Accueil » Forêts » Niveau 3 » Question 2


2. How much forest is there on the planet and at what rate is it disappearing?

  • 2.1 How much of the planet is covered with/by forests?
  • 2.2 How fast are forests disappearing?
  • 2.3 How much is there of the different kinds of forests?

The source document for this Digest states:


Extent of forest resources is the first of the thematic elements characterizing sustainable forest management. Generally speaking, it refers to the overall goal of maintaining adequate forest cover and stocking – of various forest types and characteristics including on ‘other wooded land’ and as ‘trees outside forests’ – to support the social, economic and environmental objectives related to forestry within a country or region. The ultimate aim of monitoring the extent and characteristics of forest resources is to reduce unplanned deforestation, restore and rehabilitate degraded forest landscapes, manage forests sustainably and evaluate the important function of carbon sequestration by forests, other wooded land and trees outside forests – thereby contributing to moderating the global climate (FAO, 2005d).

Information on the extent of forest resources has formed the backbone of all global forest resources assessments and continued to be a major topic in FRA 2005. Forest area is an easily understood baseline variable, which provides a first indication of the relative importance of forests in a country or region. Estimates of change in forest area over time provide an indication of the demand for land for forestry and other land uses, as well as of the impact of significant environmental disasters and disturbances on forest ecosystems. As mentioned previously, the proportion of land area under forests is also used in the Millennium Development Goals indicator process (United Nations, 2005a)

However, as was observed in FRA 2000 (FAO, 2001b), the significance of forest area as a single indicator of forest development has often been overemphasized, particularly in the public debate, where other aspects of forest resources feature less prominently. The most commonly quoted result from global forest resources assessments continues to be the global net loss of forest area. However, it is important to note that many other parameters and scales must be considered in determining the relevant trends in the extent of forest resources. Growing stock and carbon storage may be considered equally important parameters, as they indicate whether forests are degraded and to what extent they mitigate climate change. Further, the net loss of forest area is not in itself sufficient to describe land-use dynamics that include both loss of forests due to deforestation and natural disasters and gains in forest area from planting or natural expansion.

For FRA 2005, information was sought on the current status and changes over time of the following four variables:

1 See Annex 2 for exact definitions

Figure 2.1 illustrates the availability of information for these variables at the global level.

In regional and ecoregional criteria and indicator processes, as well as in national reports, more-detailed classifications of the forest area are often used, e.g. according to forest or vegetation type, age structure or diameter distribution classes. Because of the varying conditions and classification systems among countries and regions, it was not feasible to report on such classifications at the global level. However, country reports for FRA 2005 contain considerably more detail than is shown in the global tables. Moreover, thematic studies have been prepared on planted forests, mangroves and bamboo that provide in-depth knowledge on these forest types and groups of species.

In FRA 2000, an independent remote sensing survey was carried out to supplement country reporting for the pan-tropical region. The results constituted an important ingredient in the analysis of global and regional trends, leading, for example, to a calibration of reported changes in forest area for Africa. The survey also provided considerable insight into change processes in land use, including the documentation of different patterns of land-use change in tropical regions. The results have been widely acknowledged and used (e.g. Mayaux et al., 2005). While no similar project was carried out for FRA 2005 owing to lack of resources, preparations have been made for a more ambitious approach (FAO, 2003d) that takes a broader range of information requirements into account. This approach is being considered for the next global forest resources assessment (FRA 2010).

Source & ©: FAO  Global Forest Resources Assessment 2005, Progress towards sustainable forest management, Chapter 2: Extent of forest resources, p.11-12

2.1 How much of the planet is covered with/by forests?

The source document for this Digest states:


Forest area provides the first indication of the relative importance of forests in a country or region, while estimates of forest area change over time provide an indication of the demand for land for forestry and other land uses, and may also illustrate the impact of significant environmental disasters and disturbances on forest ecosystems. Forest area is relatively easy to measure, and this variable has therefore been selected as one of the 48 indicators for monitoring progress towards the Millennium Development Goals agreed by the United Nations (particularly Goal 7 – Ensuring environmental sustainability).

Data on the status of and trends in area of forest are crucial to decisions related to forest and land-use policies and resource allocations, but they need to be combined with information on the health and vitality of forests and their socio-economic and environmental functions and values. Other sections of this report deal with these aspects.

