Modelling tropical deforestation and its consequences for global climate

Ecological Modelling, 58 ( 1991 ) 217-247 Elsevier Science Publishers B.V., Amsterdam

217

Modelling tropical deforestation and its consequences for global climate Jan Rotmans and Robert J. Swart Netherlands National Institute Jbr Public Health and Environmental Protection (RIVM), P.O. Box 1, 3720 BA Bihhoven, Netherlands (Accepted 26 March 1991)

ABSTRACT Rotmans, J. and Swart, R.J., 1991. Modelling tropical deforestation and its consequences for global climate. Ecol. Modelling, 58: 217-247. Major land use changes can influence bio-geochemical cycles. Presently the conversion of tropical rainforests is considered to be the most important type of change in this respect. For use in connection with the carbon cycle module of RIVM's Integrated Model to Assess the Greenhouse Effect (IMAGE) a deforestation module has been developed. The purpose of this module is to relate deforestation rates to socioeconomic developments in order to develop long-term scenarios. Next to the major driving force: expansion of agricultural land, also expansion of pasture, commercial logging, fuelwood gathering, industrial and mining projects have been taken into account. Because of the close interrelation between growth of population and economies and the demand for agricultural products, energy and building materials, we find that the rapid destruction of the forests can only be halted by sharp increases in the productivity of agriculture and forestry. Plantation wood, as a substitute for fossil fuels, has to be used efficiently. Land expansion should take place according to land suitability for agricultural land, possibly even at the expense of pasture lands. Since data on deforestation and the underlying processes are very uncertain and at this stage in the model the tropical forests are aggregated into the three major regions the results can only be considered in a relative way. It is found that deforestation plays a significant role in the greenhouse effect, but much less than the combustion of fossil fuels.

1. INTRODUCTION

The world's tropical forests presently face tremendous pressures from increasing demands of increasing populations. More than 10 million hectares of closed tropical rainforests are destroyed annually and an equal amount is severely altered. The results are the extinction of species, increased erosion, threatening of indigenous people, the modification of regional and even global climate, and the destruction of a wide variety of possibly economically important assets. 0304-3800/91/$03.50

© 1991 - Elsevier Science Publishers B.V.

218

J. ROTHMANSAND R.J. SWART

Fig. 1. Integrated Model for the Assessment of the G r e e n h o u s e Effect ( I M A G E ) .

TROPICAL DEFORESTATION AND ITS CONSEQUENCES FOR GLOBAL CLIMATE

219

RIVM has developed the Integrated Model to Assess the Greenhouse Effect (IMAGE) as a scientifically based policy and demonstration instrument (Rotmans, 1990; Rotmans et al., 1990). The model links socio-economic causes (energy consumption, industry, agriculture) to global effects (global mean temperature and sea level rise). For the Netherlands a number of particular socio-economic impact modules has been added (den Elzen and Rotmans, 1991), dealing with coastal protection and water management (Fig. 1). The pivotal part of IMAGE is the carbon cycle, based on Goudriaan and Ketner (1984). This dynamic model has an ocean and a terrestrial biosphere part as is shown in Fig. 2. IMAGE has been used for policy analysis at the national level (Langeweg, 1988) and internationally (IPCC, 1990). Recently IMAGE has been used for determining Global Warming Potentials (Lashof and Rotmans, 1990, unpublished data). Land use change is one of the human activities influencing global change. It can be manipulated in the model by changing the elements of a land use transfer matrix, describing shifts between ecosystems. As a first step towards a global land use change module a tropical deforestation module was developed to be linked to this matrix in IMAGE. A prototype of this

decomposition of organic material

BIOSPHERE Fig. 2. Schematic of the IMAGE terrestrial biota model (Goudriaan and Ketner, 1984).

220

J ROTHMANS AND R.J. SWART

deforestation module was described by Swart and Rotmans (1990). Because the climate effect is only one among several other (possibly more important) effects of deforestation, the module can also be used as a demonstration tool of its own accord. To take other land use changes into account, like desertification, forest dieback, wetland drainage and extended lowering of groundwater tables, the module has to be upgraded in the future. In this article the module and its results will be discussed. First the model structure will be described, after which the different processes will be considered. Finally conclusions will be drawn and recommendations made. 2. SCOPE Because of the different characteristics of the areas in the module the three major tropical forest areas are considered separately: Latin America, Africa and South-East Asia. Although we realize that for a more reliable study on deforestation the module should have been disaggregated to a national or even district scale, the limited availability of data and the scope of the IMAGE-project prevented us from doing so. Therefore we do not intend to present reliable projections of forest resources at this level for the near future, but are primarily interested in the long-term possibilities for different types of development. IMAGE generates time series from 1900 to 2100. For this paper, four scenarios were developed: (A) unrestricted trends, (B) reduced trends, (C) changed trends and (D) forced trends. Differences between the scenarios are mainly based on different environmental viewpoints. From A to D increasing national and international awareness causes increasing protective measures (Swart and Rotmans, 1989). In scenario A economic growth, is not restricted by environmental concerns. Short time profits prevail over long-term assets. For tropical forests this means: a continuing exploitation as if tropical forest were 'mined' as quickly as possible. This means high estimates for logging, agricultural consumption and industrial projects. Because in such a scenario short term economic growth can be considered to generate finances enabling the use of agricultural inputs to increase productivity, the scenario can be expected not to show an exponential trend. The scenarios B and C are meant to describe two levels of developments in which forest destruction is limited by increasing use of the long-term economic potential, as for instance planned in some recent initiatives as the International Tropical Timber Agreement (ITTA) and FAO's Tropical Action Plan. Scenario B is assumed to reflect the effect of the ITTA, whereas in scenario C the Tropical Action Plan of the FAO can be considered to be implemented (Food and Agricultural Organization, 1986). Finally, scenario D is intended to describe the preservation of the forests for ecological reasons mainly. Scenario D assumes pure conservation of the forest in

T R O P I C A L D E F O R E S T A T I O N A N D ITS C O N S E Q U E N C E S FOR G L O B A L C L I M A T E

221

their primary state, even if this means, that the economic value of the area would not be used to its full extent. In this scenario, for instance, production of hardwood and meat and dairy is limited in the tropical zones and reforestation is important. To simulate scenarios the parameters in the module are varied: demand for wood for construction, fuel wood, energy and minerals, agricultural land and pasture, etc. Population growth is taken common for all scenarios. Economic development is not included explicitly, since we believe that there is no direct easy linkage between economic growth and land use change, and that long-term land use change will be driven by the demand for land and products whatever the costs. The details of the different scenarios are described in the appendix. 3. D E S C R I P T I O N O F T H E D E F O R E S T A T | O N

