The preceding analysis has described the trends and current status of forest plantation development around the world and examined a number of economic and political variables that affect the forest plantation sector. The analysis reveals that forest plantation development is often the result of a mixture of complex economic, environmental, political and, occasionally, philosophical decisions. However, a very important point that should also be made, is that these decisions are not taken in a vacuum. Rather, they are taken within a broader and more complex framework of economic, social and environmental aspirations, which are often expressed and articulated outside of the forestry sector. These broader aspirations are often difficult to identify, measure and analyse, but they are likely to influence forest plantation development far more than simple considerations of future wood supply.
Forest plantation development is part of a complex adaptive evolutionary approach to meeting future demands for both wood and non-wood products and for a range of environmental and social services that forests can provide. In some instances (for example in New Zealand), policymakers have tried to meet these demands by rigorously differentiating the functions of natural forests and forest plantations. In most countries, however, the concept of multiple-use management has been adopted, in recognition of the non-market benefits that forest plantations can provide and in recognition of the fact that natural forests probably can continue to play a role in future wood supply as long as they are managed sustainably.
In the context of this study however, two questions are of most importance to the outlook for roundwood supply from forest plantations:
1. what is the likely future supply of roundwood from forest plantations under current policies; and
2. what are the choices available to policymakers and how are these likely to influence future supply?
The final section of this paper is, therefore, devoted to quantitatively modelling future roundwood supply potential from forest plantations, and then to qualitatively appraising the scenarios and the impacts of particular policy choices.
Three scenarios for future potential roundwood supply from forest plantations have been modelled in this study. This modelling is not based on a detailed analysis of national forestry policies or statements38 or on an analysis of the economics of alternative forest plantation development strategies. Rather, the intention is to present a broad range of possible future outcomes, then to examine some of the forces that may affect the future course that is eventually chosen. This includes an examination of the ways in which future changes might be implemented and a discussion of how effective such changes would be in meeting future forestry policy objectives.
Since the objective is to examine the future from the perspective of forestry policy, rather than to attempt to model silvicultural or technological dimensions of forest plantation yields, only the rate of new planting varies between the scenarios. In other words, the modelling is designed to show the impact on future roundwood supply of different rates of forest plantation establishment. Other potentially important variables (such as changes in mortality rates39) are varied over time within each scenario, but do not vary between scenarios.
The three alternative forest plantation establishment scenarios considered in this analysis are:
1. no new forest plantation establishment, but replanting of all existing areas after harvesting;
2. new planting at a fixed annual rate equal to one percent of the current forest plantation area (plus replanting of all existing areas after harvesting); and
3. new planting at the estimated current rate of new planting for 10 years, with this rate reduced by 20 percentage points at 10 yearly intervals40 (plus replanting of all existing areas after harvesting).
The projected levels of potential41 future roundwood supply from forest plantations under each of these three scenarios are presented below for the period to 2050. (A detailed description of the modelling methodology is presented in Appendix 2). As earlier, these projections make the simple assumption that industrial forest plantations will be used in the future for the production of industrial roundwood, while non-industrial forest plantations will be used for the production of wood fuel.
The first scenario can be considered a baseline scenario, under which there is no further new forest plantation establishment in the future. This scenario does, however, assume that all harvested areas will be replanted with the same species and for the same purpose (i.e. industrial roundwood or wood fuel production) in the year that they are harvested. Thus, the total area of forest plantations remains constant in each country throughout the projection period (to 2050). Variations in projected potential roundwood production over time are, consequently, the results of changes in the forest plantation age-class distributions within countries and for different species (determined, in turn, by the existing age class structures and different rotation lengths used for each species).
Figure 22 shows the projection of future potential industrial roundwood production to 2050 under Scenario 1, by year and by geographical region. In this scenario (as in all of the others as well) the projection starts from the current estimated level of potential production of 331 million cubic metres of industrial roundwood per year (equal to about 22 percent of global industrial roundwood production in 1995).
Figure 22 Projected potential industrial roundwood production (1995 - 2050) under Scenario 1
The large proportion of young, immature forest plantations in the current plantation age-class structure is evident from the significant increases in potential production shown in this figure. This increase is due to these areas reaching maturity. A peak in the projection of potential production occurs in 2020 at approximately 670 million cubic metres per year. A second, higher peak, occurs in 2045 at around 710 million cubic metres. This occurs due to the projected coincidence of long (40+ years) rotation species reaching maturity at the same time as a second rotation of medium (20-40 years) rotation species reaches maturity. Thereafter, under this scenario, it might be expected that industrial roundwood production from forest plantations would continue to fluctuate around a level of about 600 million cubic metres per year in the long-run.
One of the most notable features of this projection is the impact of potential production in Asia on the overall projection of potential production. In 1995, industrial forest plantations in Asia are estimated to have the potential to produce about 60 million cubic metres per year (or 18 percent of the total estimated potential production). By 2045, potential production in Asia is projected to increase to 290 million cubic metres (or 40 percent of the total).
A large part of this projected increase is attributable to forest plantations in China, which has a 7 percent share of total potential production from forest plantations in 1995, projected to rise to 25 percent by 2050. Of the other regions, only South America is projected to increase its share of the global total, with a small increase from 11 percent in 1995 to just over 12 percent by 2050. Due to the age-structure of forest plantations in countries of the former-USSR, these countries could increase production from forest plantations in the medium-term (2010-2025) but, without new planting, production potential is expected to decline somewhat after this period.