Information availability

Information on the extent of forests was provided by 228 of the 229 countries and areas reporting for FRA 2005 – the exception being the Marshall Islands, for which no quantitative information was available. Antarctica and some of the smaller dependent territories, which do not have, or have no significant, forest area were not included in the list of reporting units for FRA 2005.

Four countries or areas (Guam, Guyana, Lebanon and the Occupied Palestinian Territory) did not provide an estimate of forest area for 1990. All other countries and areas provided estimates for all three reporting years (1990, 2000 and 2005). For the purpose of analysis, the 1990 forest area for each of these four countries and areas was estimated by FAO based on a linear extrapolation of the figures provided for 2000 and 2005.

Since extent of forest resources is a key variable for decisions regarding forest policy and investments in the forestry sector, almost all countries and areas provided information on this variable. However, some countries had comprehensive information from only one point in time (see Table 2 in Annex 3), while others had estimates that were incompatible, making trend analyses difficult.

Information on the extent of other wooded land as of 2005 was available from 180 countries and areas, which together account for 64.9 percent of total forest area. Only 61 countries and areas reported on current extent of other land with tree cover, which is a new variable in global forest resources assessments. It aims to capture those areas in which forest cover criteria are met, but the predominant land use is agricultural (e.g. orchards and oil-palm plantations) or urban (e.g. urban parks).


Total forest area as of 2005 is estimated at 3 952 million hectares or 30 percent of total land area. This corresponds to an average of 0.62 ha per capita. As can be seen from Figure 2.2, the area of forest is unevenly distributed. For example, 64 countries with a combined population of 2.0 billion have less than 0.1 ha of forest per capita.

Based on available information, total area of other wooded land is estimated to be at least 1 376 million hectares – about one-third of total forest area. This category suffered from reclassification problems, particularly in dry zones such as those in Australia, Kenya and the Sudan, in which the distinction between forest and other wooded land is not very clear. Total area of other land with tree cover is at least 76 million hectares. These two estimates, particularly the latter, were limited by lack of information, and the true extent of other land with tree cover is undoubtedly much higher.

Distribution of forests. A subregional summary of the distribution of forests is shown in Table 2.1. Europe accounts for one-quarter of total forest area, followed by South America and North and Central America with 21 and 18 percent respectively. Information on the area of forest and other wooded land by country can be found in Table 3 in Annex 3.

Forest-rich and forest-poor countries. The five most forest-rich countries (the Russian Federation, Brazil, Canada, the United States and China) account for more than half of total forest area (2 097 million hectares or 53 percent). The Russian Federation alone accounts for 20 percent of the world total. Seven countries have more than 100 million hectares of forest each. The ten most forest-rich countries account for 66 percent of total forest area (Figure 2.3). The remaining 34 percent is spread among 212 countries and areas. Seven countries and areas (the Falkland Islands, Gibraltar, the Holy See, Monaco, Nauru, South Georgia and South Sandwich Islands and Tokelau) reported having no areas that qualify as forests using the FRA 2005 definition.

High and low forest cover countries. Forty-five countries and areas have more than half their total land area covered by forests (Figure 2.4), and 11 of these have more than 75 percent of their total land area covered. Most of these are small island states or territories, but the list also includes three low-lying coastal states in South America and one country in the Congo Basin (Table 2.2).

Sixty-four countries and areas have less than 10 percent of their total land area covered by forests. These include many SIDS and dependent territories, as well as 17 larger countries with relatively substantial forest areas (more than 1 million hectares each). Three of these (Chad, the Islamic Republic of Iran and Mongolia) have more than 10 million hectares of forest, but still qualify as LFCCs.

At the regional level, South America is the region with the highest percentage of forest cover, followed by Europe and North and Central America. Asia is the region with the lowest percentage of forest cover (Table 2.3).

Source & ©: FAO  Global Forest Resources Assessment 2005, Progress towards sustainable forest management, Chapter 2: Extent of forest resources, p.14-17

2.2 How fast are forests disappearing?

The source document for this Digest states:


Figure 2.5 is a simplified model illustrating forest change dynamics. It has only two classes: forests and all other land. A reduction in forest area can happen through either of two processes. Deforestation, which is by far the most important, implies that forests are cleared by people and the land converted to another use, such as agriculture or infrastructure. Natural disasters may also destroy forests, and when the area is incapable of regenerating naturally and no efforts are made to replant it, it, too, reverts to other land.