MODEL

The deforestation module is an independent submodule within the carbon cycle module. In Fig. 3 the structure of the deforestation module is shown. The module distinguishes four types of ecosystems that are important for the areas concerned: tropical forests, grasslands, arable land and other land. In this study the tropical forest system was split into closed and open forest as defined by Lanly (1982). The ecosystems have different carbon densities for their different components (vegetation, litter, soil humus). Intermediate types of forests (secondary forests, forest fallow) were included in the 'open forest' system. We assume that all converted forests end up as agricultural or pasture land, degraded area, or as reestablished forests (by reforestation). Deforestation is triggered by a variety of processes, which are mainly caused by a number of demands driven by growth of population and economy. In Table 1 an enumerative description is given of all ecosystems and processes. Land transfers are described by differential equations, while processes are represented by algebraic dynamic functions. The time indication in these equations or functions is represented as follows: ECOSYSTEMj(t) or PROCESSj(t) denotes the a m o u n t of ecosystem or process j at time t, whereas ECOSYSTEMj(0) or PROCESS~(0) denotes the initial amount of ecosystem j or process j, at time t = 0, in 1900. As can be seen in Fig. 3 some simplifications have been made. Fuelwood gathering is only taken into account for open forests, assuming, that in closed forest areas the natural production capacity surpasses local demand (de Montalembert and Cl6ment, 1983). For the regions considered here, forest plantations are assumed to develop on other lands rather than on agricultural, grass or forest lands. Only in Europe or North America reforestation may take place on abandoned agricultural land or pasture. Logging is assumed to take place in closed forests only, degrading them into secondary (here open) forests. In the following sections these processes will be discussed. To

222

J. R O T H M A N S A N D R J S W A R T

I PoPo~'T'o"t--I I P~ooU~,'o~J l NU~"OF CATTLE h I CEREAL L-. CONSUMPTION/ I

~t FEED/FOODI

I CATTLE I DENSITY

SELF SUFFICIENCYI ~ -'--INoN-CEREALS I

t,

~NDDEMANDI DE~RADATIONI FOR FOOD I PRODUCTION I

ARABLELAND I

i

IDE~RADATIONI

DEMAND FOR PASTURE

LAND

I PASTURE I I LAND I

.............. o.... I

I

.

.....

1 "IDEVELOPMENTSI Fig. 3. Structure of IMAGE deforestation module.

validate the different relationships for the past decades the simulated amounts of ecosystems have been compared with the amounts presented in the literature. Used sources were the FAO Production Yearbooks as well as a number of literature estimates of present and historical values of important parameters, among which those cited by the World Resources Reports (World Resources Institute, 1986, 1987, 1988/1989). The most important assumptions and input data are summarized in the appendix.

223

T R O P I C A L D E F O R E S T A T I O N A N D ITS C O N S E Q U E N C E S F O R G L O B A L C L I M A T E

TABLE 1 Denotation of the ecosystems (in ha) and processes (in ha/year) CLSF OPNF AGRIC GRASS OTHR

= = --=

closed tropical forest open tropical forest arable land pasture land tundra and semi-desert land

DEFAGC DEFAGO DFLIVC DFL1VO YSHFTC DEFYSH YDMOTH COMM REFOR FWOOD 1ND DEGAGR DEGGRA

= -= = = = = = -= = = =

expansion of arable land into closed forest expansion of arable land into open forest expansion of pasture into closed forest expansion of pasture into open forest shifting cultivation in closed forest shifting cultivation in open forest expansion of agricult, land into other land logging by commercial wood production development of plantations fuelwood gathering industrial or mining projects degradation of arable land degradation of pasture

from from from from from from from from from from from from from

CLSF to AGRIC OPNF to AGRIC CLSF to GRASS OPNF to GRASS CLSF to OPNF OPNF to AGRIC OTHR to AGRIC CSLF to OPNF OTHR to OPNF OPNF to OTHR CLSF to OTHR AGRIG to GRASS GRASS to OTHR

T h e a m o u n t o f c l o s e d t r o p i c a l f o r e s t is d e t e r m i n e d b y the initial a m o u n t a n d the c o n v e r s i o n i n t o a r a b l e l a n d , p a s t u r e , o t h e r l a n d a n d o p e n t r o p i c a l f o r e s t respectively. T h i s leads to the f o l l o w i n g d y n a m i c e x p r e s s i o n : C L S F ( t ) = C L S F ( t - 1) + I ' , _ i ( - C O M M ( r ) - DEFAGC(r) - DFLIVC(r)dr

- IND(r) - YSHFTC(r) (1)

S i m i l a r to the c l o s e d f o r e s t the a m o u n t o f o p e n t r o p i c a l f o r e s t is d e t e r m i n e d b y the initial a m o u n t a n d the c o n v e r s i o n i n t o a r a b l e land, p a s t u r e , o t h e r l a n d a n d the r e - e s t a b l i s h m e n t o f o p e n forest; a d d i t i o n a l l y c l o s e d f o r e s t s are c o n v e r t e d i n t o o p e n forests. H e n c e the d y n a m i c a m o u n t o f o p e n t r o p i c a l f o r e s t is e x p r e s s e d as: OPNF(t)

= O P N F ( t - 1) + I ' , _ ~ ( C O M M ( r ) + R E F O R ( r ) + Y S H F T C ( r ) - DEFAGO(r) - DFLIVO(r) - FWOOD(r) - DEFYSH(r)dr

(2) T h e last e c o s y s t e m w h i c h will be d e s c r i b e d is a r a b l e land. T h e a m o u n t o f a r a b l e l a n d is specified b y the initial a m o u n t a n d the e x p a n s i o n o f a g r i c u l t u r al l a n d i n t o c l o s e d f o r e s t a n d o p e n f o r e s t a n d i n t o o t h e r l a n d m i n u s the d e g r a d a t i o n o f a r a b l e l a n d to g r a s s l a n d .

224

J. R O T H M A N S A N D R.J S W A R T

AGRIC(t) = AGRIC(t - 1) + I',_ ~ (YDMOTH(r) - DEGAGR(r) + DEFAGC(r) + DEFAGO(r) + DEFYSH(r)dr

(3)

The other ecosystems, pasture and other land are treated identically with respect to the systems dynamics (Rotmans, 1990). The expansion of agricultural land is modelled in such a way that the most suitable area is converted to agriculture first. Therefore, in cooperation with the International Soil Reference and Information Center (ISRIC) in Wageningen the suitability of soils was estimated as a function of soil type, climate and slopes (Bouwman, 1989). The results are shown in Table 2, which contains suitability classes for the different ecosystems. At any time the expansion of agricultural land is distributed over the ecosystems closed and open tropical forest, and other land. This is dynamically represented in the equations by using suitability ratios, representing the proportion between the remaining amount of closed forest, open forest and other land within a particular suitability class. This distribution mechanism will be explained later on in section 4. Following this interpretation the expansion of agricultural land into closed tropical forest can be represented by the demand for arable land and the degradation of arable land, multiplied by the suitability ratio: DEFAGC(t) = (YDMAGR(t) + DEGAGR(t)) * ALFCLF(t)

(4)

where, YDMAGR(t) = yearly demand for arable land in (ha/year) and ALFCLF(t) = ratio of current amount of closed tropical forest to the current amount of closed forest, open forest and other land according to the suitability distribution in Table 2. A similar definition holds for the expansion of agricultural land into open tropical forest, with an identical type of suitability ratio ALFOPF(t). The expansion of agricultural land into other land is simulated by the following relationship: YDMOTH(t) = YDMAGR(t) * ALFOTH(t)

