Table of Contents

ACKNOWLEDGEMENTS

ABBREVIATIONS

ABSTRACT

INTRODUCTION

MAJOR FINDINGS

Forest area

Country information
Pan-tropical remote sensing survey
Issues

Volume and Biomass

Forest plantations

Trees outside forests

Forest management

Forest fires

Protected areas

Removals

Non-wood forest products

Global maps

Other assessment parameters

FORESTRY INFORMATION SYSTEM

CONCLUSION

REFERENCES

APPENDIX 1 - BASIC DEFINITIONS FOR FRA 2000

APPENDIX 2 - COUNTRY TABLES

ACKNOWLEDGEMENTS

The FRA 2000 represented a major effort of the FAO Forestry Department, member countries, donors, partners and individual experts. Countries provided the basic data for the assessment in the form of technical reports and analyses. Many contributed staff to interpret satellite imagery and conduct technical work to assess forest and ecological conditions. Several countries donated to the assessment, including Austria, Denmark, Finland, France, Italy, Japan, Sweden, Switzerland, and the United Kingdom. Valuable in-kind contributions were provided by Sweden, India and the United States. The United Nations Economic Commission for Europe (UN-ECE) collected the basic information for industrialized countries.

Important partners included the Australian Bureau of Rural Sciences; the Autonomous National University of Mexico; the Bishop Museum (US); the Brazilian Institute of Environment and Natural Resources; the Canadian Forest Service; CIRAD-For�t (France); the Department of Remote Sensing and Landscape Information Systems, University of Freiburg (Germany); the EROS Data Center (US); the European Forest Institute (Finland); the Faculty of Agricultural and Forest Sciences of Tuscany (Italy); the Federal Agency for Nature Conservation (Germany); the Federal Research Institute for Forestry and Wood Products (Germany); the Forest National Corporation of Sudan; the Forest Survey of India; the Foundation for American Friendship (Argentina); the Institute of Remote Sensing Applications of China; the International Institute for Applied Systems Analysis (Austria); the Italian Overseas Institute; the National Aeronautics and Space Administration (US); the National Board of Forestry (Sweden); the Regional Community Forestry Training Center (Thailand); the State University of New Jersey, Rutgers; the Swedish University of Agricultural Sciences; the Tropical Agriculture Research and Higher Education Center (Costa Rica); the Tropical Science Center (Costa Rica); the United Nations Environment Programme (UNEP); the United States Department of Agriculture, Forest Service; the University of Maryland (US); the University of Michigan (US); and the UNEP-World Conservation Monitoring Centre (UNEP-WCMC).

FAO is grateful for the support of all countries, organizations and experts inside and outside of the organization that made the assessment possible.

ABBREVIATIONS

ABSTRACT

The paper presents an overview of findings from the Global Forest Resources Assessment 2000 (FRA 2000). FRA 2000 was based on an exhaustive survey of inventory reports from around the world dialogues with national forestry experts and application of a core global set of definitions. The assessment concluded that the world's forest cover at the year 2000 was 3.86 billion hectares compared to FAO's previous estimate of 3.45 billion hectares in 1995. However, the two estimates are not directly comparable due to changes in definitions and the information base.

Net annual deforestation at the global level is still high at 9 million hectares, with gross deforestation estimated at 13.5 million hectares per annum. This is a significantly lower net rate compared to FAO's previous report for the period 1990-1995 (11.3 million ha per year), partly due to improved datasets. Key factors contributing to the estimate of lower net forest loss are attributed to natural regeneration of forests in industrialized countries and high rates of plantation establishment in Asia, particularly in China and India. However, the large-scale conversion of forests to other land uses was not significantly lower in the tropics between the 1980s and the 1990s.

FRA 2000 showed that the concept of sustainable forest management (SFM) continues to gain momentum around the world. Most countries are involved in international initiatives related to SFM and areas under forest management plans are increasing. But statistics on key indicators of SFM are still missing from a large number of countries. Forest and forestry information has increased in quantity over the past decade, but studies within the current assessment indicate that primary information and first-hand knowledge may not have improved significantly at the global level.

Comprehensive reports are found at www.fao.org/forestry/fo/fra/index.jsp on the FAO website.

INTRODUCTION

1. The Global Forest Resources Assessment 2000 (FRA 2000) was the most comprehensive in its fifty year history and for the first time one global definition of forest was agreed upon and used throughout the world. The assessment was a joint endeavour carried out by FAO in co-operation with its member countries and a number of partners. The agenda for FRA 2000 was developed by some of the world's leading forest inventory specialists at an expert consultation in Kotka, Finland, in 1996 (Nyyss�nen & Ahti 1997). In 1997, the FAO Committee on Forestry (COFO) and the fourth session of the Intergovernmental Panel on Forests (IPF-IV) approved the consultation's findings and recommended that FAO serve as the lead agency for the assessment, working in partnership with other institutions in its execution. An important partner was the United Nations Economic Commission for Europe (UN-ECE), which served as the focal point for the assessment of the industrialized temperate and boreal countries. The synthesis of information from the industrialized and developing countries together constitutes the global assessment.