The total area of non-industrial forest plantations in the World is considerably smaller than the area of industrial forest plantations. Consequently, despite the generally shorter rotations used in non-industrial forest plantations, the projected level of production from these plantations is much lower. In 1995, the level of potential wood fuel production from non-industrial forest plantations was estimated to be about 86 million cubic metres. This figure is roughly equal to 5 percent of total estimated wood fuel consumption in that year.
Figure 23 shows the projection of future potential wood fuel production to 2050 under Scenario 1, by year and by geographical region. It is assumed that the countries of Europe and the former-USSR only use their forest plantations for industrial purposes and potential production in Oceania is very small, so they do not appear in this figure. As the figure shows, potential wood fuel production is projected to increase rapidly from 1995 to 2005, by which time it will have almost doubled. After this, potential production will increase much more gradually to reach a peak in 2045. This unusual change in growth in the projection is mainly the result of recent large-scale non-industrial forest plantation establishment programmes in India and China. In 2005, the projected level of potential wood fuel production from non-industrial forest plantations is 150 million cubic metres per year and the peak in 2045 is projected to reach 185 million cubic metres per year.
Figure 23 Projected potential wood fuel production (1995 - 2050) under Scenario 1
The dominance of Asia in the global total potential production of wood fuel from non-industrial forest plantations is even more marked than in the case of industrial forest plantations. In 1995, Asia's share was estimated to be around 60 percent of the total and, by 2045, this share is estimated to increase to 75 percent. The majority of potential wood fuel production in Asia can be found in two countries: China and, in particular, India.
A second "medium growth" scenario was modelled under the assumption that forest plantation areas will increase each year by an amount equal to one percent of the forest plantation area in 1995.43 Again, this scenario also assumes that all harvested areas will be replanted with the same species and for the same purpose (i.e. industrial roundwood or wood fuel production) in the year that they are harvested. It is also assumed that the species used and the purpose of newly planted areas will be in proportion to the current species mix and use, such that these variables do not change relative to each other in the future. Projected potential production under this scenario starts at the same level as in Scenario 1, but increases to a higher level due to the expansion of forest plantation areas managed on short-rotations and, later on, as other newly planted areas reach maturity.
Figure 24 shows the projection of future potential industrial roundwood production to 2050 under Scenario 2, by year and by geographical region. Comparing the projection under this scenario with the projection under Scenario 1, there is little difference between the two projections of future potential industrial roundwood production until 2015. For example, as shown in Figure 24, projected potential industrial roundwood production under Scenario 2 is around 670 million cubic metres per year by 2015, compared with a projection of 645 million cubic metres per year under Scenario 1.
Figure 24 Projected potential industrial roundwood production (1995 - 2050) under Scenario 2
After 2015 however, the two projections diverge, with the projection under Scenario 2 increasing much more rapidly to reach a peak of just over 900 million cubic metres per year in 2045 (and with a much smaller decline thereafter). In fact, the peak in the potential level of industrial roundwood production from forest plantations under this scenario is almost one-third higher than the projected peak under Scenario 1.
Under this scenario, forest plantations in Asia will again increase their share of the total potential production from industrial forest plantations to about 40 percent. However, there may be some subtle shifts in the share held by each country within each region as potential production will increase most rapidly in countries where trees grow fastest.
Compared with Scenario 1, potential production of industrial roundwood is projected to almost treble rather than slightly more than double. Consequently, under Scenario 2, potential production from industrial forest plantations in all of the seven regions is expected to increase significantly over the projection period. In Asia, South America and the countries of the former USSR, potential production is projected to more than treble between 1995 and 2050, while for North America and Oceania potential production is expected to almost double. Potential production in European industrial forest plantations is expected to increase by 75 percent. Africa has the lowest projected increase in potential production, with an increase of only 50 percent. This relatively low level of increase is mainly because industrial forest plantations in South Africa (by far the largest resource on the continent) already currently contain a high proportion of mature age-classes, so they do not benefit from the maturation of young plantations as some of the other regions do.
Figure 25 Projected potential wood fuel production (1995 - 2050) under Scenario 2
Figure 25 shows the projection of future potential wood fuel production to 2050 under Scenario 2, by year and by geographical region. Because of the generally shorter rotations used in non-industrial forest plantations, the projection under Scenario 2 diverges from the projection given under Scenario 1 much more quickly, starting in around 2005. By 2050, the potential level of wood fuel production from non-industrial forest plantations is projected to reach almost 250 million cubic metres per year (over 40 percent above the projection for this year under Scenario 1).
Again, the dominance of non-industrial forest plantations in Asia is shown in Figure 25. By 2050, Asia will account for 76 percent of the total potential production from non-industrial forest plantations under Scenario 2. South America will account for 14 percent and Africa for 8 percent.
The third scenario used for the modelling in this analysis contained the highest assumption about future rates of new planting. Under this scenario, annual rates of new planting in tropical and subtropical countries were taken from Pandey (1997)44 and current new planting rates were estimated for temperate countries. These rates of new planting were then used for the first ten years of the scenario (i.e. 1995 - 2004). For the next ten years (2005 - 2014), these rates were reduced by 20 percent (i.e. to 80 percent of the current rate). The same reduction (in absolute terms - i.e. number of hectares rather than a compounded percentage) was then applied to both of the following two ten-year periods (i.e. 2015 - 2024 and 2025 - 2034) and for the final 16 years of the projection period (2035 - 2050). Thus, in the final 16 years of the projection period, it was assumed that the annual rate of new planting in each country will have fallen to 20 percent of the current new planting rate.