(5)

where, ALFOTH(t) = ratio of current amount of other land to the current amount of closed forest, open forest and other land according to the suitability distribution in Table 2. Contrary to the distribution of agricultural land, the distribution of pasture land is simply supposed to be proportional to the remaining area. Thus the agricultural expansion of pasture into closed forest is determined by the demand for grassland, the degradation of both grassland and arable land, multiplied by the relative amount of remaining closed forest: DFLIVC(t) = (YDMGRA(t) + DEGGRA(t) - DEGGRA(t)) • ALFGRC(t)

(6)

TROPICAL DEFORESTATION AND ITS CONSEQUENCES FOR GLOBAL CLIMATE

225

where, YDMGR(t) = yearly demand for pasture (in ha/year) ALFGRC(t) = ratio of current amount of closed tropical forest to the initial amount of closed tropical forest. For an exhaustive and more detailed description of the other driving forces and processes of the deforestation module we refer to Rotmans (1990). 4. DESCRIPTION OF THE M O D E L L E D PROCESSES

Permanent agriculture Expansion of permanent agriculture has always been the most important factor in the deforestation at all latitudes. Presently the demand for agricultural land in the tropical regions is primarily caused by the following factors: - an increasing number of people - - higher levels of consumption of food, fiber and forest products due to economic growth - area and productivity loss by land degradation - - production for export (often for debt relief) Many studies have indicated, that theoretically present and future world populations can be amply fed by increases in arable land and productivity (e.g. Buringh et al., 1975). Other studies indicate, that this is not so for individual countries (FAO, 1984). Unfortunately+ in past studies on population carrying capacities the present vegetation has not been taken into account. In this study it is. To determine the net demand for additional agricultural land for each region the following parameters have been taken into account: population, cereal demand per capita, cereal productivity per hectare, present arable land, fraction of land used for non-cereals and other products, fraction of cereals used for feed, the self-sufficiency ratio to allow for food imports or aid. This is modelled as follows: POP(t) * CE(t) * SSR DMAGR(t)

=

Y(t) * o~ * ( 1 - 1 3 - ~ , )

(7)

where, DMAGR(t) = cumulative demand for agricultural land, POP(t) = population, CE(t) = cereal consumption, Y(t) = yield in kg/ha, SSR = self sufficiency ratio, c~ = fraction of cereals used for food,/3 = fraction of area for food from non-cereals and 3, = fraction of area for other agricultural products. To determine the total expansion of agricultural land into other ecosystems finally degraded lands are added, see equation (4). For lack of

226

J. R O T H M A N S A N D R.J. S W A R T

quantitative information we simulated degradation by a negative exponential function, which encompasses the assumption, that every newly converted hectare will be more susceptible to erosion than the previous: D E G A G R (t) = AGRIC(t) - AGRIC(0)

(8)

PR • exp [-(Y(O)/Y(t)) * (AGRIC(t) - AGRIC(0)/AGRIC(0))] where, D E G A G R = degradation mand for agricultural land (in ha/year) and PR = agricultural degradation constant (= 100) (in years). The demand for agricultural land has to be distributed over the different ecosystems. The following procedure has been followed. First, those areas are converted, which have been altered by logging or shifting cultivation the year before. Then the land is assumed to be colonized according to suitability for agriculture. The underlying algorithm is depicted in Fig. 4. For this project, estimates of the suitability of areas have been made by Bouwman (1989), taking into account the soil types, the climate, the topography and the input level, see Table 2. Generally the most suitable area is assumed to be converted to agriculture first. In the right column of Table 2 the total amounts of different land types are presented for the year 1985, in millions of ha. These totals are divided into four suitability classes, which are arranged in order of suitability for agricultural use. This is done for two different cases: a high input and a low input for the various key factors as population growth, food production, consumption level, etc. The high input level is assumed to be equivalent to scenarios A and B, while the low input level is comparable with scenarios C and D. Shifting or pioneer cultivation

Much controversy exists on the subject of shifting cultivation. According to Lanly (1982) shifting cultivation is increasing because of increasing numbers of people; according to Myers (1984) it is decreasing, because landless farmers from other areas convert traditional shifting cultivation areas into permanent crop land. For the carbon cycle a sustainable traditional form of shifting cultivation is not very important. Without entering the discussion on the definition of 'shifting' or 'pioneer' cultivation we choose the following approach: expanding traditional shifting cultivation in closed forest areas is considered as a catalysing agent for further forest destruction in later years, degrading them initially into open forests. First the expansion into closed forests is calculated as a function of the fraction of the rural

227

~

n

Y

I YDI~ AGR

~

F I

I

'v t = t+l

n

~OPNF

1 + OOPNF - YDMAGF

CLSF - CLSF-COMM I YDMAGR = YDMAGR-YSHFTC-COMM

n = n+1

I n

n

YDMAGR-CLSF(n) -OPNF(n)-OTHR(n) CLSF(n) = 0 J OPNF(n) = 0

~1

YMAGR

=

rI

I OTHR(n) = 0

CLSF(n) = CLSF(n) - YDMAGR "

CLSF(n) = CLSF(n) - YDMAGR"

CLSF(n) = CLSF(n). YDMAGR "

CLSF(n) CLSF(n)+OPNF(n)+OTHR(n) OPNF(n) CLSF(n)+OPNF(n)+OTHR(n) OTHR(n) CLSF(n)+OPNF(n)+OTH R(n)

CLSF

= CLSF(I)+CLSF(II)+CLSF(III)+CLSF(IV)

OPNF

= OPNF(I)+OPNF(II)+OPNF(III)+OPNF(IV)

OTHR

= OTH R(I)+OTH R(II)+OTHR(III)+OTHR(IV)

I where: t= n. CLSF ,= OPNF OTHR ,,

time suitability class closed tropical forest open tropical forest other land

YDMAGR = yearly demand for agriculture YSHFTC = yearly shifting cultivation COMM = logging

Fig. 4. Scheme for the algorithm to allocate the demand for agricultural land. t, Time; n, suitability class; CLSF, closed tropical forest; OPNF, open tropical forest; OTHR, other land; YDMAGR, yearly demand for agriculture, YSHFTC, yearly shifting cultivation, COMM, logging.

228

J. R O T H M A N S A N D R.J. S W A R T

TABLE 2 Agricultural suitability classes (derived from Bouwmam 1989) mln. ha

Suitability classes ~ Low input

Africa Closed forest Open forest b Grassland Arable land Semi-desert ¢ Total Latin America Closed forest Open forest b Temperate forest Grassland Arable land Semi-desert c Total Southeast Asia Closed forest Open forest b Grassland Arable land Semi-desert c Total