2. FRA 2000 compiled and analysed all available information about the extent, composition, protection and utilization of forests for each country. Special attention was given to estimating the rate of change of forest cover and to documenting the various factors implicated in this. The assessment was conducted as a transparent and highly participatory process, which will publish all background material and analyses. It is the most comprehensive, reliable and authoritative baseline survey of forest resources to date.

3. In addition to the country-specific survey, FRA 2000 included an independent, objective pan-tropical remote sensing survey of forest cover change, a set of global maps of forest cover and ecological zones, and the development and establishment of a forestry information system. Detailed results are available at www.fao.org/forestry/fo/country/nav_world.jsp and www.fao.org/forestry/fo/fra/index.jsp on the World Wide Web. Printed versions will be available shortly.

MAJOR FINDINGS

FOREST AREA

Country information

4. The 2000 assessment concluded that the world's forest cover at the year 2000 was about 3.86 billion hectares, or about 0.6 ha per capita. Net deforestation at the global level was estimated at approximately 9 million hectares per year and gross global deforestation at approximately 13.5 million hectares per year. Net deforestation rates were highest in Africa and South America, whereas afforestation, through forest plantations, significantly offset the loss of forests in Asia. In contrast, the forest cover in industrialized countries remained essentially stable (Table 1, Appendix 2).

5. FRA 2000 was the first global assessment to effectively use a common definition for forests in industrialized and developing countries alike (Appendix 1). The result of the uniform application of the forest definition had a significant impact on the global findings for the year 2000, which, in part, resulted in an estimated forest area which was 400 million ha greater than the global figure reported for 1995 (FAO, 1997). This was particularly evident in Australia and the Russian Federation. FAO's 1995 estimate for forests in Australia was 41 million ha, compared to the 2000 estimate of 158 million ha. The dramatic increase of over 115 million hectares of forest is a consequence of the application of a 10 percent canopy cover threshold for defining forest, as opposed to the 20 percent threshold used for industrialized countries in previous assessments. Accordingly, large expanses of Australia's sparsely-stocked forests were incorporated into the year 2000 estimate. In the Russian Federation, FAO's 1995 estimate was 764 million ha, compared to the year 2000 estimate of 850 million hectares.

Table 1. Regional forest cover and forest cover change

¹⁾ n.s. means not significant. ²⁾ according to FAO (1997), included for comparison with present estimates

6. Another factor contributing to the higher estimate for the year 2000 was the inclusion of results from inventories conducted after 1990, some of which contained improved figures that were higher than those previously reported, for example in Mozambique¹. In other cases, such as in Kenya, more detailed breakdowns of forest classes in the inventory reports facilitated an improved reclassification of national results into the FRA 2000 global standards, thereby including some areas of forests, previously classified as other wooded lands.

7. The forest vegetation and the terminology used to describe them vary widely between countries, compounding the problem of aggregating national data into a uniform global estimate. More than 650 definitions of forest types were assembled from 132 developing countries (from 110 independent surveys). Re-classifying this information into a small set of global classes was a major challenge. The details on how the national definitions were converted into global classes are maintained in the Forestry Information System and are presented on the World Wide Web.

8. Deforestation is defined as the removal of the forest and its replacement by another land use class, such as shifting or permanent agriculture, ranching, mining or water (usually from impoundments). By definition, logging does not in itself result in deforestation, if the forest is allowed to regenerate. Reforestation refers to the re-establishment of forest formations after a temporary condition with a crown cover density of less than 10 percent. However, in many tropical countries, logging roads opens up new areas to farmers who take advantage of the clearings and begin to cultivate crops. Consequently the area is deforested due to the land use change from forest to agriculture. Net deforestation is the sum of all deforestation less gains due to afforestation (forest plantations and natural regeneration in previously unforested areas) (Appendix 1).

9. While the concept of deforestation has remained essentially consistent for FAO's assessments throughout the last 20 years, its actual implementation in the global survey is difficult due to the reliance on country information which was not developed for this purpose. For example, an increased reliance on remote sensing in national surveys, without adequate field sampling or ground truthing, has made the detection of actual changes in land use difficult. Instead, changes of land cover are assessed with satellite imagery, which may not necessarily coincide with actual changes in the use of the land.

10. From the aggregation of data supplied by countries to FAO, net annual deforestation rates for the period 1990 - 2000 were estimated at 9.0 million ha globally, as compared to previous estimates of 11.3 million ha (1990 - 1995 (FAO, 1997)) and 13.0 million ha (1980 - 1990 (FAO, 1995a)). While these estimates are not fully comparable due to the different methods employed in the various global assessments, as well as changes and updates in the country information base from which the estimates were derived, net deforestation has likely decreased since the 1980s at the global level.

