figure 6). Similar r e s u l t s are available f o r o t h e r pollutant species also (e.g. Knabe, 1984).

The response of t h e potential photosynthetic r a t e , q ,

to

pollutant expo- s u r e h a s been studied in laboratory and field experiments by

a

number of authors (e.g. Keller, 1978; Katainen

et

al., 1983).

5. STRATEGIES TOWAEtDS FOREST IMPACT SCENARIOS

5.1. Preliminaries

thc

The aim of this r e p o r t is t o s e a r c h f o r a model s t r u c t u r e t h a t can b e used in producing scenarios of f o r e s t damage due t o a i r pollution in Europe.

-600

-

^{900 -1050 , 1050+} sea l e v e l

50-60

40-50 30-40

20-30

20-30

2ox 10-15

< 10

30-40

2ox

loX

2ox

Figure

0 up to 2 measurement points ( 6 trees) at least 3 measurement points

Sulfur content of l-year old spruce needles

6. Relationship between sulfur content of l - y e a r old s p r u c e nee- dles and t h e total amount of needles on t h e

tree

expressed as a percentage of t h e l - y e a r old needle class.

The preceding sections show clearly t h a t t h e problem w e want t o solve quan- titatively, i s

at

present r a t h e r poorly understood even qualitatively.

Therefore, a stepwise s t r a t e g y i s p r e f e r a b l e which allows gradual improve- ment of t h e model in pace with increasing knowledge. Secondly, i t is of par- ticular importance t h a t t h e sources of uncertainty a r e understood, and means of increasing t h e reliability of t h e r e s u l t s a r e constantly developed.

5.2. Input

By Section 2, t h e evidence does not allow t h a t any one of t h e reviewed potential causes of damage is excluded. On t h e c o n t r a r y , i t has been argued t h a t several simultaneous mechanisms contribute t o t h e damage, and t h a t

t h e impact is deeply dynamic. The synergetic impacts a r e briefly under- stood, though, and t h e i r quantification requires a lot of guesswork. I t

is

therefore considered more realistic t o start with

a

model based on one pol- lutant at

a

time. The consequent results will obviously be applicable t o a r e a s where one pollutant dominates, but they can also be used t o study how much this one factor can explain.

In Section

1,

t h e suitability of sulphur dioxide concentration a s an input variable was questioned.

A s

mentioned in Section

2, it

has been claimed t h a t sulphur cannot be t h e main cause of damage unless t h e pathway through soil acidification

is

considered. This would require considering

a

delayed damage. According t o s t r e s s theory, however, t h e delay can also be connected with t h e so-called direct impacts of sulphur dioxide, and this should therefore be regarded a s one of t h e relevant potential causes.

In a summary,

it is

suggested that the

first

stage of the

m o d e l

should be based on airborne sulphur dioxide.

A s

a f u r t h e r step, soil pH and airborne ozone should be considered.

5.3. Output

Model output should b e simple enough t o be projectable on a map.

A

simple classification of t h e output forest state

is

t h e r e f o r e preferable.

According t o Section

3,

visible injury

is

t h e most widely documented

type of forest damage in Europe and hence provides the most realistic out-

put variable f o r t h e scenario

model.

I t should b e possible a t least t o

separate between two classes, injured and non-injured.

Since growth reductions may have economic significance even b e f o r e visible injury a p p e a r s , i t would b e informative if t h e c l a s s "hidden injury"

could be distinguished. Another interesting border-line both economically and esthetically, i s t h a t between productive and non-productive land.

In a summary, a possible classification of damage i s t h e following:

[I] no damage

[2] hidden damage, d e c r e a s e d productivity possible [3] visible damage, d e c r e a s e d productivity

[4] disintegration, z e r o productivity

In terms of t h e dynamic behaviour of t h e system, c a t e g o r i e s [I]-[3]

r e p r e s e n t s t a b l e development, although classes [2] and [3] show reduced growth, whereas c a t e g o r y [4] c o r r e s p o n d s t o unstable behaviour. There- f o r e , an important b o r d e r l i n e i s t o distinguish c l a s s [4] from t h e o t h e r s .

5.4. Model structure

A s t h e model is intended f o r interactive use, a n important c r i t e r i o n i s t o keep t h e simulation time s h o r t enough. T h e r e f o r e , a multi-variable dynamic model cannot b e chosen. F u r t h e r , as w a s s e e n in Section 4. t h e r a t h e r simple models based on t h e o r e t i c a l argument are not suitable f o r extensive quantitative use, because t h e y e i t h e r have some t h e o r e t i c a l discrepancies, o r t h e y cannot b e reasonably t e s t e d against data. Hence a more o r less empirical, a g g r e g a t e d a p p r o a c h h a s t o b e taken, generalizing t h e implications of some individual d a t a sets and t h e o v e r a l l understanding of t h e f o r e s t damage p r o c e s s t o c o v e r a l a r g e area and a long period of time.

