Recent trends and current status of forest resources

World forest areas in 1995
Changes in forest cover and condition

World forest areas in 1995

Since its inception, FAO, in cooperation with other international partners, particularly the UN-Economic Commission for Europe (UN-ECE) and the UN Environment Programme (UNEP), has endeavoured to provide an overall picture of the state of the world's forest resources at five to ten year intervals. The information is largely based on an analysis of existing country data (see Box 1). The last such review, 'Forest Resources Assessment 1990' (FRA 1990) was carried out with 1990 as the reference year. The next full assessment will be FRA 2000. This edition of the State of the World's Forests provides information on forest cover as of 1995, and a revised estimate of forest cover in 1990.¹ More detailed information on data collection and analysis is provided in Annex 4 of this volume.

¹ A revised figure of forest cover in 1990 is based on updated population figures for 1990 and some new inventory reports which were available in 1996. Based on this Information, the new estimates of forest areas in 1990 are somewhat higher than those given by FRA 1990, and those of deforestation rates during the 1980s are slightly lower.

Forests are estimated to have covered 3 454 million hectares (ha), or 26.6 percent of the total land area of the world (Greenland and Antarctic excepted) in 1995. Figure 1 shows the distribution of forests (natural forests plus plantations) in 1995 by region (see Table 1 of Annex 3 for a list of countries by region). Developing countries, which for the most part are tropical, account for 58.9 percent of the total land area of the world and 56.8 percent of its forests.

Estimates of forest areas in 1995 by country are shown in Data Table 2 of Annex 3. More than 60 percent of the world's forests are located in seven countries: the Russian Federation, Brazil, Canada, United States, China, Indonesia and Zaire (see Figure 2). Twenty-nine countries have more than half of their land covered by forest. Of these, 21 are in the tropical belt. Forty-nine countries, in addition to the many non-forested small island states and territories, have less than 10 percent of their land covered by forests. Five entire subregions are in this category: North Africa (1.2 percent), Near East (1.9 percent), Temperate Oceania (6.2 percent), Non-tropical Southern Africa (6.8 percent) and West Sahelian Africa (7.5 percent).

Figure 1 - Forest areas by main regions in 1995

The distribution of forest and of forest cover change into ecological zones was not assessed for this report. The latest information of this kind was provided by FRA 1990.²

² FAO. 1993. Forest Resources Assessment 1990 - Tropical Countries. FAO Forestry Paper 112, FAO, Rome.

At that time, temperate and boreal forests occupied 1.64 billion ha and tropical forests occupied 1.76 billion ha. A breakdown by ecological zone was available only for tropical forests (see Table 1). These data show that the large majority (88 percent) of tropical forests are in lowlands. Of these, tropical rainforests are the most abundant (accounting for 47 percent of all lowland tropical forests), followed by moist deciduous forest and finally by dry forests.

Natural forests

The estimates of forest area given above include undisturbed forests, forests modified by humans through use and management (or 'semi-natural' forests) and human-made forests (i.e., forest plantations) created artificially by afforestation or reforestation.³ In most developed countries, forests are often regenerated by a combination of natural and artificial means on the same management unit, making it difficult to isolate figures for natural forest and plantations. The distinction can more easily be made for developing countries.

³ See definition of these terms in Annex 2.

Box 1 Two essential components of global assessments of forest resources: national forest inventory systems and a common classification system for forests National forest inventory systems The quantity and quality of information provided by global and regional assessments are dependent to a large extent on the capacity of national forest inventory systems to collect and analyse data at national and sub-national levels, and to adjust the information so that it is compatible with global and regional reporting parameters. Of the 143 developing countries covered by FRA 1990, all but seven carried out one reliable nationwide estimate of forest cover at some time between 1970 and 1990. However, only 25 had more than one national forest cover assessment, and very few of these had carried out more than one comprehensive national forest inventory. The database for developed countries is far more complete; practically all developed countries were able to provide detailed information not only on areas, but also on biomass, volume and other forest parameters. UNCED recognized that assessment and systematic observations were 'often neglected aspects of forest resources management, conservation and development...' and that 'in many developing countries there was a lack of structures and mechanisms to carry out these functions'. Consequently, it devoted one of the four programme areas of the forest chapter of Agenda 21 to the 'strengthening and establishment of systems for the assessment and systematic observation of forests and forest lands' and the 'provision of sound and adequate updated information on forest and forest land resources to economists, planners, decision-makers and local communities'. FAO, with the support of several donor countries, is pursuing an active programme of country capacity building in forest assessment which, ultimately, will contribute to improving the quality of global forest resources assessments. A common classification system for forests In order to make a global assessment of forest resources, data from national forest inventories must be amalgamated, adjusting them to a common classification system. A common set of concepts and classifications for all wooded lands of the world is, therefore, essential for securing consistency between regional and global forest resources assessments. The FAO and UN-ECE Secretariats have worked over the years to build consensus on a minimum core set of definitions and classifications within and between developed countries (all of which lie in the temperate and boreal zones) and developing countries (most of which fall within the tropical belt). In June 1996, FAO, in cooperation with UN-ECE and UNEP, and with the support of the Government of Finland, organized an expert meeting in the Finnish town of Kotka in which agreement was reached on a common core set of parameters to be assessed within the framework of the next global forest resources assessment for the year 2000 and on some of the concepts and classifications to be used to that end.

