Indicative forest plantation yields for hardwood species grown in the tropical and subtropical zone are shown in Table 6. As the table shows, the species with the highest yields are Eucalyptus species, Acacia mangium and Gmelina arborea. Species that produce roundwood that is generally more valuable, such as Mahogany (Swietenia macrophylla) and Teak (Tectona grandis), tend to have a lower yields than the lower value species.
The highest yields are generally found in South America, followed by Asia and Africa. This variation probably reflects variations in the intensity and quality of management rather than growing conditions.
The yield of pine species growing in the tropical and subtropical zone probably varies from around 20 m3/ha/year for Pinus radiata in the temperate parts of South America and Australia to between 12 m3/ha/year and 15 m3/ha/year for Pinus caribaea in Central and South America.
As a general rule, plants and trees tend to grow more quickly in the tropical and subtropical zone than in the temperate and boreal zone. Therefore, it would be expected that, in general, yields would tend to be higher in the tropical and subtropical zone. However, the substantial variability in the quality of planting sites in most countries can significantly affect yields. The local climate, altitude and geomorphology, the suitability of a species for a site and the impact of pests and diseases can also have significant effects on yields. Overall, the great variability in yields that are currently achieved in the tropical and subtropical zone suggests that there is considerable potential to increase the average yield of forest plantations in most countries in the tropical and subtropical zone.
Current research into growing Eucalyptus species presents a good example of this potential. Large scale forest plantations of Eucalyptus species rarely achieve yields of more than 25 m3/ha/year at present, but significant advances may well be achieved in the near future. For example, trials of rooted Eucalyptus cuttings in Brazil have already been reported to give yields of up to 100 m3/ha/year (South, 1998). However, there is uncertainty about the extent to which such results can be translated into higher yields in large-scale forest plantations. Another uncertainty is whether other problems (e.g. poor wood quality or susceptibility to disease or windthrow) may arise in forest plantations of species with very high yields.
In addition to advances in genetics and plant breeding, there is also considerable scope for increasing yields in tropical and subtropical forest plantations by improving forest plantation management and investing in silvicultural treatments. For example, planting failure rates in forest plantations in the tropical and subtropical zone are often high, with Pandey (1995) reporting success rates of only 26 percent in the Philippines, 47 percent in Laos and 57 percent in Colombia and failure rates of up to 70 percent in some individual cases. A number of other improvements could also significantly increase yields, such as: improving the matching of species to sites; improvements in plant storage, handling and planting; and investment in site preparation and soil improvement, weeding, pruning and thinning. Indeed, improvements in these areas probably have more potential to increase yields and can be applied across larger areas, than advances in genetics and plant breeding.
Table 6 Indicative forest plantation yields by species and country for hardwood species grown in the tropical and subtropical zone
Species |
Yield m3/ha/year |
Countries |
Acacia auriculiformis |
6.5 - 10.0 |
Bangladesh, Benin, Haiti, India, Madagascar, Myanmar, Philippines, Sierra Leone, Sri Lanka, Thailand and Vietnam |
Acacia mangium |
12.0 - 19.0 |
Indonesia, Malaysia and Papua New Guinea |
8.0 - 12.5 |
Bangladesh, Laos, Panama, Philippines, Sierra Leone, Sri Lanka and Vietnam | |
Casuarina species |
5.0 - 7.5 |
India and Vietnam |
1.5 - 2.5 |
Angola, Benin, Cuba, Kenya, Madagascar, Mauritius, Mozambique, Senegal, Somalia and Thailand | |
Dalbergia sissoo |
3.0 - 5.0 |