However, as t h e r e i s relatively few d a t a available f o r quantifying input-output relationships between a i r pollution and f o r e s t r e s p o n s e , some t h e o r e t i c a l understanding of t h e p r o c e s s would b e helpful in constructing t h e above-mentioned model. Models describing t h e growth dynamics of trees under a i r pollutant load c a n w e l l contribute t o such information, integrating t h e c u r r e n t

state

of knowledge about individual r e s p o n s e p r o c e s s e s t o t h e r e s p o n s e of t h e whole

tree.

5.5. Evaluation of uncertainty

The uncertainty of t h e r e s u l t s of a n empirical model i s connected with problems of generalization. In t h i s c a s e , both geographic and temporal extrapolation i s necessary. If uncertainty i s due t o s t a t i s t i c a l unreliabil- ity, i t c a n b e d e c r e a s e d by increasing t h e size of t h e sample. This way f o r instance t h e growth and yield tables developed in f o r e s t r y h a v e become successful and p r a c t i c a l . However, e a c h experiment t a k e s a long time up t o s e v e r a l decades, and varying t h e input under field conditions i s almost impossible.

A more difficult p a r t of t h e uncertainty originates in t h e possibility t h a t not all t h e important f a c t o r s have been included in t h e model s t r u c -

ture.

Secondly. t h e dynamics of t h e impact may b e more complicated t h a n assumed, f o r instance some delay s t r u c t u r e s may not have become a p p a r e n t by t h e data. When t h e model i s applied t o conditions different from t h e d a t a , e r r o n e o u s conclusions c a n b e made if t h e s e possibilities are ignored.

T h e r e i s no g e n e r a l method f o r assessing t h i s kind of uncertainty. One possibility i s t o analyze how t h e r e s u l t s would change if t h e simplifying hypotheses were substituted with some more r e a l i s t i c ones. Obviously such

a n analysis can only c o v e r p a r t of t h e problem

at

a time, since if i t w a s pos- sible t o

treat

t h e whole problem t h i s way, why should t h e model simplify

at

all? For instance, t h e dynamic behaviour of t h e model delineated h e r e could b e compared with t h a t of t h e mechanistic models presented in Section 4, t o assess t h e r e s u l t s qualitatively o v e r limited areas. This would also contribute t o t h e f u r t h e r development of f o r e s t impact scenarios, espe- cially as r e g a r d s synergistic e f f e c t s of a i r pollutants.

I t i s suggested t h a t modelling f o r e s t impacts p r o c e e d s from t h e sim- plest model towards more difficult t a s k s which can b e built upon t h e former steps. Each s t a g e utilizes a n aggregated input-output model with geographi- cally distributed p a r a m e t e r s describing

stress

resistance. Such a model s t r u c t u r e i s depicted in Figure 7. This model will b e constructed on t h e basis of empirical d a t a and t h e o r e t i c a l understanding of t h e dynamics of t h e process. More detailed "mechanistic" models will b e used t o e x t r a c t infor- mation of f o r e s t damage dynamics t o t h e a g g r e g a t e input-output model. The consequent s t e p s , specified by t h e choice of input and output, are listed

in

Table 2. The stepwise p r o c e d u r e i s shown schematically in Figure 8. Evalua- tion of uncertainty and f u r t h e r s t e p s and questions

are

a l s o indicated.

Table 2. Model inputs and outputs to be applied in a stepwise manner as Al, A2,

...,

^B3.

R (x) Spacedistributed resistance parameters

i n p u t o u t p u t

Input EMPIRICAL Output

U(X, t ) I NPUT-OUTPUT Y(X. t)

MODEL Damage

(stress rate)

A, ground c o n c e n t r a t i o n of SO2 ( a n n u a l a v e r a g e + s t a t i s t i c a l

d a t a )

B. ground c o n c e n t r a t i o n o f ozone ( a n n u a l a v e r a g e + s t a t i s t i c a l

d a t a )

Figure 7. Structure of forest impact submodels.

I. b i n a r y - v a l u e d : damage

-

no damage

2. m u l t i - v a l u e d : s e v e r a l s t a g e s o f damage

3 . c o n t i n u o u s , b a s e d on growth r e d u c t i o n

DATA

Figure 8. Suggested procedure f o r building f o r e s t impact submodels.