Figure 2 - Countries with the largest percentage of the world's forests

Interest in natural forests, particularly in their role in biological diversity conservation, has led to efforts to compare forests today with what is thought to be their original character and to provide protection for areas of forests which have had little or no human interference. Although there are difficulties in identifying the extent of natural forest, compounded by problems of definition (see Box 2), some information exists which can be used as an indication of broad patterns of natural forests in various regions.

An attempt has been made by the World Wide Fund for Nature (WWF) to quantify the amount of forests in western Europe⁴ which have been relatively undisturbed by man or which have retained much of their natural character.⁵ It found that only a small proportion (less than 1 percent) of the total forest land in northern and western Europe can be considered as virgin forest. Almost all of it is located in Sweden, Finland and Norway, with small areas in Greece, Austria and Switzerland and (according to another author) in France. In eastern Europe, Slovakia and

⁴ World Wide Fund for Nature. 1994. The status of old growth and semi-natural forests in Western Europe. WWF, Gland, Switzerland.

⁵ This report distinguishes between 'virgin forest', defined as 'forest ecosystems whose characteristics are determined exclusively by natural location and environmental factors... without human influences present or visible any more', and 'natural and ancient semi-natural forests', which 'have not been planted or sown by man for the past two centuries' and 'which continue to have a large number of the natural elements'.

Table 1
Forest cover area and rate of deforestation in the tropics by main ecological zone as estimated by FRA 1990

ecological zone		land area	population density (1990)	annual population growth (1981 -90)a	forest cover (1990)		annual deforestation (1981-90)a
		million ha	inhabitants/km2	percent p.a.	million ha	percentage land area	(million ha)	(percent p.a.)
FOREST ZONE		4 189.7	57	2.4	1 748.2	42	15.3	0.8
lowland formations		3 476.6	57	2.3	1 543.9	44	12.8	0.8
	rainforest	937.1	41	2.2	718.3	76	4.6	0.6
	moist deciduous	1 298.6	55	2.4	587.3	46	6.1	1.0
	dry and very dry	1 241.0	70	2.3	238.3	19	2.2	0.9
upland formations (hilt and mountain forest)		713.1	56	2.6	204.3	29	2.5	1.1
NON-FOREST ZONE (alpine areas, deserts)		588.6	15	3.1	8.1	1	0.1	1.0
TOTAL TROPICSb		4 778.3	52	2.4	1 756.3	37	15.4	0.8

^a Compound formulae were used for population growth and deforestation rate calculations.
^b Totals may not tally due to rounding.

Box 2 'Naturalness' of forests A concept of 'naturalness' is being used as an indicator of biological diversity in some international processes which are defining criteria and indicators for sustainable forest management (see Part 3). Naturalness is characterized by such elements as a complex spatial structure, a composition and distribution of species natural to the site, a wide range of ages in tree species and the presence of dead and decaying trees. This has also led to the use of often loosely-defined terms such as 'virgin forests', 'primary forests', 'old growth forests', 'natural forests' and (ancient) 'semi-natural forests', making a consistent review of the state of naturalness difficult.

Belarus have considerable areas of virgin forest, and Poland and Croatia have small areas. The estimate given is that there are no more than a half a million hectares of virgin forest in Europe. An additional 21 percent of forest land in northern and western Europe is considered to be in 'natural and ancient semi-natural forests'. There are an additional 3 million ha in the region in national parks and other protected areas, and another 50 000 ha in small forest reserves (mostly for nature conservation and scientific research) whose use is tightly restricted. In all, therefore, less than 6 million ha or 5 percent of the forest land of Europe is considered to be relatively undisturbed (i.e., retaining much of its natural character).