High input

Total

I

I1

II1

1V

I

I1

1II

IV

0.7 1.0 6.6 1.1 2.8 12.1

27.8 136.9 186.0 50.5 82.7 483.8

151.0 165.8 170.0 64.4 93.7 644.7

41.7 194.8 415.7 67.1 1098.0 1817.3

43.3 162.5 221.8 67.7 111.7 606.9

149.3 227.5 233.0 76.1 119.4 805.3

5.8 30.0 35.1 12.4 21.9 105.1

22.7 221.1 78.6 498.5 288.3 778.2 27.0 183.1 1024.2 1277.1 1440.7 2958.0

0.0 0.0

53.1 46.3

509.4 87.3

130.7 403.3 179.8 106.0 86.91 106.2

66.0 9.6

0.0 5.3 0.0 0.6 5.9

22.0 97.8 51.2 57.0 327.3

0.0 202.1 85.8 18.3 902.8

14.7 245.2 38.5 215.8 750.8

22.7 0.0 173.1 201.8 108.3 38.3 58.4 30.5 852.7 556.7

7.8 0.0 0.0 0.8 1.7 10.2

62.2 15.6 4.3 97.5 53.8 233.3

179.6 3.5 9.9 55.0 58.2 306.2

55.9 11.9 16.0 103.9 71.5 259.1

220.6 25.0 12.5 166.6 119.8 544.4

29.6 1.5 3.5 20.5 15.1 70.2

11.1 57.2 16.7 29.3 190.0

44.0 36.8

693.2 239.5

2.9 36.7 118.3 550.4 12.2 175.5 173.3 291.6 387.5 1986.9

35.9 3.4 5.0 25.8 22.0 92.1

19.4 1.1 9.2 44.3 28.3 102.2

305.5 31.0 30.2 257.2 185.0 808.8

a I = high, 1I = moderate, III = low and IV = marginal suitability. b Including forest fallow, logged forest, forest plantations. c Including human area.

p o p u l a t i o n i n v o l v e d i n this t y p e o f a g r i c u l t u r e , t h e f a l l o w p e r i o d a n d t h e c o n v e r s i o n p e r f a m i l y ( D e t w i l e r et al., 1985): SHFTCL(t)

= POP(t) * RUR

* SWFC

* (CLF/NFM)

* FC

(9)

w h e r e , S H F T C L ( t ) = a r e a u n d e r s h i f t i n g c u l t i v a t i o n ( i n ha), R U R = r u r a l fraction, SWFC = rural fraction shifting cultivation (swidden farmers), C L F = a m o u n t o f l a n d c l e a r e d p e r f a m i l y p e r y e a r (in h a / y e a r ) , N F M =

TROPICAL DEFORESTATION AND ITS CONSEQUENCES FOR GLOBAL CLIMATE

229

number of people per family (in cap/fm) and FC -- fallow cycle (in years). In the computer program this converted area is memorized and in the next timestep it is first used for expansion of permanent agricultural area.

Cattle breeding In all developing countries the number of cattle has gradually increased along with the population. Only in Latin America has this growth taken place at the expense of tropical forests. In Asia and Africa meat and dairy form a less important part of the diet, while no major export markets are near. The troublesome situation in Latin America, where cattle breeding is expanding, originally triggered by beef demands in the U.S.A. and maintained by a complex of socio-economic factors, is discussed extensively in the literature (Hecht, 1988; Fearnside, 1981; Repetto, 1988). In Asia, cattle are integrated in the agricultural system and in general do not require separate grazing lands. They feed on agricultural wastes and off-season feed products. In Africa a different situation exists. Many areas are less suitable for agriculture than for cattle breeding. Therefore livestock is an important source of income for many people. Although the extensive cattle breeding has expanded in Africa, it appears, that the most important problem is the land degradation by overgrazing rather than forest destruction. In the model for the different scenarios different growth rates of numbers of cattle were taken and different levels of intensity (numbers per hectare): DMGRS(t) = GRASS(0) * CAT(t)/CAT(0) * CATPRD(t)/CATPRD(O) (10) where, DMGRS(t) -- cumulative demand for pasture (grassland) (in ha), GRASS(0) = initial amount of pasture, in 1900 (in ha), CAT(t) = number of cattle, CAT(0) -- initial number of cattle, in 1900, CATPRD(t) = cattle productivity (in ha/hd) and CATPRD(0) = initial cattle productivity, in 1900 (in ha/hd). In the scenarios different logistic curves for the number of cattle are taken, increasing from scenario D to scenario A. To compute the demand for pasture different intensities are taken for cattle productivity, also decreasing from scenario D to A; the numbers are given in the Appendix.

Logging In public discussions on tropical deforestation in the developed world production of commercial tropical hardwood is often pointed at as the most important cause of deforestation. The wood traders then reply, that they do not cause more than a few percent. Both may be right. Directly commercial

230

J. ROTHMANS AND R.J. SWART

logging may not contribute much in comparison with other causes. But indirectly logging activities open up the forests by road construction, taking poor farmers in their wake to finish the job. In our model we simulated this effect similar to shifting cultivation: after logging in the next timestep the logged area is converted into agricultural area. This assumption might be extreme and is not valid everywhere, for instance in parts of Indonesia, where logs are transported by river. Generally logging can be considered as the mining of a non-renewable resource. Commercially preferred species need 40-100 years for full regrowth, if they are replanted, which does not happen yet. The future does not look very promising. The per capita demand for wood and wood products is more or less proportional to economic growth (Van der Meiden, 1988). This means, that in the future an increasing portion of the wood production of the tropical forest areas will be used internally. Even now some tropical forest countries are wood importers. Supplies in the developed world will be restricted by the application of the principle of sustained yield and maybe by the effects of acidification. As the forest resources of Southeast Asia and West Africa become depleted, one will focus the attention to hitherto relatively unexploited regions like Amazonia and the Congo basin. There is a number of ways to reduce the stress on the wood resources. First more species have to be utilized. This does however not stop the 'mining' aspect of tropical wood production. More efficient end use or the use of alternative materials is needed. Also the development of tropical hardwood plantations has to be stimulated. In the model scenarios we varied two parameters: the growth rate of hardwood production and the productivity per hectare.

COMM(t) = COMM(85) •

IND(CWP) IND(PRI)

(ll)

where, COMM(t) = forest degradation by commercial wood production (in ha/year), COMM(85) = forest degradation in 1985 by commercial wood production, IND = index, IND = 1 in 1985, CWP = commercial wood production and PRI = production intensity. In the scenarios different logistic curves are chosen to reflect increasing consumption of tropical hardwood (CWP) from scenario D to A, fit to a limited series of past consumption rates and World bank projections to 2000. In the scenarios also increasing removals per ha are assumed from A to D to reflect increasing wood usage efficiency (PRI), presented in the Appendix. Although we included different rates of plantation as well, this was not linked directly to forest degradation by logging.

TROPICAL DEFORESTATION AND ITS CONSEQUENCES FOR GLOBAL CLIMATE

231

Fuelwood gathering Combustion of fuelwood only contributes to the greenhouse effect, when extraction is larger than regrowth. This is the case in many regions, where fuelwood scarcities occur (de Montalembert and Cl6ment, 1983). Usually these areas lie outside the major forest areas. Lanly (1982) considers fuelwood gathering in forests as a less important degradation factor rather than a cause of total forest destruction. Fuelwood is the primary source of energy in rural areas. Since in the rural areas scarcities arise and wood is transported over ever longer distances, commercial fuelwood extraction from forests may increase, leaving large areas of forest severely damaged. According to Myers (1984) the demand for fuelwood already causes major deforestation (25 000 km 2 annually). The available information is insufficient to describe the influence of firewood collection on tropical forests adequately. For our study we assume that closed forests are not affected because of natural regrowth. Only open forests are taken to be deforested at a rate which is chosen proportional to rural population growth in the area: the 1985 value of the amount of land affected by fuelwood is multiplied by the relative rural population growth. No allowance for decreasing withdrawal because of plantation wood or increased combustion efficiency has been taken into account yet.