Bangladesh, Bhutan, Burkina Faso, India, Nepal, Nigeria and Pakistan |
Eucalyptus species |
16.0 - 25.0 |
Argentina, Brazil, Chile and Uruguay |
12.0 - 19.0 |
Australia, Rep Congo, Malawi, Papua New Guinea, South Africa, Swaziland, Uganda, Zambia and Zimbabwe | |
8.0 - 12.5 |
Burundi, China, Colombia, DR Congo, Costa Rica, Cuba, Ecuador, El Salvador, Ethiopia, Gabon, Guatemala, Kenya, Madagascar, Mauritius, Nicaragua, Nigeria, Pakistan, Paraguay, Peru, Philippines, Solomon Islands, Tanzania, Thailand and Venezuela | |
6.5 - 10.0 |
Malaysia and Sierra Leone | |
4.0 - 6.0 |
Algeria, Angola, Bangladesh, Benin, Bolivia, Burkina Faso, Cameroon, Cape Verde, Chad, India, Indonesia, Laos, Lesotho, Mali, Morocco, Mozambique, Myanmar, Namibia, Nepal, Niger, Rwanda, Senegal, Sri Lanka, Sudan, Togo, Tunisia and Vietnam | |
Gmelina arborea |
12.0 - 19.0 |
Belize, Bhutan, Bolivia, Brazil, Burkina Faso, Colombia, DR Congo, Costa Rica, Cote D'Ivoire, Cuba, Dominica, Gambia, Ghana, Guatemala, Guinea, Indonesia, Laos, Liberia, Malawi, Malaysia, Mali, Nicaragua, Nigeria, Philippines, Sierra Leone, Solomon Islands and Venezuela |
Swietenia macrophylla |
5.0 - 7.5 |
Bangladesh, Benin, Cameroon, Dominica, Fiji, Guatemala, Indonesia, Jamaica, Nigeria, Philippines, Solomon Islands, Sri Lanka, St Vincent and the Grenadines and Trinidad and Tobago |
Terminalia species |
8.0 - 12.5 |
Costa Rica and Cote D'Ivoire |
6.5 - 10.0 |
Rep Congo, DR Congo, Guinea, Nigeria, Papua New Guinea, Senegal, Sierra Leone and Solomon Islands | |
5.0 - 7.5 |
Bhutan, India and Jamaica | |
Tectona grandis |
8.0 - 18.0 |
Belize, Colombia, Costa Rica, Jamaica, Nicaragua, Panama and Trinidad and Tobago |
4.0 - 6.0 |
Bangladesh, Benin, Bhutan, Burkina Faso, Cote D'Ivoire, Ecuador, Ghana, India, Indonesia, Laos, Liberia, Malaysia, Myanmar, Nigeria, Papua New Guinea, Philippines, Senegal, Solomon Islands, Sri Lanka, Sudan, Tanzania, Thailand, Togo and Vietnam |
Notes: yields are expressed as mean annual increment (MAI) over the "likely" rotation length. It must be stressed that these yields, which have been used in the modelling process, are provisional and are only very broad estimates of expected average yields. Leech (1998) and the author have compiled these figures.
Information about forest plantation yields in temperate and boreal countries is presented in Table 7. The table shows that the species with the highest yields are generally Eucalyptus and Pinus species, particularly where they are grown in the warmer parts of the region. Again, the generally more valuable hardwood species (Quercus and Fagus species) tend to have lower yields.
Table 7 Indicative forest plantations yields by species and country in the temperate and boreal zone
Species |
Yield m3/ha/year |
Countries |
Pinus species |
18.0 - 24.0 |
New Zealand |
4.0 - 14.0 |
Japan, Portugal, Spain, Turkey, United Kingdom and United States of America | |
2.0 - 10.0 |
Belgium, Denmark, France, DPR Korea, Rep Korea, Latvia, Libya, Lithuania, Sweden, Syria and Ukraine | |
1.0 - 5.0 |
Russian Federation | |
Picea and Abies species |
12.0 - 18.0 |
Ireland |
8.0 - 16.0 |
Denmark, France, Turkey, Ukraine, United Kingdom and United States of America | |
4.0 - 12.0 |
Latvia, Lithuania and Russian Federation | |
Larix species |
4.0 - 12.0 |
Japan, DPR Korea, Rep Korea, United Kingdom and United States of America |
Cupressus and Chamaecyparis species |
2.0 - 8.0 |
Japan and Syria |
Cedrus and Cryptomeria species |
4.0 - 10.0 |
Japan, Latvia, Lithuania, Russian Federation, Turkey and Ukraine |
Eucalyptus species |
10.0 - 15.0 |
Spain, Portugal and United States of America |
5.0 - 10.0 |
Libya and Syria | |
Quercus species |
2.0 - 8.0 |
France, Latvia, Lithuania, Portugal, Russian Federation, Spain, Turkey, Ukraine, United Kingdom and United States of America |
Fagus species |
2.0 - 12.0 |
Denmark, France, Spain, Turkey, United Kingdom and United States of America |
Populus species |
8.0 - 25.0 |
France and Italy |
Betula species |
4.0 - 10.0 |
Rep Korea, Finland, Sweden, United Kingdom and United States of America |
Sources: various sources compiled by the author.