The situation in temperate and boreal North America is quite different from that of densely populated Europe (and Japan), where use and management of forests for many centuries have left very little forest area untouched. 'Old growth forests', as they are called in North America, still cover extensive areas. 'On the lands managed by the US National Park Service alone, old-growth forests covered 1.97 million ha in 1988. Although not a direct measure of the extent of old growth forest, it is useful to note that the area of forest included in national parks and other protected areas in the United States and Canada together was reported to be close to 49 million ha in 1990.

FRA 1990 did not distinguish between undisturbed and disturbed natural forests in developing countries. FRA 1980, however, made estimates of the areas of undisturbed closed forests (primary forests and old secondary forests where there had been no logging for the last 60 to 80 years) and of closed forests included in national parks and other protected areas (thus relatively undisturbed, at least in theory). At that time, these two categories together represented 60 percent of the total closed forest area in the tropics, a proportion ranging from 39 percent in tropical Asia, to 59 percent in tropical Africa and 69 percent in tropical America. The different proportions by region reflected slower development of large-scale harvesting in tropical America compared to tropical Africa and Asia, and also the fact that, in tropical America, spontaneous colonization did not follow in the wake of timber harvesting as systematically as in the two other regions. While the sum of these two categories does not match the concept of virgin forests of Europe or of old growth forests in North America, it nevertheless gives an indication of the amount of forest disturbance which existed around 1980 in the humid tropics. Although corresponding estimates for 1990 and 1995 are not available, the share of undisturbed forests is certainly higher in the three tropical regions than in Europe and probably also than in North America.

Forest plantations

Forest plantations are tree crops which are in many ways analogous to agricultural crops. They have a simple structure and are usually restricted to one or a few species chosen for their fast growth, yield of commercial products and ease of management, which results in higher productivity than for natural forests (see Table 2). For example, while plantations in Argentina represent only 2 percent of the forest area, they account for 60 percent of the country's industrial wood production.

Table 2
Plantation area and share of total wood production from plantations in countries with important forest plantation estates

countries	share of forest plantations (percent)
	in area	in industrial wood production
Argentina	2.2	60
Brazil	1.2	60
Chile	17.1	95
New Zealand	16.1	93
Zambia	0.5	50
Zimbabwe	0.4	50

Although most plantations are intended for the production of industrial roundwood and/or fuelwood, many are established for environmental protection or other services (e.g., mainly for soil and water conservation, enhancement of agricultural production in agroforestry systems, carbon sequestration) and many also provide non-wood forest products (e.g., fodder, various foodstuffs, medicines, etc.).

Although it is difficult to separate plantations: from natural forests in most developed countries, a rough estimate of the total forest plantation in these countries is 80-100 million ha in 1995, up from: 45-60 million ha in 1980.

In almost all developing countries, on the other hand, the distinction between natural and plantation forests is more clearly drawn. In the developing world in 1995, the total 'net' forest plantation area (i.e., reported totals of annually planted areas adjusted by a survival coefficient) is estimated at some 81 million ha out of a total forest area of 1961 million ha, i.e., 4.1 percent. In 1980, it was assessed at about 40 million ha.

More than 80 percent of the forest plantations of the developing world are in the Asia-Oceania region (see Figure 3). Five of the seven countries with a forest plantation estate greater than 1 million ha in 1990 (China, India, Indonesia, Republic of Korea and Viet Nam) are from that region, the others being Brazil and Chile (see Table 3). China, with its substantial planting programmes, accounts for over 40 percent of the total plantation area of the developing world.

In the tropics as a whole, the proportion of industrial plantations established in large blocks fell from 40 percent in 1980 to 35 percent in 1990; community woodlots, farm forestry and agroforestry plantations have grown in importance during the period, particularly in the Asia-Oceania region. Some of the farm forestry and agroforestry plantations supply industrial wood markets. In India, for example, an estimated 26 000 ha of poplar (Populus deltoides) were established by farmers in Uttar Pradesh, Punjab and Haryana States in the early 1990s, the produce from which was to be sold to factories for the manufacture of matches, plywood, and other wood products (see Box 3).

Unfortunately, area figures for small private or community plantations are frequently unreliable, often being derived from the number of seedlings leaving the forest nursery, converted to area at a fixed rate that varies by country. Some countries do not record such plantings in their plantation data. Even less complete are country-level data on trees outside the forest, including those in very small woodlots and in agroforestry systems. In many countries, particularly those with limited forest area, these trees often provide the bulk of fuelwood, poles, construction wood, utility wood, as well as fodder and other non-wood forest products for household use.