An increase in forest area can also happen in two ways: Either through afforestation, i.e. planting of trees on land that was not previously forested, or through natural expansion of forests, e.g. on abandoned agricultural land – which is quite common in some European countries. Where part of a forest is cut down but replanted (reforestation), or where the forest grows back on its own within a relatively short period (natural regeneration), there is no change in forest area.

For FRA 2005, countries were asked to provide information on their forest area for three points in time. This allows calculation of the net change in forest area over time. This net change is the sum of all negative changes due to deforestation and natural disasters and all positive changes due to afforestation and natural expansion of forests.

The total net change in forest area in the period 1990–2000 is estimated at -8.9 million hectares per year – equivalent to a loss of 0.22 percent of the remaining forest area each year during this period.

The total net change in forest area in the period 2000–2005 is estimated at -7.3 million hectares per year – an area the size of Panama or Sierra Leone – or equivalent to a loss of 200 km2 of forest per day. Compared to the 1990s, the current annual net loss is 18 percent lower and equals a loss of 0.18 percent of the remaining forest area each year during this period.

Countries were not requested to provide information on each of the four components of net change, as most countries do not have such information. This, however, makes estimation of the deforestation rate difficult and no attempt has been made to do so at the country level. Rather, an estimate of the global deforestation rate has been made as follows:

The total net loss for countries with a negative change in forest area was 13.1 million hectares per year for 1990–2000 and 12.9 million hectares per year for 2000–2005. This would indicate that annual deforestation rates were at least at this level. Since the net change rate takes into account afforestation efforts and natural expansion of forests, the rate of deforestation might be higher still. On the other hand, Brazil, which accounts for 21 percent of the total net loss in the period 1990–2000 and 24 percent in 2000–2005, calculated its forest area in 2005 and 1990 based on information from 2000 and the sum of annual figures of the area of forests cleared. It did not take into account to what extent the land use of these areas had changed and to what extent cleared lands had been abandoned and had reverted to forest through natural regeneration. Such naturally regenerated secondary forests are thought to be quite extensive, but insufficient information is available to estimate the extent. Thus the area of deforestation and the net loss of forests in Brazil are likely overestimated.

Taking these considerations into account, the global deforestation rate was estimated at 13 million hectares per year during the period 1990–2005, with few signs of a significant decrease over time.

In summary, deforestation continues at an alarming rate – but the rate of net loss is decreasing due to afforestation and natural expansion of forests in some countries and regions.

Trends in area of other wooded land were analysed, based on the 171 countries and areas providing information for all three reporting years. The analysis indicates that other wooded land is more or less constant in North and Central America and Oceania. In Europe and South America, it decreased in the period 1990–2000, but remained almost constant in the period 2000–2005. It decreased in both periods in Africa and Asia. At the global level, area of other wooded land decreased by about 3.3 million hectares per year over the past 15 years. This finding should be treated with caution, however, since many countries do not have compatible information over time for other wooded land, and thus one estimate was frequently used as the best available figure for all three reporting years. Data for other land with tree cover were too limited to allow trend analysis.

Regional and subregional comparisons. Table 2.4 and Figure 2.6 show the changes in area of forest by region and subregion. South America suffered the largest net loss of forests from 2000 to 2005 – about 4.3 million hectares per year – followed by Africa, which lost 4.0 million hectares annually. While there are signs that the net loss in Africa is decreasing, it seems to be increasing in South America – primarily due to a reported increase in the net loss of forests in Brazil. However, as indicated above, the net loss reported by Brazil for both periods may be overestimated. Efforts are currently underway to design and implement a national forest assessment on a pilot basis in Brazil, which should yield better information for the next global forest resources assessment.

North and Central America and Oceania each had a net loss of about 350 000 ha, with a decreasing trend in Oceania, and a slightly increasing trend in North and Central America – the latter primarily owing to a decrease in the plantation establishment rate in the United States (down from an average of 596 900 ha per year in 1990–2000 to an average of 157 400 ha per year in the period 2000–2005) and the continued, albeit decreasing, net loss of forests in Mexico.