Industrial projects In systems studies on deforestation industrial projects are seldom considered. The implementation of projects is primarily steered by political and financial aspects. Although the direct effect is usually rather small in terms of hectares of forests cleared, the opening up of virgin land and possible employment opportunities draw vast amounts of people to the project area. Although in all three regions these projects play a role, we only tried to include the massive plans for Amazonia, aiming at the exploitation of the mineral resources and hydropower potential. The planned size of the projects is very much dependent on government changes and opportunities for external financing and changes rapidly. In order to illustrate the potential importance of these projects, we included different rates of forest conversion for the different scenarios for Latin America.

Forest plantations Next to the abatement of deforestation re-afforestation is often brought forward as an important instrument to combat climate change. The calculated areas necessary to balance fossil fuel emissions depend very much

232

J. R O T H M A N S A N D R.J. S W A R T

on the assessment of the productivity of forest plantations. Optimists use figures taken from relatively small well managed plots~ while pessimists argue that the associated productivities cannot be applied to the large scale plantations necessary to have a significant impact on the carbon cycle. Furthermore in the developing countries forest plantations will be established for more pressing needs than carbon sequestering, such as fuelwood supply or erosion control. In our model we only included reforestation of 'other lands', leading to forest plantations with a carbon density comparable to open forests. Re-afforestation of abandoned agricultural lands in the developed countries may have significant potential but is not covered by our module. In our most optimistic scenario we started with present plantation rates as given by Lanly (1982) and let the areas logistically increase to 345 million hectares in Africa, 310 million hectares in Latin America and 100 million hectares in Southeast Asia. In the other scenarios only 75, 50 and 25%, respectively, of these areas would be forested. These figures are derived from EPA (1989). RESULTS

In Figs. 5 and 6 the remaining amounts of closed and open forests for four scenarios are presented for the three regions. In the most pessimistic scenario the closed tropical forests would disappear before the middle of the next century. Although one may expect the forests in the remote Amazon and Congo basin to last the longest, we find that they may disappear at least as fast as the Asian forests. This result can partly be explained by the fact that we used estimates of past land use changes by Houghton et al. (1983) to fit our results for the period 1900-1980. This source suggests that in Asia only a small percentage of the tropical forests would have been destroyed. Very likely this finding is not correct, taking into account that in several southeast Asian countries 30-40% of the forests have disappeared (World Resources Institute, 1988). In our model this affects also the future deforestation rates. The curves for open forests show upward trends for some scenarios, which is primarily caused by the inclusion of forest plantations, logged forests and closed forests affected by shifting cultivators in this category. The important role of soil degradation is illustrated in Fig. 7, which shows that the simulated loss of arable land increases for all regions in both absolute and relative terms. This is caused by the fact that degradation is modelled in such a way that in areas where land expansion is the fastest the area losses are the greatest. Figures 8 and 9 show the causes of deforestation as simulated in 1980 and one particular future year, namely 2020. For closed forests the demand for agricultural land, pasture and commercial wood are of similar importance,

AND ITS CONSEQUENCES

TROPICAL DEFORESTATION

FOR GLOBAL CLIMATE

233

(3

.c_ 2 -

~.~ . . . . . . . . scenario A ~

"1o 1 O "B

l'='ner! °cIscenari°D '

0-

1900

.

.

•'. .

.

.

---'1

,

I m

1950

"',

2000

\"

"'" ....

\-.,

.... :

2050 timeinyears

2100

10

=o

m 7.5--

C) -~.

E <£::

5

,9o 2.5

_

scenario B scenario C

---I ..... m

~ " ", "''~ ~.~,

"'..

0

1900

1950

2000

2050

2100

time in years 4A O

~3-


~

0

scenarioB _ _ _ m scenarioC . . . . . m

~10

~

scenarioD _ _ _ . I 0

I

1900

1950

"',

~•• •

"-..

N~,,, I

2000

I

2050

2100

timeinyears

Fig. 5. Simulated areas of closed tropical forests for three regions. while for open forests both agriculture and fuel wood demand play a dominant role. In Fig. 10 the impact of different deforestation scenarios on the atmospheric CO2-concentration is shown. For both lines we took a common

234

J ROTHMANS AND R.J. SWART

10-

I

=o

~7.5.~ 5.0"~ 2.5 ~.

o

II,--°a,io'~ !

I Iscenario B I lscenarioC Ilmr'oO

-.---.

o-I . . . .

I

1900

1950

!

I

2000

8-

I

I

I

I

I

I

2050 time in years

21 O0

scenario A / scenario B - - - II

eiO

s~na,io C

6_

I

.....

t~


4-

/ ,J

,o°°''''''°*-..

! 0_

1900

1950

2000

2050 time in years

21 O0

1-~ ":

] |

Iscenario A II Iscenario B - - - II

," - /

""

0.75-r~ "~" I ~

IscenerioC . . . . . II ; /'''',, Iscenari°D --''1/, "'

-:. o.5o-I

\

i

c,,25

~.""

I

/';'/"

I

\j7

1900

1950

-I

V

\

',

"',\', ~

" , t ~

~ '

2000

2050 time in years

",'

',\

21 O0

Fig. 6. Simulated areas of open tropical forest for three regions.

energy scenario (a forced trend scenario, see Rotmans et al., 1990). The difference between deforestation scenario A and D is limited to about 50 ppm, or roughly 10%. For other energy scenarios the differences are of the same order of magnitude. For the higher energy scenarios the relative differences are slightly smaller. These findings indicate that deforestation is important,

T R O P I C A L D E F O R E S T A T I O N A N D ITS C O N S E Q U E N C E S F O R G L O B A L C L I M A T E

12_

_i

235

i

Isc..,.,io I''°ha''°"B _ _ _ ilI

9-Iscenario C . . . . . II

/

Iscen.,o 0 -- -- |

+~(13

"...

,s

- ~ E

i

i

1900

,

-

1

,

|

i

i

1950

,

i

,

2000

i

. . . . . ._Z_-2-

oO

|

|

,

/ /

i

,

i

i

2050 time in years

2100

12E

9_

scenarioA scenarioB - - - I

/

:-:J

3O A

§~,

Jl

6-

E

--- 8 3 ~E O_ 1900

,

-o t3

1950

2000

2050 time in years

2000

2050

8_ 6-

scenarioA scenarioB scenarioC

2100

j ""~-i

scenario D

.to1

oo

,

4_

o 2ca<:

~.~_

O_ i i i | i i , ~ 19O0

1950

ii~i" t time in years

2100

Fig. 7. Simulated degradation of agricultural land for three regions.

but small in comparison to the contribution of fossil fuel combustion to the greenhouse effect. It is important to note that this result is not the separate and direct effect of the destruction of tropical forests alone, but the net impact of all changes of land use and atmospheric composition. In the

236

J. R O T H M A N S A N D R.J. S W A R T

24.000.

•

industrialdevelopments

~ livestock

"20.000, ,=..-

B permanentagriculture

~ 16.000.

