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.