Asia, which had a net loss of some 800 000 ha per year in the 1990s, reported a net gain of 1 million hectares per year from 2000 to 2005, primarily as a result of the largescale afforestation reported by China. Forest areas in Europe continued to expand, although at a slower rate than in the 1990s. For information on changes in forest area by country, see Table 4 in Annex 3.

Countries with large positive or negative changes. In the Caribbean, Europe, North America, Oceania and Western and Central Asia, a majority of countries have no major changes over the last five years, while in Africa a majority of countries have a negative change rate (Figure 2.7).

A large number of countries in Oceania and the Caribbean have reported no major change, primarily because of lack of data and particularly for more than one point in time. The ten countries with the largest net loss per year in the period 2000–2005 had a combined net loss of forest area of 8.2 million hectares per year (Table 2.5).

The ten countries with the largest net gain per year in the period 2000–2005 had a combined net gain of forest area of 5.1 million hectares per year due to afforestation efforts and natural expansion of forests (Table 2.6). The large increase in forest area for China is due to recent, large-scale afforestation programmes.

Thirty-seven countries and areas have an estimated net negative change rate of 1 percent or more per year. The ten countries with the largest annual net negative change rates for 2000–2005 are: Comoros (7.4 percent); Burundi (5.2 percent); Togo (4.5 percent); Mauritania (3.4 percent); Nigeria (3.3 percent); Afghanistan (3.1 percent); Honduras (3.1 percent); Benin (2.5 percent); Uganda (2.2 percent) and the Philippines (2.1 percent).

Eighteen countries have an estimated annual positive change rate of 1 percent or more due to natural expansion of forests and afforestation. The ten countries with the largest estimated annual positive change rates for 2000–2005 are: Rwanda (6.9 percent); Iceland (3.9 percent); Bahrain (3.8 percent); Lesotho (2.7 percent); Kuwait (2.7 percent); Egypt (2.6 percent); China (2.2 percent); Cuba (2.2 percent); Viet Nam (2.0 percent) and Tunisia (1.9 percent).

Most but not all of the countries with large change rates measured in percentages are LFCCs or countries with a limited forest area, where a relatively small change in absolute values results in a large change in relative or percentage terms.

Comparison with previous estimates

Countries were asked to provide estimates for three points in time for FRA 2005: 1990, 2000 and 2005. The figures provided for 1990 and 2000 are likely to differ slightly from those reported for the previous assessment (FRA 2000) for the following reasons:

First, the estimates presented in both assessments are derived primarily through linear interpolation and extrapolation of the results from two or more recent assessments. National forest resources assessments are fairly expensive, thus they are often carried out at infrequent intervals and a new data set can significantly change previous forecasts based, for example, on estimates from the 1970s or 1980s.

Second, many more countries were actively involved in the FRA 2005 process than in previous assessments, and the national correspondents helped provide access to better and more recent information, while their detailed knowledge of forest types helped improve the reclassification of data into FRA 2005 categories.

Table 2.7 shows a comparison of the results provided in FRA 2000 and those reported in FRA 2005 for reporting years 1990 and 2000.

Globally, total forest area estimated in FRA 2005 for 1990 and 2000 was about 3 percent higher than that in FRA 2000. This was primarily owing to reclassification of unproductive forests in Canada and the United States (previously classified as other wooded land), but also to new and better information from other countries.

Most countries provided estimates of forest area that differed from those provided for FRA 2000. Many differences were minor and due to calibration of areas to match the official land areas as found in the FAO database FAOSTAT (FAO, 2005a). Others were due to reclassifications or to new and better information and, in some cases, resulted in significantly different figures.

A total of 79 countries provided estimates for 1990 for FRA 2005 that differed by more than 10 percent from those presented for FRA 2000. Similarly, a total of 85 countries provided new figures for 2000 that differed by more than 10 percent from those presented for FRA 2000. A separate working paper has been prepared explaining these differences (FAO, 2006a).

Annual net loss of forests in the 1990s appears to have been overestimated in previous studies. FRA 2000 estimated the annual net change in global forest area to be -9.4 million hectares per year for the period 1990–2000. FRA 2005 estimates the rate for the same period to be -8.9 million hectares per year, i.e. half a million hectares less per year.