Figure 3 - Forest plantations in the developing regions, 1980

Figure 3 - Forest plantations in the developing regions, 1995

Table 3
Area of plantations in developing countries with the largest forest plantation estate

countries	net estimated areaa
	(thousand ha)
China	33 800
India	14 620
Indonesia	6 125
Brazil	4 900
Viet Nam	1 470
Republic of Korea	1 400
Chile	1 015
Argentina	547
Thailand	529
Morocco	321
Bangladesh	310
Myanmar	276
Venezuela	253
Cuba	245
Philippines	203

^a These figures represent the latest reported plantation areas as of 1995, adjusted by an estimated survival rate of 70 percent. There are few large-scale inventories of forest plantations which allow for a comprehensive assessment of their condition and the level of damage caused by diseases, pests, animals and fire. A reliable assessment of plantation survival is therefore difficult, if not impossible, in most cases. FRA 1990 identified and analysed 56 plantation inventories in 18 developing countries, from which a global estimate of survival rate of 70 percent was determined,

In the FRA 1990 assessment of forest plantations in tropical countries, eucalypts (Eucalyptus spp.) accounted for 23 percent of the area, followed by pines (Pinus spp.) (10.5 percent), acacias (Acacia spp.) (7.7 percent), teak (Tectona grandis) (5.0 percent) and others (see also Box 3). Eucalypts and pines dominate in tropical America, occupying more than 80 percent of the total plantation area. Tropical African plantations are more diverse, with eucalypts and pines (mainly Pinus patula) making up about 50 percent of the forest plantation areas, with the balance composed of Acacia mearnsii (for tannin) or A. Senegal (gum arabic) and, in West Africa, Gmelina arborea and the indigenous Terminalia spp. and mahoganies (Khaya spp.). Forest plantations in the Asia-Oceania region are more diversified than either tropical Africa or America, with eucalypts and pines comprising 20 percent of the total area and teak (Tectona grandis), Acacia spp., Dalbergia sissoo, Paraserianthes falcataria, Casuarina spp., Gmelina arborea and Swietenia spp. occupying much of the rest.

The figures for forest plantations do not include those areas planted with tree species but which are generally considered in the agriculture sector. In tropical developing countries, such plantations are mainly in the Asia-Oceania region, (principally in Indonesia, Malaysia and Thailand). Of their total area of 14.1 million ha in 1990, 7.2 million ha were occupied by rubber and the remainder by coconut (4.2 million ha) and oil palm (2.7 million ha). Supply shortages and recent advances in conversion technology have made rubber wood and coconut wood important new sources of woody raw material for forest industries.

Forest plantations are expected to play an increasingly important role in national forest programmes in the future. Many developing countries, provided updated information on their present and future plantation programmes to FAO in 1996. Most of the countries with large plantation estates indicated that they intended to double their plantation areas between 1995 and 2010.

The expansion of forest plantations in all developing regions is expected to be accompanied by qualitative changes in the programmes, including:

· increasing sensitivity on the part of planners to the need for consultation and involvement of all concerned parties in planning plantation programmes;

· more integration of trees into agricultural systems to supply domestic needs and, where the price is attractive, industrial roundwood; and

· enhanced development of tree breeding programmes, some using advanced techniques but frequently incorporating ex-situ and in-situ genetic conservation of important populations.

Changes in forest cover and condition

A meaningful analysis of changes in the world's forests requires a differentiation between:

(i) the increase (by natural colonization of non-forest land with trees or by new plantings) or decrease (by deforestation) of forest area; and

(ii) the changes in forest condition, which may be positive (recovery of degraded stands, stand improvement treatments) or negative (defoliation/dieback, forest fires, unsustainable exploitation for wood, overgrazing, pests and diseases).

Changes in forest cover from 1980 to 1995 Between 1980 and 1995, the extent of the world's forests decreased by some 180 million ha, an area about the size of Indonesia or Mexico. This represents an annual loss of 12 million ha. During the 15-year period, the developing world lost nearly 200 million ha of natural forests, mostly through clearing for agriculture (shifting cultivation, other subsistence agriculture, cash crops, ranching). This was only partially compensated for by new forest plantations. Over the same period, forests in the developed world expanded slowly (by some 20 million ha) through afforestation and reforestation, including natural regrowth on land abandoned by agriculture. This increase more than compensated for the loss of forests in developed countries due mainly to urbanization and infrastructure development. The change in forest area by region over this period is shown in Figure 4.