11. Reductions in net deforestation (or gains in forest area) in both developing and industrialized countries were mainly due to significant increases in forest plantations and the succession of forests on abandoned agricultural lands. The current annual rate of plantation establishment is 4.5 million ha worldwide. For the 1990's as a whole, however, it was estimated that about 30 million ha (i.e. 3 million ha/year) of plantations were successfully established, half of which constitute reforestation of previously forested lands. There were no significant transitions from plantations into natural forests, or conversions into agricultural lands.

12. Natural regrowth of forests is occurring in many industrialized countries in areas where agriculture is no longer an economically viable land use. This is a process that has been occurring for several decades, and is particularly evident in some countries within the Commonwealth of Independent States (CIS), including the Russian Federation. The FRA 2000 pan-tropical remote sensing survey also revealed that about 1 million ha of other lands in the tropics revert to forest each year. Along with the establishment of plantations in Asia, these new forests contribute significantly to a lower net deforestation.

13. The above findings are illustrated in Table 2 that shows the broad fluxes of land use by major domain and in Table 3 where the gross and net changes in forest cover are summarized.

Table 2. Annual transitions between natural forest, plantations and other land 1990-2000 in millions of hectares per year [Mha/year], globally and for tropical / non-tropical countries.

Notes: Tropical countries include those covered by the FRA 2000 pan-tropical remote sensing survey. The estimated rate of successful plantations for the 1990's as a whole (3 million ha/year) was used.

Table 3. Gross and net change of total forest cover respective gross and net change of natural forest cover by major domain. Million hectares per year 1990-2000

Notes: Tropical countries include those covered by the FRA 2000 pan-tropical remote sensing survey. The estimated rate of successful plantations for the 1990's as a whole (3 million ha/year) was used.

Pan-tropical remote sensing survey

14. FRA 2000 included a remote sensing survey for tropical forests to assess forest change. The survey was based on sampling techniques and the use of satellite imagery and produces information at pan-tropical and regional levels. Results of the study at the pan-tropical level indicate that the world's tropical forests continue to be lost at about 8.6 million ha annually in the 1990's, compared to a loss of around 9.2 million ha during the previous decade. During the same period, the loss of closed forests dropped from 8.0 million ha in the 1980's to 7.1 million ha in the 1990's. With standard errors on these estimates at 15%, the reduction in deforestation rates between the two decades is not significant. Analysis of these results and those of specific countries is continuing, particularly on how to compare the results to the country-specific findings.

15. Stratified random sampling (10%) of the world's tropical forests was employed through 117 sample units distributed throughout the tropics^{2 3}. For each of the sample units, three Landsat satellite images from different dates provided the raw material for producing statistics on forest and other land-cover changes from the period 1980 to 1990 and from 1990 to 2000. The survey was an update of the study made for FRA 1990 (FAO, 1996), using the same sample, but observing at three (instead of two) points in time over the twenty year period 1980-2000.

16. According to the remote sensing survey, during the last decade, most of the conversion of forests were into agriculture, pastures and shifting agriculture. The rates and causes of deforestation are essentially consistent for natural forests during the two time periods (Figure 1). Important products generated through the survey include change-matrixes for the tropics as a whole and for each region. The matrices show the various forest and land-cover classes and how they change over the last two decades. This study is the first to provide a consistent methodology for assessing forest change between two assessment periods.

Figure 1. Summary of results from the Remote Sensing Survey. The figure shows the total area change for each land classification as interpreted in the survey, for each of the two ten-year periods studied. Note that although a decrease in the loss of forest can be detected, this is not statistically significant in the study.

Issues

17. Comparable time series were absent in most countries, including many industrialized ones making precise estimates of forest change difficult at both national and global levels. The lack of fully compatible multi-date information at the national level for most countries continues to be one of the biggest constraints in the global assessments. In fact, the vast majority of national inventories today are still the result of single-shot exercises.

18. Because national forest inventories are long-term endeavours, much of the country information that was used for FAO's periodic assessments was several years old by the time it was incorporated into the assessment. The bulk of the national inventories used for FRA 2000 span a period of more than ten years; the average year for inventories in developing countries and industrialized countries was 1994 and 1996 respectively. The oldest survey used was from 1981. Projecting these figures forward (and backward) to the reference years 1990 and 2000 had major implications for the reliability of the calculation of forest change and deforestation.

19. In addition to the inventory reports from the developing countries and the remote sensing survey, several initiatives were undertaken to review the state-of-the-art knowledge on deforestation including an Expert Consultation on methodologies for forest change estimations (FAO, 2001), a comprehensive review of literature on deforestation and in-depth studies for 12 countries.

20. The Expert Consultation concluded "...the process of deforestation is such a complex process, involving physical, climatic, political, and socio-economic forces which are themselves very complex, that simple generalised models of forest change have so far not been developed. Current models are oversimplified and yield similar predictions of forest cover change rates for countries which are known to be very different (P�ivinen and Gillespie, 2000)". Based on these findings, FRA 2000 discontinued using the FRA 1990 model for predicting or extrapolating forest loss based on population dynamics. The correlation between some basic national socio-economic indicators and the rate of forest cover change is shown in Table 4. The only variable that comes near significance is the proportion of rural population. Still, it only accounts for 14% (r²) of the variation in forest cover change at national level.