Populus species and Salix species (willows) also have potentially high yields. For example, yields of over 40 m3/ha/year have been achieved in research plots planted with some Populus species. However, these species are not generally used for industrial roundwood production (except occasionally for pulpwood production), although Populus species are often planted to provide shelter and soil and water protection and are sometimes used to produce wood fuel. High-yielding Salix species are also attracting attention as potential bioenergy crops, in situations where they can be grown under short rotation coppice management systems on cycles of three to five years.
Compared with the tropical and subtropical zone, forest plantations in the temperate and boreal zone generally achieve lower yields. Indeed, as a general rule, the maximum potential yield that can be achieved with any particular species is strongly and negatively correlated with latitude. Thus, in general, forest plantation yields in temperate parts of the zone are lower than in the subtropics and tropics, but generally higher than in the boreal region.
The main factor limiting yields in forest plantations in the temperate and boreal zone is temperature and the length of the growing season. Therefore, latitude, aspect and altitude are important constraints to yield, although other factors can also be the main constraint in some areas. Such factors include general climatic conditions, such as wind speed and average rainfall, along with site-specific factors, such as drainage, nutrient availability and soil depth.
Compared with tropical and subtropical zone, there is probably little scope for significantly increasing yields in most countries in the temperate and boreal zone, with current levels of technology. Improvements in plant breeding and genetics have produced some gains in yields, but nothing like what has been achieved with Eucalyptus species in the topical and subtropical zone. Again, improved forest management and investment in silviculture are probably the two routes that could bring the greatest and most widespread gains in yields. However, even these gains might be quite limited because temperate and boreal forest plantations are already, on the whole, fairly well managed.
Where forest plantations are managed under a system of clearfelling and replanting, the rotation length is the length of time between establishment (i.e. planting the trees) and clearfelling the final crop. Under coppice and continuous cover or selection forestry systems, the rotation length or cutting cycle is the length of time between major harvests of roundwood.
Rotation lengths are determined by a number of factors, including:
· growth rates (which are, in turn, determined by site productivity, silviculture, species, thinning regimes and spacing);
· desired wood and fibre properties;
· site constraints (such as susceptibility to damage from wind);
· socio-economic factors (such as amenity values and income from recreational activities, that tend to increase as trees get older); and
· the rate of return or profitability from roundwood production over the rotation.
The last of these factors (i.e. profitability) is usually the most important factor affecting the choice of rotation age in industrial forest plantations, closely followed by the growth rate or expected yield of the forest plantation.
The literature on forest economics contains many references to the calculation of economically optimal rotation lengths (for example, see: Johansson and Lofgren, 1985). Generally speaking, rotation lengths tend to be shorter where forest plantation owners have a higher preference towards the present rather than the future (i.e. they use a high interest rate or discount rate to calculate the economically optimal rotation length) or, more generally, where they seek to maximise the profitability of their forest plantations.
Figure 14 The effect of yield on the economically optimal rotation length for Pinus species in forest plantations in Lithuania
Notes: cost and price data collected to produce these figures comes from Whiteman (1999). The exchange rate in 1999 was 1 US$ = 4 Litas (Lt). The yield classes are expressed as top height in metres at 100 years old (i.e. Pine-100-18 means 18 metres high at 100 years old). A rough estimate of maximum mean MAI in m3/ha/year is also given in the legend (i.e. YC 4.2 means a maximum mean MAI of 4.2 m3/ha/year).