Again, this scenario also assumes that all harvested areas will be replanted with the same species and for the same purpose (i.e. industrial roundwood or wood fuel production) in the year that they are harvested. It is also assumed that the species used and the purpose of newly planted areas will be in proportion to the current species mix and use, such that these variables do not change relative to each other in the future.
In most countries, current new planting rates are much higher than one percent of the current total area of forest plantations (i.e. the new planting assumption used in Scenario 2). Consequently, Scenario 3 has, by far, the greatest rate of growth in potential production from industrial forest plantations and reaches the highest projected potential production level of all three scenarios.
Under the new planting assumption made in Scenario 3, the projection for potential industrial roundwood production begins to diverge significantly from the previous two scenarios as early as 2005. By 2015, the projection under Scenario 3 is 23 percent above the projection under Scenario 1 and 18 percent above the projection under Scenario 2. Potential production in 2050 reaches 1,500 million cubic metres per year (see Figure 26) or about four and a half times the estimated level of potential production in 1995. Furthermore, the projection would continue to increase beyond this point due to the maturation of industrial forest plantations planted in the latter part of the projection period.
Figure 26 Projected potential industrial roundwood production (1995 - 2050) under Scenario 3
The current area of non-industrial forest plantations is much smaller than the area of industrial forest plantations so, under Scenario 3, the rate of expansion of non-industrial forest plantations is also comparatively slow. However, compared with the past, the current rate of new planting of non-industrial forest plantations is relatively high, particularly in India. Consequently, under Scenario 3, projected wood fuel production potential increases significantly over the projection period. In fact, the projected level of wood fuel production potential in 2050 under this senario, 487 million cubic metres, is 5.6 times larger than estimated production potential in 1995 (see Figure 27 below). In comparison with the other two scenarios, this scenario produces a projection for 2050 that is 97 percent higher than under Scenario 2 and 2.8 times that of Scenario 1.
Asia has by far the largest share of projected wood fuel production potential from non-industrial forest plantations throughout Scenario 3. By 2050, Asia is projected to account for 80 percent of production potential across the five geographical regions (i.e. excluding Europe and countries of the former USSR). India and China will continue to contain by far the largest areas of non-industrial forest plantations and are projected to account for 67 percent of the total projected wood fuel production potential in 2050. Wood fuel production potential in India is projected to increase from 30 million cubic metres (in 1995) to 226 million cubic metres under Scenario 3. This would require the area of non-industrial forest plantations to increase from a current 8.3 million hectares to 24 million hectares by 2050. Wood fuel production potential in China is projected to increase from a (probably very low) estimate of 5 million cubic metres in 1995 to 100 million cubic metres in 2050.
Figure 27 Projected potential wood fuel production (1995 - 2050) under Scenario 3
The three scenarios presented above describe a range of possible futures that outline the likely boundaries for future forest plantation establishment and potential roundwood production form forest plantations. The scenarios are, however, based on relatively simple extrapolations of forest plantation establishment in the past rather than any attempt to model the almost infinite number of policy and economic choices that could affect countries plans for forest plantations in the future. In other words, the scenarios have been presented to show the likely range of possible futures rather than to prescribe any "most likely" outcome.
However, within the modelling framework used to produce the scenarios, it is worthwhile to examine some of the factors that are most likely to influence future developments. In doing so, it should be noted that future developments will be physically constrained by the inputs in shortest supply. In some countries it is likely that physical inputs such as: land; capital; water; length of growing seasons; and fertility, will limit the potential for future forest plantation expansion. In others, economic factors may limit development (i.e. investors in forest plantations will need to earn a satisfactory rate of return on their investment and this may, perhaps, be affected by the ability and willingness of governments to provide financial incentives to encourage forest plantation development).
The remainder of this section discusses the projections of future production potential from industrial forest plantations in the overall context of future global industrial roundwood supply and demand.45 It then goes on to examine some of the factors that are most likely to constrain forest plantation development in the future.
Comprehensive modelling of future production, consumption and trade in forest products has already been completed as a core component of the Global Forest Products Outlook Study (GFPOS). The modelling has been carried out independently of this study, using the Global Forest Products Model (GFPM), which is a projection model based on a price endogenous linear programming system (for further details, see: Tomberlin et al, 1999).
Because of the increased uncertainty associated with making long-term economic forecasts, the published projections only extend to the year 2010 (Zhu et al, 1998). However, for comparison with the results of this study, it would be more illuminating to have a longer projection of future world supply and demand. Therefore, the projection of total global industrial roundwood consumption (see footnote) has been extended to 2050 using simple extrapolations (see Box 12). It must be borne in mind that the uncertainty surrounding these extrapolations is quite high.
Box 12 Extending the GFPM industrial roundwood consumption projection to 2050
In order to extend the industrial roundwood consumption projection produced by the GFPM from 2010 to 2050, three alternative extrapolations have been used.
Extrapolation 1: is based on the projected growth rate in industrial roundwood consumption produced by the GFPM over the period 2005 - 2010 (1.27 percent per annum). By simply increasing consumption at this rate to the year 2050, this gives a projection of global industrial roundwood consumption of 3.1 billion cubic metres in 2050. This is the highest of the three extrapolations and is thought to be the least likely to occur in the future.