21. The contradiction between the emphasis on deforestation in the global forest agenda and the shortage of reliable information led FRA 2000 to conduct a separate study on the availability of forest change information for the tropics. One study (Rudel et al., 2000) indicated that the number of publications on tropical deforestation has grown at such a rapid rate over the past ten years that it has become difficult to `keep up' with it. Yet, the uncoordinated ways in which the studies are conducted and the disparate outlets in which they are published, limit the use of the information. In addition, and despite the larger number of publications, the study indicates that first hand (field) knowledge on forest change may have decreased. This represents a challenge for organizations that would like to use such data to monitor tropical forest change or to support policy initiatives.

Table 4. Correlation coefficients (r) between forest cover change rate and selected variables (n=203, data not weighted). All data at national level and from FRA 2000 global tables available at www.fao.org/forestry/fo/fra/index.jsp

VOLUME AND BIOMASS

22. Total forest volume (over bark) and above ground woody biomass were estimated for 166 countries, representing 99 percent of the world's forest area. The world totals, subject to validation, in process at the time of publication, are 500 billion m³ of wood equivalent to 350 billion tons of woody biomass. Almost one third is located in South America and 18 percent is in Brazil alone. The world-wide average standing volume is 126 m³/ha equivalent to 92 tons per ha. South America had the highest average standing volume at 172 m³/ha, North and Central America 129 m³/ha, Africa 124 m³/ha, Europe 112 m³/ha, Asia 99 m³/ha and Oceania the lowest with 73 m³/ha. South America is also the highest in terms of average biomass per ha, at 128 t/ha.

23. Many Central American and Central European countries are among the highest in terms of standing volume by hectare, the former having high-volume tropical rain forests and the latter intensively managed temperate forests.

FOREST PLANTATIONS

24. Overall findings indicated that the world's natural forests were still subject to net decreases due to deforestation, while new forest plantation areas were reported as being established globally at the rate of 4.5 million hectares per year, with Asia and South America accounting for more new plantations than the other regions. For the 1990's as a whole, it was estimated that about 3 million ha per year were successfully established (see also Tables 2 and 3). Of the estimated 187 million ha of plantations worldwide at year 2000, Asia had by far, the largest areas in forest plantation development. In terms of genera composition, Pinus (20 percent) and Eucalyptus (10 percent) remain dominant genera worldwide, although overall diversity of species planted was shown to be increasing. Industrial plantations account for 48 percent, non-industrial 26 percent and unspecified 26 percent, of the global forest plantation estate.

25. The results above were the first global estimates with a uniform definition of forest plantations and can therefore not be directly compared to previous estimates. The FRA 2000 country statistics on plantations may also differ from those reported in prior FAO publications (e.g. FAO 1995b), partly due to changes in definitions. For example, rubber tree plantations are now considered as forest plantations, whereas in past assessments they were not. Previous assessments also used general reduction factors to indicate the successful proportion of plantations remaining after establishment. In fact, the 2000 assessment applied reduction factors according to the best available data from each country independently. Another reason for the change in the FAO statistics is the change in the information base from which the estimates were derived. FRA 2000 has the most extensive and up-to-date database on plantations ever compiled. With over 802 references on plantations, the current information base on plantations surpasses all other assessments. These include data from many industrialized countries as well, none of which were included in the prior global assessment reports.

26. At the same time, information about the extent of plantations in many industrialized countries, particularly in Europe, was less clear than the situation in developing countries. Many European countries made no distinction between planted and natural forests in their inventories. Additionally, because trees have been planted over long periods of time in these areas, frequently have long rotation periods (up to 100 years) and commonly use naturally occurring species, the distinction between natural and planted stands is not readily discernible. Nevertheless, FRA 2000 identified the ten countries with the largest reported plantation development programmes (by area) as China with 24 percent of the global area; India with 18 percent; the Russian Federation and the US each with 9 percent; Japan with 6 percent; Indonesia with 5 percent; Brazil and Thailand each with 3 percent; Ukraine with 2 percent and the Islamic Republic of Iran with 1 percent. Together, these account for 80 percent of the global forest plantation area.

27. Within the same ten countries, an estimated 52 percent of forest plantations are grown for industrial purposes to supply raw material for industry; 26 percent for non-industrial uses ; and the purpose was not specified in 22 percent, particularly in the Russian Federation, Japan and the Ukraine. The industrial forest estate ownership, where specified in these ten countries was: 33 percent public; 26 percent private; and 41 percent other or unspecified. For the non-industrial estate, figures were 39 percent public; 39 percent private; and 22 percent other or unspecified. The countries with major industrial plantation areas (expressed as a percentage of national forest plantation area) included the US (100 percent); China (83 percent); and India (37 percent). These three countries account for 73 percent of all industrial forest plantations globally. The countries with major proportions of non-industrial plantation areas included Thailand, 76 percent; India, 63 percent; Indonesia, 42 percent; and China, 17 percent. These four countries accounted for 75 percent of all non-industrial forest plantations globally.