The main differences are found in Africa, where the net loss is 1 million hectares lower than previously estimated, and in Asia, where FRA 2005 estimates a higher loss for the 1990s than previously reported, primarily due to a revised change rate for Indonesia, based on more recent information.

For Africa, the results for FRA 2005 are closer to the results of the independent remote sensing analysis done for FRA 2000, which indicated that the net annual loss was -2.2 million hectares, while the reports indicated a net loss of -5.5 million hectares. However, the net loss of 4.3 million hectares reported for FRA 2005, which is based on national reports, may still be overestimated.

Source & ©: FAO  Global Forest Resources Assessment 2005, Progress towards sustainable forest management, Chapter 2: Extent of forest resources, p.18-23

2.3 How much is there of the different kinds of forests?

The source document for this Digest states:


The request for information on forest characteristics aimed to provide more detailed information on the kinds of forest that exist, in terms of their ‘naturalness’ or the intensity of silviculture and management practices. A continuum exists from primary forests with no – or no visible – indications of past or present human activity to intensively managed forest plantations of introduced species, primarily managed for a single product, often on a relatively short rotation. Between these two extremes lies a range of scenarios, and there are no clear cut-off points between possible classes along the continuum.

Countries were asked to characterize their forests and other wooded land according to five classes: primary, modified natural, semi-natural, protective forest plantation and productive forest plantation.

The first three classes comprise native forest tree species only, with the possible exception of small areas of natural regeneration of introduced or naturalized species in the semi-natural class. While the origin of primary and modified natural forests is natural regeneration, semi-natural forests are established through assisted natural regeneration, planting or seeding, while all forest plantations are established through planting or seeding.

Planted forests thus comprise all forest plantations and parts of semi-natural forests. All planted forests of introduced species were classified as forest plantations in FRA 2005. Planted forests of native species were classified as forest plantations if characterized by few species, straight, regularly spaced rows and/or even-aged stands. If they resembled natural forests of the same species mix, such as many planted forests in Europe, they were classified as semi-natural forests.

A thematic study on planted forests, including the planted-forest component of both semi-natural forests and forest plantations, is being completed for release during 2006 to complement the data available in FRA 2005 (Box 2.1).

The use of the five different classes helps clarify the extent to which forests are human-made or -modified, while at the same time providing an indication of the intensity of management and the potential for wood production, e.g. for use in global fibre supply models.

The typical modified forest is a tropical forest in which selective logging has taken place, but no silvicultural measures have influenced the natural regeneration of species. The typical semi-natural forest might be a temperate forest in Europe or a teak forest in Asia, in which the harvesting is much more intense, removing a larger volume and number of trees per hectare, and with specific interventions aimed at securing a desirable future species mix through assisted natural regeneration, seeding or planting of native species.

Forest plantations may be established for different purposes and have been divided into two classes, with protective forest plantations typically being unavailable for wood supply (or at least having wood production as a secondary objective only) and often consisting of a mix of species managed on long rotations or under continuous cover.

This section provides an overview of status and trends as related to forest characteristics. More detailed information on primary forests can be found in the chapter on biological diversity, while analyses of productive and protective forest plantations can be found in the respective chapters on these themes.

Information availability

Although a large number of countries reported on the characteristics of their forests, information on all five classes was not always readily available, because countries either did not collect information or used a different national classification system. Proxy values have often been used, which makes a detailed analysis of status and trends difficult.

Information was unavailable for many of the countries in the Congo Basin, the second largest expanse of tropical forest, and this should be kept in mind when analysing the findings.

Few countries had information on the area of primary forests. Some used the current area of forests in national parks and other protected areas as a proxy value or provided an expert estimate of the percentage of natural forests that could be considered primary according to the definition used for FRA 2005. There were also some inconsistencies in reporting planted forests of native species: some countries reported these as semi-natural forests, while others preferred to include them as forest plantations. Thus it may not be possible to directly compare figures for different countries, owing to differences in interpretation of the classification systems.

Of 229 countries and areas reporting, 174 reported on the characteristics of their forests. Their combined forest area was estimated at 3 678 million hectares – equivalent to 93 percent of the total forest area of the world (Figure 2.8).

Of the 180 countries providing information on the area of other wooded land, 114 provided information on characteristics.