FITTING MODEL TO DATA

MODELLING

RESISTANCE, CLIMATE & SO1 L

OVER EUROPE I = O

GATHER*

QUALITATIVE INFORMATION

POLLUTION u ( x . t ) DAMAGE

y (x, t )

USE MODEL ANALYZE

UNCERTAINTY, EVALUATE

SKETCH FURTHER STRATEG l ES

REFERENCES

Agren, G.I. 1983. Model analysis of some consequences of acid precipitation on f o r e s t growth. In: Ecological Effects of Acid Deposition. National Swedish Environment P r o t e c t i o n Board

-

R e p o r t PM 1636: 233-244.

Agren, G., and P . Kauppi. 1983. Nitrogen dynamics in European f o r e s t ecosystems: considerations r e g a r d i n g anthropogenic nitrogen deposi- tion. IIASA CP-83-28.

Alcamo, J . , P . Kauppi, M. Posch and E. Runca. 1984. Acid Rain in Europe: A Framework t o Assist Decision Making. IIASA WP-84-32.

Arovaara, H., P . Hari and K. Kuusela. 1984. Possible e f f e c t of changes in atmospheric composition and acid r a i n on tree growth. An analysis based on t h e r e s u l t s of Finnish National F o r e s t Inventories. Communica- tions Instituti Forestalis. Fenniae. p. 16.

Bossel, H., D. G o c k e r t , H. K r i e g e r , W. Metzler and H. S c h a f e r . 1984a. Simu- lation d e s Systems BAUM u n t e r Schadstoffbelastung. In: Bossel et al.

1984b.

Bossel, H., W. Metzler and H. S c h a f e r . (Ed.). 1984b. Dynamik d e s Waldster- bens. Matematisches Model1 und Computer simulation. Interdisziplinare Arbeitsgruppe Mathematisierung (IAGM). Gesamthochschule Kassel. p.

267.

Forsurning Idag Och I Morgon. En Svensk Studie Infor Stockholms- Konferensen 1982 om Forsurningen a v Miljon. J o r d b r u k s d e p a r t e m e n t e t , Kommitten Miljo 1982.

Guderian, R. 1977. Air Pollution Phytotoxicity of Acidic Gases and its signi- ficance in a i r pollution control. Ecological Studies Vol. 22. S p r i n g e r - Verlag, B e r l W H e i d e l b e r g , 1 2 7 pp.

Guderian, R. (ed. ). 1985. Air Pollution by Photochemical Oxidants. Forma- tion, t r a n s p o r t , c o n t r o l and e f f e c t s on plants. (Ecological Studies.

Analysis and Synthesis, Vol. 52). Springer-Verlag.

Guderian, R., D.T. Tingey and R. Rabe. 1984. Effects of photochemical oxi- dants on plants. In: Guderian, R. (Ed.) Air pollution by photochemical oxidants. Formation, t r a n s p o r t , control and e f f e c t s on plants. Ecologi- c a l Studies 52, S p r i n g e r Verlag Berlin, Heidelberg, N e w York.

Guderian, R., H. van Haut and H. Stratmann. 1960. Probleme d e r Erfassung und Beurteilung von Wirkungen gasformiger Luftverunreinigungen auf die Vegetation. Z. Planzenkrankh. Pflanzenschutz 67257-264.

Hari, P . , H. A r o v a a r a , T. Raunemaa and A. Hautojarvi. 1984. F o r e s t growth and t h e e f f e c t s of e n e r g y producton: a method f o r detecting t r e n d s in t h e growth potential of trees. Can. J. For. Res. 14:437-440.

Huttunen, S. 1975. The influence of a i r pollution on t h e vegetation around Oulu. Acta Univ. Ouluensis. 33.

Kamari, J. 1984. Methods f o r predicting f r e s h w a t e r acidification. IIASA WP-84-27.

Kotainen, H.-S., T. Pohja, E. Maki, P. Hari and S. Kellomaki. 1983. Rikki- dioksidin vaikutus fotosynteesiin. (Impact of sulfur dioxide on pho- tosynthesis). In: Energiantuotannosta peraisin olevien ilman epapuhtauksien vaikutus metsan tuotokseen (Effect of pollutants from e n e r g y production o n f o r e s t productivity - in Finnish). Helsinki University, Dept. of Silviculture.

Kauppi, P . 1984. S t r e s s , s t r a i n and injury: S c o t s pine t r a n s p l a n t s from lift- ing t o acclimation on t h e planting site. Acta Forestalia Fennica 1 8 5 : l - 49.

Kauppi, P., M. Posch, E. Matzner, L. Kauppi and J. Kamari. 1984. Acidifica- tion of f o r e s t soils: model development and application f o r analysing impacts of acid deposition in Europe. IIASA CP-84-16.

Keller, Th. 1978. Frostschaden a l s Folge e i n e r "latenten" Immissionsschadi- gung. Staub 28:24-26.