28. Industrial plantations provide the raw material for wood processing for commercial purposes, including timber for construction, panel products and furniture, and pulpwood for paper. In contrast, non-industrial plantations are aimed for example at supplying fuelwood, providing soil and water conservation, wind protection, biological diversity conservation and other non-commercial purposes. In many countries, particularly in the developing world, the end purpose of the plantations is not clearly defined at the onset. In some of these cases, valuable tree resources are established which coincidentally match future needs. However, in others the lack of planning may result in plantations which have little commercial value and a low potential for local use.

TREES OUTSIDE FORESTS

29. Trees outside forests are trees and tree environments on land not defined as forest or other wooded land (Appendix 1). Tree resources outside forests have not been systematically taken into account in forest resource assessments and are usually overlooked in policy and decision-making processes regarding sustainable natural resources management. However, the role of these resources in providing goods and services in rural and urban environments is important.

30. In countries with little forest cover, trees outside forests constitute the main source of wood and non-wood forest products. For example, in Morocco, where forests are less than 5 percent of the land and other wooded lands only 7 percent, nearly 20 percent of the land may be occupied by trees outside forests, namely as wooded pasture (84 percent of the land occupied by trees outside forests) and fruit tree plantations (12 percent). However, in many areas, the use of trees in conjunction with other land uses is often constrained by insecure access to land and trees, unfavourable and contradictory policies, non-competitive economic return on products and inadequate technology transfer.

31. On agricultural lands in densely populated areas, fruit tree plantations, home gardens and hedgerows may often cover a high proportion of the land. Communities, farmers and herders who do not have access to forests diversify their production and protect their land by maintaining various tree systems on their farms. In Kerala, one of the most densely inhabited States of India, a study in 1998 estimated that about 83 percent of the wood was from homesteads, 10 percent from estates and only about 7 percent from forest areas, when 26.6% of the State area is under forest cover. Trees outside forests met about 90 percent of the fuelwood requirements of the State.

32. Much information on trees outside forests exists, but it is site specific and scattered among different institutions and sectors and was collected to serve a wide range of purposes. The data collection method is rarely reported, complicating the interpretation of the results. Much of the available information does not relate products to resources. Because an important part of trees outside forests belongs to the informal sector, they are frequently not included in forest resource statistics. The economic value of their products is often underestimated and their environmental benefits often completely overlooked. The information related to trees in settlements and cities is almost non-existent.

33. Though statistics on trees outside forests do not exist for most countries, some countries are now addressing the theme in their new surveys. For example, in France, two complementary programmes provide accurate information on the trees outside forests. The National Forest Inventory of 1996 included special inventories for hedgerows, line plantations and isolated trees. Over the last thirty years, the methodologies and scope of inventories aimed at assessing trees outside forests have been refined and are presently part of the National Forest Inventory's mandate. Since 1981, the Teruti survey realized by the "Service central des enqu�tes et �tudes statistiques" (SCEES) of the Ministry of Agriculture monitors land use changes and includes trees outside forests. Results from these studies contributed to recognizing the important role of hedgerows in regulating annual run-off and in decreasing peak flow velocity. Today, several national and European Community policies and programmes support the restoration of the landscape with trees outside forests.

34. Many examples exist where the means are being developed to address the assessment of trees outside of forests. For instance, the Tropical Agriculture Research and Higher Education Center (CATIE), Costa Rica, in collaboration with Freiburg University (Germany), is developing a regional methodology for Central America to assess tree resources outside forests. A mix of satellite remote sensing, aerial photos and ground sampling is used to assess and monitor this complex resource at national and regional levels. Parameters monitored include the number of species, their distribution and structure. Increased knowledge and improvements in the tools needed for the assessment of this important emerging issue will allow sustainable use of the resource.

FOREST MANAGEMENT

35. Several developments around the world over the last decade have focused on achieving sustainable forest management (SFM). SFM balances environmental, socio-cultural and economic objectives of management in accordance with the Forest Principles agreed at UNCED in 1992. The SFM concept has stimulated changes in forest policy and legislation in many countries. One indicator of political commitment to the concept of sustainable forest management is the number of countries (149) currently involved in international initiatives to develop and implement criteria and indicators for sustainable forest management. On the ground, changes are occurring in management objectives and practices, and in the involvement of partners in planning and managing forests. FRA 2000 results from industrialized countries (accounting for 45 percent of the total forest area in the world, most of it in the temperate and boreal zone) indicated that 88.7 percent of these forests are being managed according to a formal or informal management plan.