B shiftingcultivation B logging

~ 12.000. ~5 ~

8.O0O ~

~ 4.000,

1980 2020A2020B2020C2020D scenarios

Fig. 8. Conversion of closed tropical forests by cause for selected years.

25.000 D livestock 20.000 -c ~'

10.000

5.000 o =

0

~" 0

(5.000)

._o

8

~ permanentagriculture

15.000 B fuelwood B forest plantations B shifting cultivation B logging

(lO.OOO) (15.ooo) (20.000) (25.000) 1980 2020A 2020B 2020C 2020D scenarios

Fig. 9. Conversion of open tropical forests by cause for selected years.

TROPICAL DEFORESTATION AND ITS CONSEQUENCES FOR GLOBAL CLIMATE

1200_

237

a

scenarioA / scenario D -"'1 1000_

J ~

800_

400_ i

1900

!

I

l

l

!

1950

I

I

I

I

!

2000

s

i

I

!

l

J

.

=

2050

2100

time in years

500 -

[scenarioA / 450_ Iscenari°D - - - - |

b

.~ ~0_

I

ff NO_ 1900

1950

2000

2050

2100

time in years

Fig. 10. (a) Simulated atmospheric CO~concentrations for the highest and lowest deforestation scenario combined with an unrestricted trends energy scenario. (b) Simulated atmospheric CO 2 concentrations for the highest and lowest deforestation scenario combined with a forced trends energy scenario.

dynamics of the carbon cycle by Goudriaan and Ketner (1984) the CO2-release by deforestation is counteracted by carbon sinks like the COz-fertilization effect and charcoal formation after burning of vegetation. Furthermore it is important to realize that, in addition to tropical forests, all other ecosystems can also sequester considerable amounts of carbon. These factors account largely for the limited impact of deforestation in the carbon cycle module used here.

238

J. R O T H M A N S A N D R.J. S W A R T

A B scenario C scenario

scenario tO ',1=

,=

scenario

D

o

o '~"~'~'l

"ID C .o

._=

\ \

E i

O o

0

I

'

'

1900

I

,

I

1950

'

'

'

'

I

'

'

2000

'

'

I

2050

I

I

I

I

2100

time in years Fig. 11. The emission of CO_, by deforestation.

Therefore in Fig. 11 the 'gross' contribution of deforestation in terms of carbon dioxide emission is presented. In scenario A the emission of CO_, by deforestation rapidly increases, but after having cleared closed tropical forests, goes to zero. In the most optimistic case a direct but gradually decreasing emission takes place. CONCLUSIONS

(1) Simulation results with IMAGE show that if present deforestation rates continue a total destruction of the tropical forests will occur halfway through the next century. This would be disastrous although its effect on global climate is only one reason to stop tropical deforestation. Other possibly more important reasons are the inefficient use of the resource leading to erosion, the loss of species, the threat to the indigenous population and the effects on local and regional climate. (2) The contribution of biospheric changes to the greenhouse effect is important but small as compared to that of fossil fuel combustion. Model

TROPICAL DEFORESTATION AND ITS CONSEQUENCES FOR GLOBAL. CLIMATE

239

simulations showed a difference of maximum 10% in CO2-concentrations over the next century between optimistic and pessimistic deforestation cases. The result that even rapid deforestation has only a moderate net effect on the carbon cycle is primarily caused in IMAGE by an increasing soil carbon pool due to the conversion of organic matter into charcoal during burning, by CO_, fertilization and to possible underestimation of soil losses after conversion. To some extent the biospheric CO~ emissions by deforestation will be counteracted by a CO~ fertilization effect. However, the less vegetation that remains, the less potential there is for this effect. (3) Simulations with IMAGE show that in all scenarios the most important direct cause of deforestation is the demand for agricultural land to satisfy demand for food, feed or debt-resolving export products. The combination between growth of population and their consumption level will increase the pressure on the forests when no drastic corrective measures are taken. (4) Although reforestation can play a significant role in a transition period, a gradual shift towards a world energy system based on renewables away from fossil fuels is a more effective strategy for the mitigation of climate change. (5) The IMAGE deforestation module shows that the development of a global land use change model is feasible and can provide valuable insights into interactions between human activities and the earth system. Such a model should, in addition to deforestation, also take into account the impact of desertification, soil alteration by intensive agriculture and wetland drainage. (6) Large scale industrial, infrastructural, mining or livestock projects have an enormous potential to accelerate the deforestation process, which is, however, difficult to quantify. Integrated economic and ecological evaluation is crucial to avoid unnecessary ecological disasters. (7) Modeling socio-economic causes for deforestation is hampered by the impossibility of linking resource supply and demand by a (price) mechanism, taking resource scarcity into account. In our model we simulated resource limitations by not using exponential but logistic functions for the most important processes. (8) Decrease of agricultural area and productivity by land degradation is an important factor, catalysing deforestation. Erosion control and conversion of soils less susceptible to erosion should have priority. The importance of this factor warrants a more thorough analysis of this process than applied in this study. Further work should not only address the loss of land and fertility in relation to agricultural production but also the carbon 'lost' from the carbon cycle through river and wind transport. (9) Not only international, but also internal demand for wood and wood products will continue to increase the pressure on the forests, shifting the

240

J. R O T H M A N S A N D R.J S W A R T

m a j o r p r o d u c t i o n a r e a s f r o m S o u t h e a s t Asia to A m a z o n i a . Successful plant a t i o n s c a n c h a n g e the p r e s e n t ' m i n i n g ' o f the forests into ' s u s t a i n a b l e use'. C e r t a i n l y with respect to the g r e e n h o u s e effect, w o o d is an excellent source o f energy as l o n g as it is p r o d u c e d in a s u s t a i n a b l e way. T h e efficient p r o d u c tion a n d utilization o f w o o d f r o m trees in p l a n t a t i o n s or in the rural e n v i r o n m e n t as a r e p l a c e m e n t o f ( r a t h e r t h a n in a d d i t i o n to) fossil fuels should be s t i m u l a t e d a l o n g with efforts to reduce o t h e r e n v i r o n m e n t a l p r o b l e m s a s s o c i a t e d with w o o d b u r n i n g . ACKNOWLEDGEM ENTS T h e a u t h o r s wish to t h a n k Lex B o u w m a n o f the I n t e r n a t i o n a l Soil R e f e r e n c e a n d I n f o r m a t i o n C e n t e r in W a g e n i n g e n for his w o r k on the suitability o f the t r o p i c a l a r e a s for a g r i c u l t u r e a c c o r d i n g to soils, t o p o g r a p h y , climate a n d i n p u t levels. REFERENCES Bouwman, A.F., 1989. Land Evaluation for Dry Farming, working paper, International Soil Reference and Information Centre, Wageningen. Buringh, P., van Heemst, H.D.J. and Staring, G.J., 1975. Computation of the Maximum Food Production of the World, Agricultural University, Wageningen. Detwiler, R.P., Hall, C.A.S. and Bogdanoff, P., Land use change and carbon exchange in the Tropics: II. Estimates for the entire region, Environ. Manage., 9 (6): 335-344. den Elzen, M.G.J., Rotmans, J., 1991. A scenario study on the socio-economic impact of the greenhouse effect for the Netherlands, Climatic Change, in press. EPA, 1989. Options for Stabilizing Global Climate, Draft Report to Congress, Washington. Fearnside, P.M., 1987. Causes of deforestation in the Brazilian Amazon. In: R.E. Dickenson (Editor), The Geophysiology of Amazonia, Wiley. Food and Agricultural Organization of the United Nations, 1984. Land, Food and People, Rome. Food and Agricultural Organization of the United Nations, 1986. FAO's Tropical Forest Action Plan, Unasylva 152: 38. Goudriaan, J. and Ketner, P., 1986. A simulation study for the global carbon cycle, including man's impact on the biosphere, Climatic Change, 6: 167-192. Hecht, S., 1988. Livestock Expansion in the Brazilian Tropics: Dynamics and Consequences, paper presented at the 46th International Congress of Americanists, Amsterdam. Houghton, R.A., Hobbie, J.E., Melillo, J.M., Moore, B., Peterson, B.J., Shaver, G.R. and Woodwell, G.M., 1983. Changes in the carbon content of terrestrial biota and soils between 1860 and 1980: a net release of CO, to the atmosphere, Ecol. Monogr., 53 (3): 235-262. IPCC, 1990. Emissions Scenarios of the Response Strategies Working Group of the lntergovernmental Panel on Climate Change, US Netherlands Expert Group on Emission Scenarios, draft report. Langeweg, F. (Editor), 1988. Concern for Tomorrow: A National Environmental Survey 1985-2010, Bilthoven.