Extrapolation 2: takes the average growth rate in actual industrial roundwood consumption for the period 1961 - 1998 (1.1 percent per annum) and increases consumption at this rate from 2010 to the year 2050. This results in a projection of 2.9 billion cubic metres of industrial roundwood consumption (globally) by 2050. This extrapolation is also believed to be probably too high.
Extrapolation 3: is also based on the growth rate in actual industrial roundwood consumption for the period 1961 - 1998, but takes into account that annual growth has declined by about 0.03 percentage points each year. (This was confirmed by fitting a simple OLS regression line through the data). Projecting forwards on this basis, this extrapolation suggests that growth will fall to zero by 2050, at which point global industrial roundwood consumption will be around 2.34 billion cubic metres. This is probably the most realistic of the three extrapolations.
Figure 28 A comparison between the three extrapolations of future industrial roundwood consumption and the projected levels of potential industrial roundwood production from industrial forest plantations under each of the three scenarios about future forest plantation development
Figure 28 compares the three extrapolations of future industrial roundwood consumption with the different projected levels of future potential industrial roundwood production from industrial forest plantations, under each of the three scenarios about future forest plantation development. The figure demonstrates a few points concerning the likely contribution of industrial forest plantations to future industrial roundwood production.
Until 2010, potential industrial roundwood production from industrial forest plantations increases by about the same amount (in volume terms) as projected total future consumption. This occurs whatever scenario or extrapolation is chosen. It implies that existing forest plantations have the potential to meet the projected increase in demand for industrial roundwood (at the broad level) in the near-term. However, this increase is unlikely to be sufficient enough to substitute (in aggregate) for production from the natural forest in any significant way.
Beyond 2010, the contribution that industrial forest plantations will make is dependent on the extent to which consumption continues to increase and on future levels of forest plantation expansion.
If future industrial roundwood consumption continues to grow at a high rate (i.e. like in either of the first two extrapolations) then potential industrial roundwood production from industrial forest plantations would only keep up with the growth in consumption if large additional areas of forest plantations were established (i.e. as in under scenario 3). The other two forest plantation scenarios would fail to keep pace with the growth in consumption (i.e. they would increase their share of total production, but volume of industrial roundwood required from other sources would also have to increase). In other words, if either of these higher extrapolations turned-out to be true, then the area of forest plantations would have to increase quite dramatically to fully satisfy projected additional demand. Alternatively, production from the natural forest would have to increase to meet demand (probably unlikely) or other ways of improving supply would have to be found (e.g. efficiency gains, new or additional sources of fibre, or intensification of management46).
If, however, future industrial roundwood consumption grows less rapidly, for example at rates similar to Extrapolation 3, then industrial forest plantations are likely to play a much greater role in future industrial roundwood production. Under Extrapolation 3, Scenario 1 would result in forest plantations having the potential to meet 30 percent of industrial roundwood demand in 2050. This is slightly higher than at present, but not a major improvement over the current situation. Furthermore, it would also require production from other sources to increase in the future, because the growth in potential production (in volume terms) would not be sufficient to meet the total growth in consumption.
Under Scenario 2, forest plantations could meet 37 percent of industrial roundwood demand by 2050. This would be a significantly higher percentage than at present and would also mean that production from forest plantations could keep up with increases in consumption (in volume terms). In other words, it would not be necessary to look to alternative ways of meeting the growth in demand (although it still would not probably reduce pressures on the natural forest for industrial roundwood production).
The combination of Extrapolation 3 and Scenario 3 gives the most dramatic results of all. If growth in consumption is relatively modest (i.e. like Extrapolation 3), but forest plantation expansion is high (Scenario 3), the projected potential industrial roundwood production from industrial forest plantations (1,500 million cubic metres in 2050) could account for 64 percent of total consumption by 2050. This would represent both a three-fold increase in the share of total consumption that might come from industrial forest plantations and an increase, in volume terms, that is sufficient to significantly reduce the pressure to obtain industrial roundwood supplies from elsewhere.
Table 14 shows the share of projected total industrial roundwood production/consumption that might be accounted for by production from forest plantations in the future under all of the combinations of the three extrapolations and scenarios.
Table 14 Projected future potential industrial roundwood production from forest plantations as a percentage of total production/consumption
Forest |
Current |
Alternative extrapolations of production/consumption growth | ||||||||
plantation scenario |
estimate |
Extrapolation 1 |
Extrapolation 2 |
Extrapolation 3 | ||||||
(1995) |
2010 |
2020 |
2050 |
2010 |
2020 |
2050 |
2010 |
2020 |
2050 | |
Scenario 1 |
22.2 |
30.6 |
31.5 |
19.7 |
30.6 |
32.1 |
21.1 |
30.6 |
32.5 |
29.6 |
Scenario 2 |
22.2 |
31.2 |
34.1 |
28.0 |
31.2 |
34.7 |
29.9 |
31.2 |
35.1 |
37.0 |
Scenario 3 |
22.2 |
34.1 |
45.1 |
48.4 |
34.1 |
45.9 |
51.7 |
34.1 |
46.5 |
64.0 |
Comparisons with the current (1995) level of total industrial roundwood production are also interesting. In 1995, global industrial roundwood production totalled 1,482 million cubic metres. Thus, by 2050, potential industrial roundwood production from forest plantations under Scenario 3 is projected to rise to a level that is approximately equal to current global industrial roundwood production. However, there are significant regional variations.