36. National statistics on forest management were not available from a fairly large number of developing countries, including many of the larger countries in Africa and some key countries in Asia. Nevertheless, preliminary results from developing countries show that, out of a total forest area of 2,128 million ha, at least 123 million ha, or about 6%, were covered by a formal, nationally approved forest management plan covering a period of at least 5 years. It must be emphasised that the total area reported to be subject to a formal or informal forest management plan is not necessarily equivalent to the total area of forest under sustainable forest management. Some areas covered by a management plan may not be under sustainable forest management. Conversely, areas may be under sustainable forest management without the existence of a formal management plan.

37. One way of demonstrating that a particular forest is being managed sustainably for wood production purposes is through the act of third party certification. A number of international, regional and national forest certification schemes now exist. Depending on how the term "area certified" is defined, the total global area of certified forests may be estimated to be somewhere between 80 and 110 million hectares. Notably, most certified forests are located in a limited number of temperate countries.

38. A direct comparison with previous estimates of the forest management status in tropical and temperate/boreal regions is not possible due to differences in definitions used. However, in 1980 an estimated 42 million ha of forest in 76 tropical countries were under intensive management for wood production purposes. Looking at the situation in the same 76 countries in 2000, 117 million ha⁴ of the forests in these countries are reportedly covered by a formal, nationally approved forest management plan of a duration of more than 5 years - most, but not all, of these for wood production purposes. 2.2 million ha of forests in these countries have been certified by third parties according to standards set by the Forest Stewardship Council.

39. As regards temperate and boreal forests, the situation appears to have remained stable or has improved in the last 20 years. All areas classified as closed forests in the former USSR were reported as covered by management plans in the early 1980s and in 2000 the Russian Federation and most of the CIS states reported that all forests were being managed according to a formal or informal plan. In Europe, 19 countries provided information for forest management assessments in the early 1980s, 1990 and 2000. For these countries the percentage of closed forests "managed according to a forest management plan" in 1980 was 62.5 percent; in 1990, the percentage of forests "under active management" was 72.7 percent; and in 2000, the area reported to be "managed in accordance with a formal or informal management plan" was equivalent to 94.4 percent of the forest area. The percentage of the forest area under management in Canada and the US has increased from 60.0 and 41.4 percent respectively in 1990 to 70.9 and 55.6 percent respectively in 2000.

FOREST FIRES

40. Severe forest fires around the world gained international attention during the 1990's. Millions of hectares burned in 1997 and 1998 and smoke blanketed large regions of the Amazon Basin, Central America, Mexico and Southeast Asia, disrupting air and sea navigation and causing serious public health problems. Significant losses of forest vegetation and biomass were incurred. Ecosystems generally not subject to fires, such as the Amazon rain forest in Brazil and the cloud forest of Chiapas in Mexico, sustained considerable damage. Although on a smaller scale, the global wildfire situation in 1999-2000 was again serious. Fires were widespread in Indonesia in 1999 and 2000, but not on a scale comparable to 1997-1998. The major fires of 2000 occurred in Ethiopia, the eastern Mediterranean and the western United States.

41. Fires were used commonly to clear areas of forest for agriculture in some countries, some of which contributed to wildfires as they escaped the control of the farmers. Other countries and regions have become more susceptible to larger and more damaging fires as a result of long-term exclusion of fire from forests. Fire exclusion in National Forests in the United States has resulted in unnatural accumulations of fuel loads and thus increases in forest fires since the 1980s. Another important consideration was the large areas of degraded forests and other wooded lands that had been converted to grassland and shrubland through repeated fires, which were much more prone to burn frequently, and therefore prevented a return to tree cover.

42. Comprehensive global statistics on wildland fires required to make a reliable comparison of global fire occurrence in the 1980s and 1990s do not exist⁵. However, some general observations can be made. Both decades experienced high annual variability in regional and national fire occurrence and impacts. El Ni�o episodes, such as in 1982-1983 and 1997-1998, were the most important climatic factors affecting total area burned and fire impacts in both decades. In these years, most of tropical Asia, Africa, the Americas and Oceania experienced extreme wildfire situations. During 1997-1998, the number of land-clearing fires and other fires which escaped control increased in the equatorial forest regions of Southeast Asia and South America. The northern temperate and boreal forest zones also experienced extremely dry years in both decades. Central eastern Asia was affected most severely in 1987, particularly Siberia and northeastern China. The Russian Far East was also severely affected during the 1998 drought.

43. Policy makers are beginning to realize that continued emphasis only on emergency response will not prevent large and damaging fires in the future. Emergency preparedness and response programmes must be coupled with prevention through better land use policies and practices. Actively working toward reduced fire risk with community involvement is an important strategy to better conserve natural resources while reducing the impacts of wildfires. Between 1998 and 2000, several international initiatives related to sustainable development and wildland fire prevention, preparedness, management and response were started or continued. Many countries are now starting to develop policies and practices to improve their institutional capacity to prevent, prepare for, and combat forest fires. At the same time, it should be remembered that fire is one of the natural forces that has influenced plant communities over time and as a natural process it serves an important function in maintaining the health of certain ecosystems. Consequently, the traditional view of fire as a destructive agent requiring immediate suppression has given way to the view that fire can and should be used to meet land management goals under specific ecological conditions.