TROPICAL DEFORESTATION AND ITS CONSEQUENCES FOR GLOBAL CLIMATE

241

Lanly, J.P., 1982. Tropical Forest Resources, FAO, Rome. Meiden, H.A., van der, 1988. Per capita wood consumption (ll) Bos en Hout Berichten, 4 (in Dutch). Montalembert, M.R. de and C16ment, J., 1983. Fuelwood Supplies in the Developing Countries, FAO, Rome. Myers, N., 1984. Conversion of Tropical Moist Forests, Norton. Repetto, R., 1988. The Forest for the Trees? Government Policies and the Misuse of Forest Resources, World Resources Institute, Washington. Rotmans, J., de Boois, H. and Swart, R.J., 1990. An Integrated Model for the Assessment of the Greenhouse Effect: the Dutch Approach, Climatic Change, 16 (3): 331-355. Rotmans, J., 1990. IMAGE: an Integrated Model to Assess the Greenhouse Effect, Kluwer Academic Publishers, The Netherlands. Swart, R.J. and Rotmans, J., 1990. Food or Forest? Can the tropical forests survive along with continuing growth of population and economy? In: Soils and the Greenhouse Effect, Wageningen. Swart, R.J. and Rotmans, J., 1989. A scenario study on causes of tropical deforestation and effects on the global carbon cycle, RIVM-report 758471007, Bilthovcn. World Resources Institute, World Resources 1986, World Resources 1987 and World Resources 1988/1989, World Resources Institute, Washington, 1986, 1987 and 1988.

Remaining closed forest Remaining open forest Fallow total A F closed Swidden cycle closed Fallow total A F open Fallow period open Population function a Percentage rural Swidden farmers closed Swidden farmers open Number in family Cleared per family Human area (0.05 ha/cap) Yield total cereals sc. A n Yield total cereals sc. B a Yield total cereals sc. C a Yield total cereals sc. D a Consumption food cereals

Africa

1000 ha 1000 ha 1000 ha years 1000 ha years mln. % % % cap/fm ha/year 1000 ha kg/ha kg/ha kg/ha kg/ha kg/cap

unit

110

70 2.5 10 8 1.3

10

294 000 539 000

1900

In 2100 scenario to

221 079 498 479 61 646 10 104 335 !5 555 70 4 10 7 1.3 27 750 800 800 800 800 110

1985

APPENDIX: IMAGE DEFORESTATION MODULE DATA SHEET

15

B

10

C

21 525 0.0073 0.0410 80 70 60 3.5 2.5 2.0 10 8 6 6 7 8 1.3 1.1 0.9 proportional to population growth 306 0.087 0.018 297 0.119 0.020 241 0.161 0.029 153 0.135 0.047 170 150 120

to be calculated to be calculated (+ 1.4%/a) 20

A

D

3500 2500 1500 1150 110

2935 50 1.0 4 9 0.7

4~ to

A B C D

ha/year

% % % 1000 h a 1000 h a 1000 h a 1000 ha 1000 1000 1000 1000 ha/hd ha/year ha/year ha/year ha/year 50 50 50 50

60 60 60 60

100 61 I00 1 1 1 1 000 000 000 000 8.0 400 400 400 400 0.09

84 61 79 2411 2471 2411 2411 140 000 140 000 140 000 140 000 4.5 639 000 639 000 639 000 639 000 1 700 000

65 75 85 80 70 60 90 80 70 15.1 !.8E-4 0.06 2.12 1.6E-5 0.08 0.50 1.9E-6 0.10 0.09 2.7E-7 0.12 59 100 0.118 0.014 55 000 0.157 0.017 58 130 0.233 0.020 55 700 0.279 0.025 4 3.5 3.0 0.2001 0.0801 0.0265 0.1572 0.0786 0.0318 0.0898 0.0561 0.0424 0.0408 0.0292 0.0556 1.0 1.3 1.6 prop. to rural p o p u l a t i o n g r o w t h 500 350 250 200

000 000 000 000 2.5 2.50 2.00 1.60 1.40 2.0

95 50 60

A,B,C in formula pop = A/I B + exp[-Ct] ], D: s a t u r a t i o n level; base year 1900. b Self-sufficiency ratio = c o n s u m p t i o n / p r o d u c t i o n (when < 100: i m p o r t / f o o d aid). c Agricultural expansion into non-forest areas. a Wood p r o d u c t i o n index is 1 in 1985. FAO: in 1983 forest + w o o d l a n d s 719.6 mln. ha., pasture 778.2 mln. ha., agricultural land 183.1 mln. ha. a n d other lands 1277.1 mln. ha.

Percentage food (vs. feed) Area non-cereals Self-sufficiency ratio b Forest p l a n t a t i o n s sc. A ~ Forest p l a n t a t i o n s sc. B ~ Forest p l a n t a t i o n s sc. C a Forest p l a n t a t i o n s sc. D a Cattle n u m b e r scen. A a Cattle n u m b e r scen. B ~ Cattle n u m b e r scen. C a Cattle n u m b e r scen. D ~ Cattle productivity C o m m . wood prod. index ~'d C o m m . wood prod. index ~'J Comm. wood prod. index ~'J C o m m . wood prod. index ~'J W o o d prod. int. index Affected by fuelwood

©

© z

Z > Z ~7

>

e% > r"

7=

Remaining closed forest Remaining open forest Fallow total AM closed Swidden cycle closed Fallow total AM open Swidden cycle open Population function a Percentage rural Swidden farmers closed Swidden farmers open Number in family Cleared per family Human area (0.05 ha/cap) Yield total cereals sc. A ~ Yield total cereals sc. B a Yield total cereals sc. C a Yield total cereals sc. D ~ Consumption food cereals Percentage food (vs. feed)