Box 3 Planting eucalypts and poplars Eucalypts and poplars illustrate the opportunities offered by fast-growing tree species when grown in plantations. Eucalypts are mainly, but not exclusively, grown in the tropical zones; poplars are best known in the temperate zones but have an increasingly important role in the subtropics. In 1990 an estimated 10 million ha, or approximately one-quarter of tropical forest plantations, were planted with eucalypts. More than half of these plantations were located in tropical Asia and Oceania, mainly in India (4.8 million ha). It is more difficult to estimate the area of poplar plantations. It is known that in 1995 there were more than 1 million ha of poplar plantations in Europe and 1.3 million ha in China. Both eucalypts and poplars provide woody material with a wide range of end uses such as sawn timber, plywood, paper and poles, animal fodder (poplars) and medicinal products (eucalypts). There is growing interest in their cultivation for biomass energy crops. Both can, under the right circumstances, be grown successfully in agroforestry systems. These species are characterized by fast growth (when correctly matched to the site), and can be cross-bred for more rapid growth, good form, or favourable wood properties. Intensive selection for a few important economic characteristics may, however, lead to increased susceptibility to insect or disease attack and, although resistance to pests must be included in the breeding strategy, such resistance may be relatively short-lived. Breeding programmes for industrial roundwood species rely increasingly on the development of new hybrids, and on the ability to return to native or large ex-situ populations to introduce new genetic combinations. Plantations, therefore, are dependent on the conservation of a wide range of native populations. In some countries, there has been strong local opposition to eucalypts and poplar. For eucalypts, this has been due to a failure to appreciate that their high demand for soil moisture makes them unsuitable for some agroforestry systems in dry (unirrigated) areas. The alienation of villagers' land or failure to consult and involve people when creating eucalypt plantations in Southeast Asia has led to criticism and even condemnation of the genus rather than of shortfalls in the planning process. In Thailand, many farmers have planted eucalyptus on their own land for income, but have opposed its planting in forest reserves instead of native species on which they rely for a wide range of goods (e.g., fodder or non-wood forest products) that eucalyptus would not provide. Some European countries have prohibited the planting of poplars on floodplains to prevent the loss of natural vegetation and wildlife habitat and to maintain a natural appearance.

Figure 4 - Comparison of 1980 and 1995 forest areas (change shown as percentage of 1980 forest areas by region, except former USSR)

Changes in forest cover in the 1990-95 period, and comparison with 1980-90. The estimates of forest cover change in the 1990-95 period indicate a net loss of 56.3 million ha of forests (natural forests plus plantations) worldwide, representing a decrease of 65.1 million ha in developing countries which was partly offset by an increase of 8.8 million ha in developed countries (see Table 4 for developed country estimates, Table 5 for developing country estimates, and Table 3 of Annex 3 for forest cover change estimates by country). The estimated amount of forest lost annually in the developing world - 13.03 million ha - was equivalent to an annual rate of loss of 0.65 percent. Deforestation was highest in the tropical zone of the developing world. The highest annual rate of loss over this period was in tropical Asia-Oceania (0.98 percent) and the lowest (0.14 percent) was in non-tropical Asia-Oceania (see Table 3 of Annex 3).

Looking only at cover trends in natural forests in developing countries, one sees that the estimated annual loss over the 1990-95 period was 13.7 million ha, which was considerably lower than over the 1980-90 period (15.5 million ha).⁶ In other words, although the rate of loss of natural forests in developing countries is still high, it appears to have slowed somewhat since the last decade.

⁶ It should be noted that this estimate for annual loss of natural forests in developing countries during the 1980-90 period is based on new estimates for forest cover in 1980 and 1990 (see footnote 1) which are different from those made previously by FRA 1990. The hew estimate of annual loss of natural forests in developing countries - 15.5 million ha - is lower than the estimate of 16,3 million ha made by FRA 1990. In other words, although deforestation in natural forests of developing countries over the 1980s was high, it was slightly lower than previously thought.

Table 4
Forest area change in developed countries, 1990-95

regions	1990-95 (million ha)	forest area change
		annual (million ha)	(percent)
Europe	+1.94	+0.89	+0.27
Former USSR without the Russian Federation a	+2.78	+0.56	+1.12
Temperate/boreal North America	+3.82	+0.76	+0.17
Developed Asia-Oceania	+0.24	+0.05	+0.06
All developed countries (without the Russian Federation)	+8.78	+1.76	+0.12
All developed countries (without former USSR)	+6.00	+1.20	+0.18

^a The Russian Federation is not included in this table since no reliable estimate for change in forest area between 1990 and 1995 could be made for this country.