More than one-third (36 percent) of total forest area is classified as primary forest, i.e. forest of native species, in which there are no clearly visible indications of human activity and ecological processes are not significantly disturbed (Figure 2.9).

Great variation exists in terms of the distribution of primary forests, with limited areas reported from the Caribbean, Europe (excluding the Russian Federation) and the arid zones of Eastern and Southern Africa, Northern Africa and Western and Central Asia. The largest expanse of primary forest is found in South America (the Amazon). Countries in North and Central America and the Russian Federation have also classified a relatively high proportion of their forests as primary.

Slightly more than half of all forests (53 percent) are considered modified natural forests (forests of naturally regenerated native species in which there are clearly visible indications of human activity) and 7 percent are classified as semi-natural forests (forests comprising native species, established through planting, seeding or assisted natural regeneration).

Forest plantations constitute an estimated 4 percent of forest area (forests of introduced species, and in some cases native species, established through planting or seeding), classified either as productive (3 percent of total forest area) or protective (0.8 percent of total forest area).

The vast majority of other wooded land (69 percent) was classified as modified natural, 28 percent as primary and the remaining 3 percent as semi-natural.


A trend analysis was generated based on the 167 countries providing estimates for all three reporting years,2 including those reporting no primary forest.

2 This list of countries excludes the Russian Federation (see comment related to primary forests in footnote 3). Australia did not provide information for all categories for 1990; its primary forest has been assumed to be constant and the remaining forest area not classified as forest plantation has been assumed to be modified natural forest based on information from 2000 and 2005.

As can be seen in Figure 2.10, the areas of primary forest and modified natural forest are decreasing, while the areas of semi-natural forest and forest plantation are increasing.

About 6 million hectares of primary forest have been lost or modified each year since 1990,3 and there is no indication that the rate of change is slowing down. This rapid decrease stems not only from deforestation, but also from modification of forests due to selective logging and other human interventions – whereby primary forests move into the class of modified natural forests. The rate of loss of primary forests is stable or slightly decreasing in most subregions, but is increasing in South America and, to a lesser extent, in North America.

3 This estimated net loss excluded the Russian Federation, in which a large difference in the change rate (from -1.6 million hectares per year in the 1990s to +0.5 million hectares per year in the last five years) is likely due to a change in the methodology used, rather than being a reflection of actual change.

Brazil and Indonesia alone account for an annual loss of primary forest of 4.9 million hectares. The data collected do not permit an analysis of how much of this net loss is due to deforestation and how much is owing to areas of forest moving into the modified natural forest class.

A number of countries registered positive change rates in the area of primary forests, including several European countries and Japan (see Table 9 in Annex 3). In most of these cases, countries have been setting aside natural forest areas in which no intervention is to take place. With time, these areas evolve into forests in which there are no clearly visible indications of human activity and ecological processes are not significantly disturbed, which is the definition of primary forests used in FRA 2005. Japan and some of the European countries, for example, classified all natural forests over a certain age or size as primary forests if no interventions had been conducted in the last 25 years. There has been an increase in the area of forest plantations of about 14 million hectares in the last five years, or about 2.8 million hectares per year, 87 percent of which are in the productive class.

Information availability on the characteristics of other wooded land was insufficient for analysing trends over time.

Forest types and species groups. In addition to the thematic study on planted forests mentioned above, two studies on specific forest types and species groups were undertaken to complement the FRA 2005 main report – one on mangroves (Box 2.2) and another on bamboo (Box 2.3).

The total area of mangroves was estimated at 15.2 million hectares, down from 18.8 million hectares in 1980. An estimated 47 percent of this area was found in five countries: Indonesia, Australia, Brazil, Nigeria and Mexico.

As mentioned earlier, the area of bamboo is difficult to assess, as these species often occur as patches within or outside forests. Nevertheless, preliminary findings based on reports from 30 of the main bamboo-rich countries indicate that total area of bamboo amounts to some 40 million hectares – or 1 percent of the global forest area – and it is increasing.

Source & ©: FAO  Global Forest Resources Assessment 2005, Progress towards sustainable forest management, Chapter 2: Extent of forest resources, p.23-28

Publications A-Z

Get involved!

This summary is free and ad-free, as is all of our content. You can help us remain free and independant as well as to develop new ways to communicate science by becoming a Patron!