Latin America

1000 ha 1000 ha 1000 ha years 1000 ha years mln °/o % % cap/fm ha/year 1000 ha kg/ha kg/ha kg/ha kg/ha kg/cap °/o

unit

140 95

40 9 8 7 1.0

15

10

873 000 307 000

1900

In 2100 scenario to

IMAGE D E F O R E S T A T I O N M O D U L E DATA SHEET

693 155 239 520 108 612 10 61 650 15 406 4O 9 8 7 1.0 20 300 1800 1800 1800 1800 140 60

1985 B

15

C

10

31 836 0.025 0.0346 60 55 50 8.0 6.0 4.0 8 6 4 7 7 7 1.3 1.I 1.0 proportional to population growth 628 0.150 0.019 546 0.162 0.023 394 0.141 0.030 355 0.158 0.038 200 170 150 50 60 70

to be calculated to be calculated (+ 1.1 "/,,/a) 200

A

D

4200 3400 2800 2250 130 80

1259 45 3.0 2 7 0.8

I,,J

ha/year

% 1000 ha % 1000 ha 1000 ha 1000 ha 1000 ha 1000 1000 1000 1000 ha/hd 1000 ha ha/year ha/year ha/year ha/year

48 48 48 48

100 100 100 100

50

90 7303 7303 7303 7303 250 000 250 000 250 000 250 000 2.4 1000 400 2003000 400 2003000 000 2003000 400 2003000 0.03 1 300 000

125 1.5 1.5 1.5 1.5 000 000 000 000 4.4

50

65

55

115 105 90 49.2 6.3E-4 0.06 8.54 5.5E-5 0.08 1.53 6.6E-6 0.10 0.28 9.0E-7 0.12 52 000 0.041 0.021 53 600 0.071 0.023 52 060 0.116 0.028 58 300 0.167 0.032 3.5 3.25 3.0 750 500 250 0.0770 0.0193 0.0356 0.0607 0.0202 0.0401 0.0278 0.0139 0.0535 0.0049 0.0033 0.0802 1.0 1.3 1.6 prop. to rural population growth

75 85 79 000 155 000 235 000 310 000 1270000 755 000 450 000 350 000 2.5 0* 4.0 3.0 2.0 1.5 2.0

45

a A,B,C in formula pop = A / I B + exp[-Ct]}, D: saturation level; base year 1900. b Self-sufficiency ratio = consumption/production (when < 100: import/food aid). c Agricultural expansion into non-forest areas. a Wood production index is 1 in 1985. FAO: in 1983 forest + woodlands 932.7 rain. ha, agricultural land 175.5 mln. ha, pasture 550.4 rain. ha and other lands 291.6 rain. ha.

Area non-cereals Irrigated area Self-sufficiency ratio b Forest plantations sc. A a Forest plantations sc. B a Forest plantations sc. C a Forest plantations sc. D a Cattle number scen. Aa Cattle number seen. B a Cattle number scen. C ~ Cattle number seen. D a Cattle productivity Other deforestation Comm. wood prod. index ~'d A Comm. wood prod. index ~'d B Comm. wood prod. index ~'J C Comm. wood prod. index a'd D Wood prod. int. index Affected by fuelwood

t~

r-

r~ O Z

Z

Z

5

).

r-

O

Yield total cereals sc. D b

Yield total cereals sc. C b

Remaining closed forest Remaining open forest Closed 9 countries Open 9 countries Fallow total AS closed Swidden cycle closed Fallow total AS open Swidden cycle open Population function 2 Percentage rural % swidden farmers closed % swidden farmers open Number in family Cleared per family Human area (0.05 ha/cap) Yield total cereals sc. A b Yield total cereals sc. B b

Tropical Asia

cap/fm ha/year 1000 ha kg/ha kg/ha kg/ha kg/ha

1000 ha 1000 ha % % I000 ha years 1000 ha years mln. % % %

unit

50 2 0.4 9 0.7

9

10

381 000 85 000

1900

13 countries a In 2100 scenario to

IMAGE D E F O R E S T A T I O N M O D U L E DATA SHEET

301 344 30 653 92 98 69 225 10 3990 9 1274 50 2 0.4 9 0.7 63 700 2500 2500 2500 2500

1985

15

B

10

C

0.0433 0.0320 65 55 3.0 2.0 0.3 0.2 7 8 1.0 0.8 0.7 proportional to population growth 563 0.095 0.024 348 0.088 0.035 279 0.081 0.041 226 0.075 0.049

138 988 70 4.0 0.4 7

to be calculated to be calculated to be calculated to be calculated (+ 1.25°/da) 20

A

0.6

8

0.1

D

6000 4000 3500 3000

3206 50 1.0

7~ -q

k..

> Z ~7 7~

z

-r ~r >

©

I-O 4~

ha/year

1000 ha 1000 ha 1000 ha 1000 ha 1000 1000 1000 1000 ha/hd ha/year ha/year ha/year ha/year

%

1000 ha

%

kg/cap %

170 80 50 62 000 100 95 2 7320 2 7320 2 7320 2 7320 135 000 303 000 135 000 303 000 135 000 303 000 135 000 303 000 0.107 0.10 68 160 1 755 000 68 160 1 755 000 68 160 1 755 000 68 160 1 755 000 0.04 1 1 700 000

170 95 50

230 210 190 65 75 85 70 60 50 150 000 125 000 100 000 110 100 90 62.9 2.5E-3 0.06 9.53 1.9E-4 0.08 1.65 2.2E-5 0.10 0.29 2.9E-6 0.12 150 000 0.040 8.6E-3 148 360 0.099 0.011 109 970 0.220 0.023 4462 0.013 0.075 0.30 0.25 0.15 0.1846 0.1055 0.0317 0.1150 0.0767 0.0408 0.0390 0.03000.0589 1.0E-4 1.0E-40.1495 1.0 1.3 1.6 prop. to rural population growth

170 95 40 85 000 80 25 160 50 200 75 000 100 000 3 750 000 1 500 000 500 000 350 000 0. I 0 1.75 1.50 1.30 1.00 2.0

Indonesia, Papua New Guinea, Burma, Malaysia, Philippines, Thailand, Vietnam, Kampuchea, Laos, Bhutan, Nepal, India, Bangladesh. b A,B,C in formula pop = A / I B + exp[-(C/*/t)], D: saturation level; base year 1900. c Self-sufficiency ratio-consumption/production (when < 100: import/food aid). d Agricultural expansion into non-forest areas. Wood production index is 1 in 1985. FAO: in 1980 forest +woodlands 336.5 mln. ha., pasture 30.2 mln. ha., agricultural land 255.2 mln. ha. and other lands 185.0 mln. ha.

Consumption food cereals Percentage food (vs. feed) Area non-cereals Irrigated area Self-sufficiency ratio ~ Forest plantations sc. A b Forest plantations sc. B b Forest plantations sc. C b Forest plantations sc. D b Cattle number scen. A b Cattle number scen. B b Cattle number scen. C b Cattle number scen. D b Cattle productivity Comm. wood prod. index b'~ A Comm. wood prod. index b'~ B Comm. wood prod. index b'~ C Comm. wood prod. index b'~ D Wood prod. int. index Affected by fuelwood

> -4

r-

?z

7~

r~

c

©

?.

z > z

>

7~

in

> t-