Table 5
Annual forest area changes in the developing regions, estimated for 1990-95 and 1980-90

regions	annual forest area change
	1990-95				1980-90c
	natural forests		total forestsb		natural forestsa	total forestsb
	(million ha)	(percentage of 1990 area)	(million ha)	(percentage of 1990 area)	(million ha)	(million ha)
Africa	-3.75	-0.71	-3.75	-0,71	-4.28	-4.12
tropical	-3.70	-0.72	-3.70	-0.72	-4.19	-4.10
non-tropical	-0.05	-0.41	-0.05	-0.37	-0.09	-0.02
Asia-Oceania (developing)	-4.17	-0.89	-3.47	-0.67	-4.41	-1.70
tropical	-3.51	-1.14	-3.21	-0.98	-3.97	-2.49
non-tropical	-0.66	-0.42	-0.26	-0.14	-0.45	0.79
Latin America and Caribbean	-5.81	-0.61	-5.81	-0.60	-6.77	-6.44
tropical	-5.69	-0.62	-5.69	-0.62	-6.48	-6.21
non-tropical	-0.12	-0.29	-0.12	-0.28	-0.29	-0.23
developing world	-13.73	-0.70	-13.03	-0.65	-15.46	-12.26
tropical	-12.91	-0.74	-12.59	-0.71	-14.63	-12.80
non-tropical	-0.83	-0.39	-0.43	-0.18	-0.82	0.54

^a The negative figures denote deforestation.

^b The difference between an increase in area due to plantation establishment and a decrease in area due to deforestation

^c Note that the estimates for 1980-90 are based on new information and differ slightly from those produced by FRA 1990. See footnote 1 of this section.

Conversion of forests to other land cover FRA 1990 carried out a study of forest land cover change in the tropics between 1980 and 1990 which gives some indication of the relative importance of various deforestation factors. The results have recently been published.⁷ Among the most significant outputs of the study are the so-called 'area transition matrices' of the type shown in Table 6, which indicate transfers from one land cover class to another between 1980 and 1990 over the 3 068 million ha of the tropical zone. The first row, for example, shows how much of the 1368 million ha of closed forest in 1980 remained as closed forest up to 1990 (i.e., 1 275.9 million ha), and what happened to the 92.1 million ha which were converted to other land cover classes (i.e., open forest; long fallow; fragmented forest; shrubs and short fallow; other, i.e., non-wooded land cover; and forest plantations or agricultural tree plantations). Figure 5 provides these results graphically, showing the area lost or gained by each class, while Figure 6 indicates the various types of change in forest cover and condition which occurred in each of the three tropical regions. All these transfers involve changes of woody biomass, which Figure 7 attempts to capture by assigning to each land cover class an average biomass per hectare. This study will be repeated in the next global assessment (FRA 2000) in order to analyse trends over the 1990s.

⁷ FAO, 1996, FRA 1990: Survey of Tropical Forest Cover and Study of Change Processes Forestry Paper 130. FAO, Rome,

Figure 5 - Pan-tropical summary of net changes during 1980-90 by land cover class

Table 6
Area transition matrix for the tropical zone, 1980-90

land cover classes in 1980	land cover classes in 1990 (million ha)
	closed forest	open forest	long fallow	fragmented forest	shrubs and short fallow	other land cover	plantation (agricultural and forest)	total 1980 (million ha)
closed forest	1 275.9	9.0	9,3	9.2	24.1	36.6	3.9	1 368.0
open forest	0.9	283.3	1.3	5.2	3.8	10.2	0.2	304.9
long fallow	1.1	0.3	48.6	1.1	3.2	2.2	n.s.	56,5
fragmented forest	0.6	0.6	0.6	159.33	1.9	11.7	0.4	175.1
shrubs and short fallow	0.7	0.5	0.7	0.5	273.0	26.9	0.3	302.6
other land cover	0,8	0.7	0.3	1.4	4.0	837.3	0.5	845.0
plantation	0.1	n.s.	0.0	0.0	n.s	0.1	15.7	15.9
total 1990	1 280.1	294.4	60.8	176.79	310.0	925.0	21.0	3 068.0

The figures indicate that, during the decade, and at pan-tropical level, the closed forest underwent the greatest change of all land cover classes. Of the 92.1 million ha of closed forest that underwent a change in land cover class:

· 40 percent (36.6 million ha) went to 'other land cover' which includes, among other things, permanent agriculture, cattle ranching, water reservoirs, etc., representing a complete loss of the cover and woody biomass;

· 26 percent (24.1 million ha) went into shrubs and short fallow agriculture which represents deforestation, but some woody biomass remains on the site;

· 10 percent (9.2 million ha) went to fragmented forest, which represents partial deforestation, with a loss of approximately two-thirds of the original forest cover;

· 20 percent (18.2 million ha) went into open forest or long fallow shifting cultivation, representing degradation involving the loss of biomass (approximately half);

· 4 percent (4 million ha) went to agricultural or forestry plantations.

Figure 6 - Main categories of forest cover change by geographic regions, 1980-90 (changes as percentages of 1980 regional forest area; closed and open forest)

Figure 7 - Pan-tropical area and change in forest area and above-ground woody biomass

The other land cover classes that underwent significant changes are open forest, shrubs, and fragmented forest. Almost all (90 percent) of the changes to the four land cover classes resulted in a decrease in biomass.

The nature of the changes detected by the study provides some useful indications of major ultimate causes of change. Distinct patterns are evident in the different regions, namely:

· Changes in forests in Africa are dominated by transition from closed forest, through intermediary stages of depletion, to shrub and short fallow. This would indicate an extension of subsistence farming under the pressure of rural population growth.

· Changes in forests in Latin America are dominated by a more abrupt change involving deforestation from closed forest to other land cover, seemingly due to the effect of centrally-planned operations such as government resettlement schemes, large-scale cattle ranching and hydroelectric reservoirs.

· Forests in Asia are undergoing two types of changes of nearly equal measure: gradual changes as a result of rural population pressure (i.e., intensification and expansion of shifting cultivation practices); and abrupt changes due to centrally-planned operations including government resettlement schemes and large plantation programmes.

The above-mentioned study provides valuable information on the nature of vegetation change in tropical forests that can be detected through remote sensing imagery. Certain information on the processes and effects of forest cover change, however, cannot be obtained through a global study, and, therefore, should be complemented by other types of studies carried out at the national level. National and sub-national studies are needed to increase the understanding of the underlying causes of change and change processes, the factors contributing to forest cover degradation, and the effects of vegetation change on other forest resources.

Changes in forest condition

Developed countries

Although total forest area is slowly increasing in all parts of the developed world, there has been no significant improvement in some aspects of forest condition. Even though the widespread death of European forests due to air pollution which was predicted by many in the 1980s did not occur, deteriorating forest condition remains a serious concern in Europe and North America. A 1995 study carried out by UN-ECE and EU suggests that the direct causes of the decline of forest condition in Europe are drought, heat, pests, game and grazing, and human action, whereas pollution is an indirect cause, predisposing forests to decline (see the 'Regional highlights' for Europe and the former USSR in Part 4). Forest damage, which may appear moderate on a continental scale, is severe in some parts of central and eastern Europe where at least 100 000 ha of forests have died in the last 20 years. Most of these forests have subsequently been reforested with species or provenances better adapted to the site. Air pollution is said to play a role in the decline of various forest species in other industrialized countries, for example: sugar maple (Acer saccharum) in eastern Canada; the high elevation forests of red spruce (Picea rubens) in the eastern USA; and Cryptomeria japonica stands in Japan.

Forest fires have become smaller but more frequent in the industrialized countries over the last decade. There has been a slight reduction in total area burned per year in the developed world as a result of improved prevention, detection and control systems (although there has been an increase of 15 percent in the former USSR). The average area of forest and other wooded land affected per year by fires in Europe, North America and the former USSR together was about 4.26 million ha, or 0.22 percent of their total area of forest and other wooded land in 1990.

Pests and diseases are often the main agents of forest decline, and remain constant threats to forests. Until recently, plantation growing stock was based on relatively limited genetic variability which increased the vulnerability of the plantations to pests and diseases. Recent problems include the decline of various species of oaks in many parts of Europe and central Russia, the recurrent infestation of forests in Poland by the nun moth, the attack on beech by the beech scale in western and central Europe, birch and ash diebacks in the north-eastern USA, the 'little-leaf disease of short-leaf pine in the southern USA, and the 'x-disease' of pines in southern California.