PROTECTED AREAS

44. Interest in the formal protection of forests, mainly for biological diversity conservation, has increased considerably during the past decade. Many countries have committed themselves to designate considerable portions of their natural forests, with resulting increases in the number and area of national parks and reserves. For FRA 2000, the World Conservation Union (IUCN) classification of protected areas was used (Table 5). The classification scheme has six different classes according to the type of area and the degree of protection afforded. While being the only international standard classification available, the interpretation of the concept of protected areas frequently differs substantially between countries, making the aggregation of statistics unreliable at the global level. As an example, for some countries it is argued that virtually all forests are protected according to IUCN category V as a consequence of general forestry legislation.

45. FRA 2000 relied on two independent sets of statistics, one submitted by countries in response to the FRA 2000 questionnaire and the other from a spatial database developed by the United Nations Environment Programme-World Conservation Monitoring Centre (UNEP-WCMC). The FRA 2000 questionnaire sent to countries was generally directed at forestry agencies, while UNEP-WCMC collected its information from national agencies involved in nature conservation, parks and protected areas. Comparison of these two data sources indicated that much work is still needed to harmonize national and international data, and even data coming from different agencies in the same country (Figure 2).

Table 5. IUCN categories for nature protection

Figure 2. Proportion of protected forests in industrialized countries, country-wise comparison between FRA 2000 global map data and FRA 2000 questionnaire response for industrialized countries (UN-ECE 2000)

46. Consistent global data, broken down by country, could only be established through the UNEP-WCMC spatial database, overlaid on the FRA 2000 global forest cover map. World-wide, about 10 percent of the forests are then found to be protected. Regional results were that North and Central America had 17 percent of their forests under protection; South America, 16 percent; Africa, 11 percent; Oceania, 10 percent; Asia, 9 percent; and Europe (including the Russian Federation) 5 percent.

47. The increased commitment to expanding the area of forest under protection is a positive development, but it needs to be complemented by efforts to upgrade the effectiveness of protected area management. This issue has received substantial attention in the past few years and its importance is likely to grow as pressures on forests continue to increase.

REMOVALS

48. Whether the goal of SFM is reached in production of forest, depends to a large extent on the implementation of logging practices which ensure that the potential production of goods and services of the forest are sustained over the long term. For FRA 2000, information on harvesting intensities and harvesting plans was studied in 43 tropical countries. One objective with the study was to establish a new system for the collection, analysis and dissemination of information specific to the impact of timber harvesting operations on tropical forests to identify the location and severity of non-sustainable timber harvesting practices.

49. The countries studied represented the forested tropical regions in Africa (19 countries), Asia and Oceania (10 countries) and America (14 countries). The countries surveyed had a forest cover either greater than 5 percent of the total land area, or greater than 1 million hectares. Collectively they accounted for approximately 90 percent of the world’s tropical forest resources. Information was gathered through an extensive literature review, and, in some cases, through in-country visits. The literature research focused on data published in government reports and in documentation citing in which the original source of the data could be recorded.

50. Results from the year 2000 show a high variation in harvesting intensity in the countries studied. In Africa, 3.3 million ha were harvested annually, out of the 5.9 million ha under a timber-harvesting scheme. The harvesting intensity was highly variable in the countries and varied from 1 m³/ha in Zambia to 13m³/ha in Gabon. In Asia and Oceania, the total forest area under harvesting scheme in the year 2000 was 27.3 million ha, out of which 6.2 million ha were actually harvested each year. The harvesting intensity was much higher compared to Africa and ranged from 5m³/ha in Myanmar to 23m³/ha in Vietnam. In tropical America, a total of 16.7 million ha were under a timber harvesting scheme, while 1.9 million ha were harvested annually. The harvesting intensity ranged from 1 m³/ha in Bolivia to 34m³/ha in Brazil.

51. Data from industrialized countries supplied by UN ECE to FAO on removals covered 40 countries. However, only total volume extracted per year was reported. For these countries, an estimated 1.2 billion m³ per year were harvested each year.

NON-WOOD FOREST PRODUCTS

52. For centuries, non-wood forest products (NWFP) have played an important role in the daily life and well being of millions of people worldwide, but the last decade has witnessed a steep increase in interest and activities concerning NWFP. There were numerous projects to promote NWFP use and commercialization as a means to improve the well being of rural populations and at the same time conserve existing forests. Rural and poor people in particular depend on NWFP as sources of food, fodder, medicines, gums, resins, construction material, etc.

53. Although significant advances in research on both the socio-economics and the biology of NWFP have occurred in the last few years, the assessment of NWFP and of the resources providing them is still a difficult task. The multitude and variety of NWFP, the different uses at local, national and international levels, the multiplicity of interests and disciplines involved in NWFP assessment, organizational and financial constraints, the lack of globally, or even nationally, recognized common terminology and units of measurement all contribute to lack and unreliability of information on NWFP.