In three regions, Asia, Oceania and South America, projected potential industrial roundwood production from forest plantations under Scenario 3 would exceed current total industrial roundwood production by 2050. In Asia, potential production would amount to 2.8 times current production, in Oceania the figure would be 2.2 times current production and in South America the figure would be 1.3 times current production. Interestingly, even under Scenario 1, projected potential industrial roundwood production from forest plantations in Asia and Oceania would exceed current total industrial roundwood production by 2050.
Conversely, projected potential industrial roundwood production from forest plantations in the other four regions is expected to remain significantly lower than current industrial roundwood production. In Africa, potential production under Scenario 3 would reach 75 percent of current production by 2050. In Europe, the figure would be 36 percent, countries of the former USSR would reach 40 percent and North and Central America would reach 59 percent. The reasons that the share of total production coming from forest plantations is unlikely to increase significantly in these regions are that production from natural and semi-natural forests currently dominates production in these regions and that projected potential industrial roundwood production from forest plantations will grow only relatively slowly there.
Scarcity of suitable land for new planting is probably the most common physical constraint on forest plantation development. This scarcity arises both because the physical terrain of large parts of the remaining available land is unsuitable for forest plantation development (e.g. due to factors such as altitude, slope, fertility, salinity, water table and aridity) or, more frequently, because large parts of the remaining available land is more valuable in alternative uses such as agriculture, urban development or industry. Institutional and policy restrictions may also play a crucial role.
Competition for land is greatest in developing countries with high population densities. Countries with few technological resources and high demand for agricultural and urban land tend to place a relatively low priority on forest land and resources. For example, in 1990, the five developing countries/territories with the highest population densities were: Bangladesh, Bahrain, Puerto Rico, Rwanda and India (FAO, 1995a). Of these, only Puerto Rico (with 37 percent forest cover) had an above average level of forest cover (the global average is 27 percent). Systems of land tenure are also a problem in some countries, particularly in terms of developing a large-scale industrial forest plantation resource.
Both scenarios 2 and 3 imply that forest plantation areas will expand. In particular, Scenario 3 requires a significant expansion in forest plantation areas. For example, the industrial forest plantation area in China would have to increase from 17.5 million hectares to 68.3 million hectares (equal to 7.3 percent of the total land area of China or around half of the country's current total forest area). This would also be equal to an annual average rate of new planting of 918 000 hectares per annum compared with the estimated current rate of new planting of 1.64 million hectares. The total area of industrial forest plantations required under this scenario is substantially greater than the 40.35 million hectares currently planned in China (forest plantations of all types), but is not substantially out of line with the total area of land believed to be available for afforestation (63 million hectares - Shi et al (1997)).
Table 15 The total area of industrial forest plantations required under each scenario
Country or region |
Current area (1995) |
Projected area in 2050 required under each scenario (in million ha) | ||
Scenario 1 |
Scenario 2 |
Scenario 3 | ||
North and Central America |
18.9 |
18.9 |
29.3 |
43.2 |
United States |
18.4 |
18.4 |
28.5 |
41.2 |
South America |
5.4 |
5.4 |
8.4 |
13.6 |
Asia |
41.8 |
41.8 |
64.8 |
119.5 |
China |
17.5 |
17.5 |
27.1 |
68.3 |
India |
4.1 |
4.1 |
6.4 |
11.7 |
Japan |
10.7 |
10.7 |
16.6 |
12.4 |
Oceania |
2.7 |
2.7 |
4.2 |
5.7 |
Africa |
3.6 |
3.6 |
5.6 |
8.9 |
Europe |
8.7 |
8.7 |
13.5 |
15.3 |
Countries of the former USSR |
22.2 |
22.2 |
34.4 |
28.0 |
Russian Federation |
17.1 |
17.1 |
26.5 |
21.1 |
Total |
103.3 |
103.3 |
160.2 |
234.2 |
Note: in many countries industrial forest plantation areas are currently increasing at much more than 1% per annum and Scenario 3 results in a higher total forest plantation area in 2050 than Scenario 2. In a few cases, this is not the case however, and Scenario 3 results in less new planting (e.g. Japan and Russian Federation shown above).
Table 15 shows the area of industrial forest plantations that would be required to meet the potential production projections made under each of the forest plantation development scenarios. As already explained, Scenario 1 is based on no increase from the current area. Scenario 2 requires only a relatively modest, and seemingly plausible, increase in industrial forest plantation areas. For example, the 27.1 million hectares that would be required in China by 2050 is markedly less than the 40.35 million hectares noted above. On average, Scenario 2 requires a 55% increase in the area of industrial forest plantations, but expansion is unlikely to be uniform across countries. Countries such as Chile and New Zealand have, for example, achieved isolated increases in forest plantation areas of 5-10 percent in a single year. Other countries have gone for extended periods with little or no new forest plantation establishment. South Africa, for example, is not encouraging further afforestation because of water scarcity. Conversely, there are plans in Australia to develop a 3 million hectare forest plantation estate (a trebling of the current area) by 2020.
The areas required to meet the potential production projections made under Scenario 3 also seem to be generally achievable in physical terms. However, institutional and policy constraints may play a significant role in limiting new planting to a level below the required levels. Two notable cases are China and the United States of America, both of which would be required to maintain rates of new forest plantation establishment higher (or for longer) than seems likely at present. Some of this additional required new planting could, however, be spread across other countries without markedly affecting the projections under Scenario 3.