Developing countries

Because reduction in forest area is the most striking aspect of forest change in developing countries, less attention is paid to forest condition. Although the extent and degree of overharvesting of fuelwood and overgrazing are difficult to quantify, they are recognized as major causes of forest degradation in some developing countries, particularly in arid and semi-arid zones. Fires, pests and diseases are also critical factors in forest health in these countries.

Every year, very large areas of savannah-type, mixed forest/grassland formations are affected by fires, particularly in the dry zones of Africa and South America, but no reliable data are available on their extent. Although less prone to fire, forest areas in the humid tropics are also affected by large fires, the most serious in recent history being in East Kalimantan (Indonesia) in 1993 which burned 3.6 million ha. Coniferous forests in the humid tropics are often affected by fires; in the 1980s the area of pine forest in Honduras and Nicaragua burned annually amounted to some 3.5 percent of the countries' total pine forest area, and important fires have burned natural and human-made tropical pine forests in Mexico and Indonesia (Northern Sumatra). Fire is also a permanent threat to forests in the sub-temperate and temperate zones of North Africa and South America (including Southern Brazil). From 1950 to 1990, fires in China are reported to have affected annually an average of 890 000 ha, the most damaging having been the 'May 6' fire which burned some 1.85 million ha in the north-eastern province of Heilongjiang in 1987. In the absence of a global statistical fire database, it is difficult to provide an overall estimate of the annual extent of fires in forests and other wooded lands. A crude estimate for the temperate/sub-temperate and humid tropical zones of the developing world - leaving aside the significant dry tropical zone for which little reliable information exists - would be of the order of 2 million ha of forest and other wooded land per year during the 1980s. Given the lack of adequate fire prevention and control capacity in most developing countries, no significant reduction of wooded areas burned is likely to occur in the near future.

Pest and disease outbreaks in developing countries are surveyed and reported mostly for plantations and planted trees. Examples include: the Leucaena psyllid, Heteropsylla cubana, which has spread into Asia and the Pacific islands and is now spreading across Africa; the cypress aphid, Cinara cupressi, which is established in eight eastern and southern African countries and is causing heavy mortality among a number of exotic and indigenous species, especially the important plantation species Cupressus lusitanica; and the European woodwasp, Sirex noctilio, which has spread into Argentina, Uruguay and southern Brazil, affecting principally Pinus taeda, but which may become a threat to the large Chilean plantation estate of Pinus radiata.

There is evidence of forest decline associated with air pollution also in the developing world. This is likely to be increasingly important as the industrial and transportation infrastructure develops. Examples include the decline of neem (Azadirachta indica) in the Sahel, of framiré (Terminalia ivorensis) in Cote d'Ivoire and Ghana, of forests near large cities and industrial areas in China (e.g., Pinus massoniana stands near Nanshan), and of Eucalyptus globulus plantations in Colombia and Peru. Despite the overall significance of the corresponding damage and losses, surveys of forest declines and diebacks in developing countries are relatively rare.

Effects of timber harvesting on forest condition

An issue of intense discussion and considerable polarization of opinion in both developed and developing countries is the effect of timber harvesting on the condition of natural and semi-natural forests. In the words of one recent publication: 'The most urgent, and certainly the most emotive, environmental issue relating to the timber trade concerns the logging of natural or semi-natural forests.... Here is the real front line in the forest debate. On one side are the conservation activists, blocking logging roads, hugging trees, lobbying governments and organizing petitions, often backed up by sympathetic media and significant sections of public opinion. On the other side are timber traders, logging companies and often the forest workers themselves, arguing that forest management is essential for jobs, economic security and resources, and that modern logging techniques allow a sustainable yield of timber and cause little long-term environmental damage'.⁸

⁸ Dudley, N., Jeanrenaud J-P and Sullivan F. 1995, Bad Harvest? The Timber Trade and the Degradation of the World's Forests. Earthscan Publications Limited, London, UK

The concerns revolve around, among other things: the effects of overharvesting and poor harvesting practices on the future productive capacity of the forest through changes in the tree species composition and damage to the soil (see the discussion of forest harvesting in the section 'Trends in forest management and utilization'); the impact on forest plants and animals through habitat disruption caused by harvesting, other forest management operations, and by forest roads; the effect on water quality and aquatic organisms due to siltation of waterbodies and harvesting debris; and, in some cases, the increased risk of windblow, forest pests and diseases. While it is difficult to quantify the extent and degree of such degradation on a global scale, the forestry literature cites many examples of poor harvesting methods and of overharvesting leading to negative environmental impacts in many parts of the world.⁹

⁹ ibid (see citations within)