54. Few NWFP are significant at global level, which is one of the reasons why they are not reported in the global forest products statistics published by FAO. However, when taken at the category level (medicinal plants, fibres, etc) they have global relevance. FRA 2000 data collection activities covered the categories of food; fodder; medicinal plants; perfumes and cosmetics; dying and tanning; utensils, handicrafts, and construction materials; ornamentals; exudates (for plant products); living animals; honey and beeswax; bushmeat; other edible animal products; hides and skins; medicines; colorants; and other non-edible animal products. There is an urgent need to collect statistics on bamboo and rattan as major NWFP. For the industrialized temperate and boreal countries, data on quantities and monetary values are available on Christmas trees, cork, mushrooms, truffles and berries, medicinal plants and decorative foliage, game meat, hides and pelts, honey and nuts were also reported.

55. Traded NWFP contributed to the fulfilment of daily needs and provided employment and income, particularly for rural people and especially women. Internationally traded NWFP, such as bamboo, rattan, cork, gum arabic, aromatic oils and medicinal plants, can achieve high prices in comparison with NWFP traded on national markets and thus contributed to national economic development. However, the majority of NWFP were used for subsistence and in support of small-scale, household-based enterprises. Despite their real and potential importance, no regular monitoring of the resources and evaluation of the socio-economic contribution of NWFP was being carried out by national institutions.

56. Between the multitude of locally traded NWFP (with low extraction volumes, low impact on the resources, and lesser monitoring needs) and the few products which have reached such economic significance to be exploited as commercial crops outside their original natural areas, (rubber and palm oil are two notable examples), are a number of NWFP of commercial significance, the exploitation of which is still largely based on wild sources and forest ecosystems. These forest products and their source species need monitoring and management in the forestry sector to ensure sustainable management and to provide information for policy development.

57. Rattan, for example, is one of the most important commercial non-wood forest products in Asia. Over 700 million people worldwide traded or used rattan for a variety of purposes. Domestic trade and subsistence use of rattan and rattan products were valued at an estimated US$ 3 billion per annum, and another US$4 billion were generated through international trade, according to estimations made by the International Rattan and Bamboo Network (INBAR). In some producer countries, rattan resources are today in short supply. Consequently, rattan processing industries are in decline in these areas with a negative impact on the economy, particularly in rural areas.

58. Data collected for the FRA 2000 NWFP component demonstrated a significant lack of quantitative information at the national level on both the products and the resources. Information was scarce and often mixed with agricultural production statistics. Statistical data, if available at all, was limited to selected internationally traded NWFP and to export quantities. Information on the resource base and on the subsistence use of NWFP is non-existent; this was mainly due to the multitude of products that are used by local people and to the costs of measuring and reporting on subsistence products. For these reasons, data collected were presented as national results in FRA 2000, aggregated to regional and subregional, but not global, levels.

GLOBAL MAPS

59. New global forest and ecological maps provide spatial representation to area statistics of the survey findings from the individual countries and regions, providing a synoptic view of worldwide forest cover. The global ecological zoning map provides an important means of aggregating global information on forests or other natural resources according to their ecological type. The maps can be used in conjunction to show and analyse worldwide forest cover by ecological type or zone (Table 6). The forest cover map was developed using coarse resolution satellite imagery (AVHRR) (Figure 3), while the ecological map was developed from existing national and regional potential vegetation maps. In previous global assessments, the means and technology did not exist to produce a global map based on satellite imagery.

60. Each map is generated from a corresponding computerised Geographical Information System (GIS) database, which makes it possible to combine them with different spatial and statistical data, permitting new perspectives on the world’s forests. Digital versions of the maps that can be used for further analyses are publicly available through the FAO website at www.fao.org/forestry/fo/fra/index.jsp .

Table 6. Total forest area by ecological zone and distribution between regions, according to FRA 2000 global ecological zoning and global forest cover map. Distribution of percentages do not exactly tally with other area statistics due to systematic distortions in the remote sensing classification of forests in the global map.

Figure 3. FRA 2000 Global Forest map produced during FRA 2000.

OTHER ASSESSMENT PARAMETERS

61. Special studies on biological diversity and other components were completed in the year 2000 and are undergoing final data analysis and compilation for reporting.

_______________________________

1  The upward revision of forest cover for 1990 due to more recent and improved national assessments in developing countries is consistent with findings reported in the interim 1995 assessment presented in the State of the World’s Forests 1997.

2  The Kotka III meeting advised FAO to consider conducting the remote sensing survey at the global level with about 350 sample units. However, due to financial restrictions, no work was conducted outside the tropics (aside from some pilot activities).

3  Due to high rates of cloud cover in two sample units in year 2000 imagery, and a missing image for the 1990 sample in another unit, the number of samples used for FRA 2000 was reduced to 113.

4  National data missing from some countries.

5  FRA 2000 compiled country reports on forest fire statistics for countries with national reporting systems, and developed profiles on the ecology and management of forest fires for specific countries.