For countries establishing forest plantations with medium and long rotation species, forest plantations established after 2025 will have little or no effect on projected potential production to 2050. Consequently, a global industrial forest plantation estate of 234 million hectares would have the potential to produce rather more than the 1,526 million cubic metres of industrial roundwood projected for 2050 under Scenario 3. It is estimated that this volume could be sustainably produced in the long-term from an industrial forest plantation estate of around 180 million hectares, given the current global distribution of industrial forest plantations.
In reality, however, it is likely that the amount of plating required under Scenario 3 represents the upper boundary of new planting that might be achieved in the foreseeable future. Achieving the levels of new planting required under Scenario 3 would be equivalent to a widespread application of the forest plantation development model adopted by the "Southern plantation producers" (see page 49). This would probably require a significant paradigm shift from current forestry practices and ecological philosophy for most countries of the world. For example, Clapp (1995b), in discussing the costs and benefits of forest plantation development in Chile, is sceptical of the efficacy of this model:
"I argue that the environmental balance [resulting from plantation establishment] is positive, but not uniformly so, and that the [current] plantation model of forestry minimises those benefits."
This view is prevalent in many countries, particularly in Europe and North America. It suggests, however, that even ignoring physical constraints and questions of future roles of natural forests, a forest plantation development model as extensive as that required under Scenario 3 would be unlikely to occur without major policy changes.
In several countries with large forest plantation resources, shortages of water are the important factor constraining additional development. This is particularly the case in many African countries. As already noted, the countries with the largest forest plantation resources in Africa are South Africa and the North African countries of Morocco, Tunisia, Libya, Algeria, Ethiopia and Sudan. In all of these countries forest plantation development is somewhat constrained by water shortages (see Box 13). Also, in many countries, particularly Morocco, Tunisia and Algeria, the objective of establishing greater areas of forest plantations has been primarily to combat desertification rather than to produce wood, reinforcing the importance of water issues in these countries.
Box 13 Forest plantations and water scarcity in South Africa
South Africa provides a good example of a country where water scarcity is limiting future forest plantation expansion. The importance of water as a constraining factor on afforestation is highlighted by the fact that forestry and water issues are brought together in the government administration (The Department of Water Affairs and Forestry). Another factor pointing to the fact that water scarcity is probably the most important constraint on further forest plantation development is that forest plantation development is the only land-based activity that is regulated by the government.
For example, the South African White Paper on Forestry (DWAF, 1996) notes:
"Controversy about the effects of afforestation on water supplies began in the 1920s, and continues today. This led to the implementation of controls on afforestation that have been applied since 1972 through the afforestation permit system. In 1986 the industrial forests in South Africa were estimated to consume about 1.2 billion cubic metres of water that would otherwise have entered rivers and streams, and been available for other uses. This volume equated to about 30% of the amount used for urban and industrial purposes, or about onetenth the volume used in irrigated agriculture. The water consumed is a cost required to support the forestry sector as a contributor to our economy."
The South African Government is currently examining ways in which water charges might be introduced for activities using this resource without currently having to pay for it.
The main economic constraint to future forest plantation expansion concerns the rate of return that might be achieved on additional forest plantation projects. The rate of return on each new project will depend crucially on the product prices and the speed with which the return will be generated. In general, this latter factor (i.e. how fast a tree can grow) will also be related to some of the ecological constraints already described above. Thus, for example, forest plantations in countries in the boreal zone, such as Canada; Sweden; Finland; and the Russian Federation, may simply not grow fast enough to compete with countries that can establish fast growing species, such as: Brazil; Chile; New Zealand; and Indonesia.
Indeed, Sohngen et al (1997) summarise this (natural) advantage as follows:
"Because land is generally available for these plantations and because relatively small management inputs can lead to substantial future gains, subtropical plantations are a better investment than are temperate forests."
At present this view, perhaps, pays insufficient attention to many of the other elements of competitive advantage. For example, many developed temperate countries continue to have significant advantages in infrastructure, technology and labour skills, over developing countries and have the advantage of the presence of strong clusters of supporting industries. Another important point is that several large countries in the boreal zone, for example, Canada and the Russian Federation, will continue to have a large comparative advantage in land availability. In both of these countries, very low rural population densities and limits on alternative uses for forest land mean that the opportunity cost of land used for forest plantation development is likely to remain very low. Finally, developed countries will also continue to have an important advantage arising from the integration of processing and marketing functions that is quite common in their forestry sectors.47
The crucial point is that the ability to grow trees quickly is only one, in a complex set of factors that will determine success in the forestry sector. An excessive focus on the exploitation of natural advantage is unlikely to be a good substitute for strongly based systemic development. Thus, despite the advantages pointed out by Sohngen et al, the future distribution of industrial roundwood production and forest plantation establishment may not vary greatly from the present situation.
Another important feature to take into account is the private-sector's reaction to external events (e.g. response to policy changes) and in response to competitors' actions. A major policy change in one country may trigger significant changes in other regions. For example, changes to (and uncertainty about) forest policies in the united States of America in response to the Spotted Owl issue had significant effects on wood prices around the Pacific Rim. In several countries, this stimulated additional forest plantation establishment. Similarly, there remains a variety of conflicting views over the future scarcity of wood and fibre. To a large extent, this divergence of opinion is the result of the poor quality, and scarcity, of globally aggregated data. As data availability and quality improves, private sector and governments may alter their forest plantation strategies.
