3. COMPATIBILITY OF THE DATA SETS: REASONS FOR CHANGES IN FIGURES

3.1. CHANGES IN GEOGRAPHICAL BOUNDARIES
3.2. CHANGES IN DATA SET BREAK-DOWN
3.3. CHANGE IN DEFINITION
3.4. REPORTING THE NET AND THE GROSS AREA PLANTED
3.5. REPORTED AND EXTRAPOLATED DATA

Although all the data have been published by FAO, most of them are not fully compatible or directly comparable for different reasons. The most important reasons are summarised below.

3.1. CHANGES IN GEOGRAPHICAL BOUNDARIES

During the past 20 years country boundaries have changed. Since most of the data sets are based on country level, the boundary changes also affect the data sets. Below are selected examples that have resulted from country boundary changes.

Table 3: Changes in Country Boundaries

FRA 1980	Interim 88	FRA 1990	SOFO 97	SOFO 99	SOFO2001	FRA2000
Ethiopia PDR	Ethiopia PDR	Ethiopia PDR	Ethiopia Eritrea	Ethiopia Eritrea	Ethiopia Eritrea	Ethiopia Eritrea
Morocco	Morocco	Morocco	Morocco	Morocco	Morocco Western Sahara	Morocco Western Sahara
Indonesia	Indonesia	Indonesia	Indonesia	Indonesia	Indonesia East Timor	Indonesia East Timor

3.2. CHANGES IN DATA SET BREAK-DOWN

3.2.1. Ownership
3.2.2. Industrial and Non Industrial Purpose
3.2.3. Species

As previously mentioned the four data sets provide different breakdown and statistics for total forests, at global and regional levels, and only some reports provide a more detailed breakdown of information on forest plantations at country level.

3.2.1. Ownership

FRA 2000 is the only assessment that provides breakdown on ownership at country level for developing countries. A comparison of data sets, for developing countries, based on ownership is therefore not possible.

3.2.2. Industrial and Non Industrial Purpose

FRA 1980 and FRA 2000 provide a breakdown by purpose (industrial and non-industrial) at country level. Similar breakdown for industrialised countries was not done.

3.2.3. Species

Unlike earlier assessments, FRA 2000 provides, for developing countries (not for the developed reported by UN-ECE), a breakdown based on the main species group composition at country and global level. FRA 1980 provides a rudimentary species breakdown according to Hardwood, Hardwood Fast Growing, and Softwood classes at the country level only FRA 1980 provides plantation data by Hardwood and Softwood classes at country and regional levels

3.3. CHANGE IN DEFINITION

Following expert advice, definitions have changed during the years (Appendix 1) although the main body of the forest plantation definition has remained the same. In summary all definitions refer to plantation as:

• forest stands established artificially by afforestation on land which previously did not carry forest;
• forest stands established artificially by reforestation on land which carried forest for certain time before (e.g. 10 - 50 years).

FRA 2000 has further gone ahead in the definitions of plantations (Appendix 1) and has better described the concept of afforestation as: [...] the conversion from other land uses into forest, or the increase of the canopy cover to above the 10% threshold [...]. The concept of afforestation also [...] includes areas that are actively converted from other land uses into forest through silvicultural measures [...] and implies that the conversion process should be long-term. The concept “long-term” is central in the definition and is defined as ten years.

Finally FRA 2000 introduced in definitions of forest plantation also Hevea brasiliensis (rubber) and Quercus suber (cork oak) stands.

3.4. REPORTING THE NET AND THE GROSS AREA PLANTED

The adoption of the net area planted from part of the FRA 1990 is a cause of confusion when trying to compare data (Appendix 1 and Appendix 3). FRA 1980 and FRA 2000 reported on gross area planted. Although only the “FRA 1990 - global synthesis” reported net area for plantations it also influenced the figures of SOFO 1995, 1997 and 1999. The three SOFOs were based on the FRA 1990 data set and even if they do not directly report on forest plantations, they provide figures on total forest area as the sum of natural forest and net plantation area.

For the same reasons given above (see paragraphs 2.2.8, 2.2.12, 2.2.14 and 2.3) the FRA 2000 data set can not easily be compared with “Tropical Forest Plantation Areas 1995” data set.

The difference between Gross and Net Area Planted is the most important cause of differences between data sets and it must always be taken into consideration when trying to compare figures. The FRA 1990 data set applied an average 0.7 reduction factor. In the “Tropical Forest Plantation Areas 1995” data set reduction factors ranged from 1.0 to less than 0.3 at country level.

3.5. REPORTED AND EXTRAPOLATED DATA

When insufficient country based inventory or resource data is reported, data has to be extrapolated from a known data point. FRA does the assessment of plantation from a known reference year and applies data from an annual rate. FRA 1990 also used the same method to assess changes in the period 1990 - 80. Once obtained the assessment of forest plantation at year 1990 the annual rate of plantation establishment multiplied by 10, was applied backwards to assess 1980.

FRA 1980 also provided the status and change for forest plantations at year 1985 by using the same extrapolation method.

The result is that FRA has now different figures for a given year that have been obtained with different methods and that, therefore, belong to different data set. In particular two data sets provide forest plantation figures at year 1980 that can not be compared.

Table 4: Data sets and Available Figures Classified by the Methodology Applied to Obtain Them

Data set	Status	Extrapolated
FRA 1980	1980	1985
FRA 1990	1990	1980
Tropical Forest Plantation Areas 1995	1995	-------------
FRA 2000	2000	-------------