In terms of future forest plantation establishment, the prevailing economic principle may well be the law of diminishing returns. A theoretical case can be made, for example, that the best forest plantation sites (in comparative advantage terms) are already occupied. Consequently, future forest plantations will be both less profitable and will have to appear attractive to potential investors with portfolios already containing significant areas of forest plantations.
Similarly, market perceptions of roundwood from forest plantations are that it is generally inferior to roundwood from natural forests. While forest plantations may be able to expand their role in fibre-based industries, there remain considerable barriers to expanding market share in, for example, the markets for luxury hardwoods (plantation grown teak (Tectona grandis) may be the exception to this rule).
To summarise, policy and economic constraints may also, therefore, play a role in restricting forest plantation establishment to a level below that anticipated under Scenario 3. Conversely, current supply and demand modelling (see, for example, FAO, 1998b) anticipates an increasing need for forest plantations to meet future consumption requirements. This suggests that forest plantation areas will continue to expand and that the potential level of future roundwood production from forest plantations is at least likely to exceed that suggested by Scenario 1.
A case can be made for two additional future development scenarios, both of which lie outside the boundaries of the scenarios presented above. Both of these scenarios are related to structural and institutional changes that might have a dramatic effect on the forest plantation sector. These scenarios relate to:
1. the potential for forest plantations to quality for subsidies as part of a package of policy measures to reduce net carbon emissions; and
2. the possibility of plantation areas declining due to decreases in reforestation.
One of the major current uncertainties about the outlook for forest plantation development in the future, is the role that forest plantation projects might play as carbon-offset projects. With most developed countries agreeing to meet greenhouse gas emission reductions targets of 5-8 percent below 1990 levels, there is scope for massive increases in tree planting to act as carbon sinks.
The signing of the Kyoto Protocol also provides a basis for the formalisation of a market (between countries) for carbon-sequestration projects, but whether such a market will emerge as a significant force in forestry remains a moot point and will depend on a number of important developments. The first of these will be the extent to which countries enforce measures to ensure the achievement of emissions targets. The second is whether tree planting to sequester carbon will be relatively cost-effective when compared with other mitigation measures. A third uncertainty concerns whether forest plantations established under such programmes could be used for roundwood production.
Making predictions about the impact of policies to reduce net carbon emissions on future forest plantation development would, at the current time, be highly speculative. It seems likely, however, that carbon sequestration in forest plantations would be a relatively attractive option for a number of industries, particularly where reducing industrial carbon emissions requires large investments in technology or redesign. In this instance, the outcomes described in Scenario 3 would become considerably more likely and Scenario 3 might even understate future growth in new planting (although the amount of roundwood that might eventually be produced from such forest plantations remains uncertain). Thus, it is possible that carbon-offset schemes could encourage the creation of a forest plantation resource significantly larger than any described in the formal scenarios. This may result in greater production of roundwood from forest plantations, probably less production from natural forests and probably lower prices for industrial roundwood.
In a number of countries, most notably in Europe and South America, a large proportion of planting is, or has been, carried out under forestry incentive programmes. In many cases incentives have provided an important stimulus to planting, and may have been a more important driver than expectations of future prices and earnings. Some of these programmes have now been eliminated, or significantly reduced and there is a possibility that planting rates, both afforestation and reforestation, in these countries could fall. Consequently, it is conceivable that the global forest plantation could fall in the future.
A second factor that might lead to declining reforestation is the gradual privatisation of forest plantations and the shift in emphasis in new planting programmes from direct action by the government to encouragement of the private-sector. To date, there is little evidence that privatisation and similar community devolution schemes have resulted in declining planting or plantation areas. In fact, in many instances, the rate of forest plantation establishment has actually accelerated. There are, nonetheless, circumstances (for example, a prolonged period of low wood prices) that could result in marked declines in forest plantation investment.
At the moment, it seems unlikely that such a scenario will arise in the next few years. However, over the next 50 years, it is quite possible that periods of net reduction in the total area of forest plantations could arise.
This analysis supports a central conclusion that the role of forest plantations in meeting future wood and fibre demands will increase in the near-term, irrespective of future rates of forest plantation establishment. Projected roundwood production potential from forest plantations for the next decade is largely determined by trees already in the ground and, in many countries, a considerable increase in areas of forest plantations reaching harvestable age is expected. Thus, by 2010, the industrial roundwood production potential from industrial forest plantations is estimated to increase from the current 388 million cubic metres to around 600 million cubic metres, whatever happens. Potential production of wood fuel from non-industrial forest plantations is expected to roughly double from the current figure of around 80 million cubic metres.
Beyond 2010, the production forecast depends largely on what assumptions are made about new planting rates (considered here) and on assumptions about improvements in annual increment or yield (assumed to equal zero here). There is tremendous scope for forest plantations to play a dominant role in industrial roundwood production, but much will depend upon future policy decisions and market shifts. A reasonably likely scenario is that the proportion of industrial roundwood coming from forest plantations will probably increase, but that natural forests will probably continue to account for over half of all industrial roundwood production. In terms of wood fuel, production will continue to largely come from trees outside forests and forest plantations should be expected to continue to supply only a small proportion of total wood fuel consumption.
The question of where future forest plantation development is most likely to occur also remains unclear. At present, many governments remain active in forest plantation establishment, either directly through state planting programmes, or indirectly by providing incentives to the private-sector. In some instances, incentives can be justified by the non-market benefits produced by forest plantations. In others, the incentives are merely maintaining wood production capacity. In any event, under these circumstances, competitive and comparative advantages are not clearly emerging.
The most significant increases in forest plantation areas in the immediate future will be in countries where specific state planting programmes are currently in place, most notably China and India. In Europe, forest plantation establishment is likely to be mainly dictated by the life span of current incentive policies. Europe is largely self-sufficient in terms of roundwood requirements and the development of a significantly larger resource base seems a relatively unlikely development.
South America and Oceania are likely to continue to expand forest plantation areas due to the perception that these areas have a significant competitive advantage in the forest plantation sector. The extent to which forest plantations increase in these regions is likely to be dictated by the extent to which this perception is sustained. If returns on forest plantation investments fall then afforestation rates in these regions are likely to slow.
Profitability in the Southern plantation producers is likely to be determined by conditions in North American and Asian markets and by future production from forests in these regions. If the current trend towards increasing regulation of harvesting in the natural forest increases in these regions then forest plantation establishment there is also likely to accelerate. Of important interest will be developments in production from the natural forest in the US Pacific-Northwest, Canada, Indonesia and Malaysia.
Forest plantation establishment in countries of the former USSR and Africa seems unlikely to accelerate in the immediate future. In the former, economic restructuring is likely to mean that forest plantation investment will be of relatively low priority, particularly given the extensive natural forest resource in several of these countries. In Africa, the absence of strong infrastructure is likely to remain a significant competitive disadvantage for many countries. It is difficult to see true competitive advantage in forest plantation investment emerging in many African countries, including even those that have existing major industries based on production from the natural forest. Non-industrial forest plantation establishment may, however, accelerate in Africa, particularly where population pressures demand more intensive production of wood fuel.
38 Any attempt to guess the future policies of the large number of countries included in this study is unlikely to be successful.
39 Mortality rates here do not refer to mortality in the silvicultural sense (i.e. the reduction in stems per hectare due to mortality throughout the growing cycle). Rather it refers to the process of reducing reported planting areas to account for inaccuracies in reported statistics and crop failure. Two adjustments have been made as part of this process. Firstly, the area of forest plantations in each age-class (from the literature search) has been reduced so that, in total, they equal the netted-down total area reported by Pandey (1997). In this process, an exponential weighting function was used to reduce areas in the older age-classes by relatively more than in the younger age-classes (on the grounds that statistics on the areas most recently planted are likely to be more accurate). Secondly, statistics on planting for each country report the area planted but, in many countries, a proportion of that area dies after the first few years. Another exponential function was, therefore, used to simulate the likely losses from crop failure across the age classes derived for a country as forest plantation areas mature throughout the projection period. This function placed most of the losses in areas that have been planted recently, because the probability of crop failure is likely to decline over time as forest plantations mature. These adjustment factors therefore, vary over time, but they do not vary between the different scenarios. For further details, see: Appendix 1 on page 131.
40 For example, if the area of forest plantations in a country has increased from 1 million hectares in 1990 to 1.5 million hectares in 1995, this would imply a current establishment rate of 100,000 hectares per year. Under this scenario, it would be assumed that this country will establish 100,000 hectares per year from 1995 to 2004, 80,000 hectares per year from 2005 to 2014, 60,000 hectares per year from 2015 to 2024, 40,000 hectares per year from 2025 to 2034 and 20,000 hectares per year from 2035 to 2050.
41 Note that these projections are projections of potential production (i.e. the volume of wood that will reach the end of its rotation period in each year). Actual production in any year may be different to this for a number of factors. Consequently, these projections should be viewed as projections of volumes that forest plantations could produce rather than as the level of production that they actually will produce.
42 Projecting future potential production from non-industrial forest plantations is subject to a number of additional uncertainties. Firstly, it is often difficult to clearly identify whether non-industrial forest plantations are managed wood fuel production or are managed for water or soil protection, recreation or similar non-productive purposes. Secondly, even when it is clear that a forest plantation will be used for wood fuel production, it is often impossible to know whether the main stem will be harvested and/or whether other parts of the woody biomass will be utilised. Wood fuel plantations may also be managed under management systems that are more difficult to model at this scale (e.g. coppice, coppice with standards, or continuous cover management systems).
The projections presented here are projections of potential production (from main stem volume) from the total area of non-industrial forest plantations. This is based on the assumption that all of the wood and fibre produced in these forest plantations will eventually be used for something (although that may not be for wood fuel). If anything, these projections are probably an underestimate of potential production because of the large amount of non-stem biomass that is often collected and used as wood fuel.
43 i.e. by an absolute amount each year equal to one percent of the area in 1995, not one percent compounded.
44 Note: due to difficulties with separating data, annual planting rates cited in Pandey include elements of both new planting and restocking for some countries and other information had to be used to solve this problem.
45 Note that this discussion only considers future projections of industrial roundwood supply and demand. Estimates of current wood fuel production and consumption are believed to be very unreliable and any projections based on current information are likely to be highly uncertain. Wood fuel supply and demand and a revision to current estimates will be considered in a separate study to this (GFPOS/WP/05).
46 These alternative options are quite feasible and it is likely that gains in several of these areas will be made in the future anyway, but have not been taken into account here.
47 That is, significant advantages (for example, supply security, stable raw material pricing and material scheduling advantages) generally accrue to companies that own both forests and processing facilities. Similarly, processors located close to their markets generally capture marketing advantages. The present structure of the forest industry suggests, therefore, that developed countries will continue to have this important competitive advantage, where forest plantations are owned as part of a vertically-integrated system and are located close to their final market.