Economically optimal rotation lengths also tend to be shorter where yields are higher. For example, Figure 14 shows the effect of yield on the economically optimal rotation length (i.e. the rotation length at which net present value - in this case calculated using a 3% discount rate - is maximised) for Pinus species in Lithuania. This falls from around 80 years in the lowest yield class to 55 years in the highest yield class.
The concept of an economically optimal rotation length is mostly used in long-term planning to determine when forest plantations should be harvested. In the short-run, rotation lengths are often varied to take into account current market conditions. This is particularly the case in single plantation blocks, where it is relatively easy to bring forward or delay final harvesting, but may occur less in larger forest plantations of mixed species and age-classes or where forest owners face roundwood supply or cash-flow constraints.19
Other factors that can affect rotation ages are government regulations and changes in technology. In some countries, rotation ages are specified in government regulations or are implicitly enforced through harvesting regulations. For example, in Lithuania, current regulations require forest owners to manage their forest plantations on rotations that are somewhat longer than what would probably be economically optimal. A number of other Central and Eastern European countries have similar regulations, which tend to be justified on the grounds of the non-market benefits (i.e. amenity, protection and biodiversity values) associated with older trees. In contrast, trends in technology tend to have the opposite effect and are making it increasingly profitable to manage forest plantations with shorter rotation lengths, by increasing the value and marketability of smaller sized roundwood (see, for example, Box 2).
Box 2 "Millennium Forestry" - changing rotation lengths in New Zealand
In 1998, the New Zealand company Carter Holt Harvey Ltd. (CHH), announced a change in its management strategy for forest plantations of Pinus radiata over the next 20 years. Under its "Millennium Forestry" strategy, CHH plans to plant 555 trees per hectare, with no pruning or thinning and harvest the trees when they are 20 years old. This is a marked departure from the traditional New Zealand "Direct Sawlog Regimes", which tended to prune trees to 6 metres, thin to a final stocking of around 250 stems per hectare, and harvest at ages 28-32. CHH believes that in 20 years price premiums for clearwood Pinus radiata will be smaller and improvements in processing technology will enable unpruned wood and fibre to be profitably converted into products that can compete directly with clearwood sawn timber.
In essence, the CHH strategy centres on economic rationality. The strategy seeks to maximise fibre production and shorten the period between investment and realisation of returns. The strategy has created considerable controversy in New Zealand, with critics arguing that assumptions on the value (and properties) of shorter rotation trees might well be faulty and that, in terms of ecology, shorter rotation plantations move further from "desirable" natural forest traits and closer to "undesirable" cropping paradigms.
Source: adapted from New Zealand Journal of Forestry (1999).
As the above discussion implies, there is currently an enormous amount of variability in rotation lengths in forest plantations. In forest plantations managed under the clearfelling and replanting system, rotation lengths range from about seven years (for some forest plantations of Eucalyptus species grown for pulpwood production in South America) to more than 100 years (for many hardwood species and some softwood species grown in Europe). Cutting cycles in coppice systems generally vary from five to 25 years, while cutting cycles under continuous cover or selection forestry systems are usually at least 25 years in length.
Information collected as part of the literature search for this exercise included information about rotation lengths typically used in each country by species and yield class. This information has been used in the modelling exercise to estimate final felling ages and volumes by species and yield class in each country in the model.20 There is some uncertainty associated with this information, but such uncertainty is likely to be less problematic for the modelling of future roundwood production than, say, uncertainty about the estimated mean annual increment, which would be compounded throughout the duration of a forest plantation rotation.
Based on all of the information described above, a simple production forecasting model has been constructed for this exercise. This model produces projections of the volume of roundwood that could be produced from the world's forest plantations based on their area, species, type, yield and age structure. It must be stressed that this is a projection of potential production and that actual production may differ from this for a number of reasons. However, due to the high levels of investment in forest plantations, it seems likely that most forest plantations will be fully utilised for wood production and thus, that actual production will be quite close to potential production.
Current statistics collected about roundwood production do not differentiate between roundwood produced from the natural forest and roundwood produced from forest plantations. Thus, the production forecast model was first used to estimate how much industrial roundwood might have been produced in 1995 in industrial forest plantations. Based on the age structure of industrial forest plantations in 1995, the model suggested that industrial forest plantations could have accounted for about 331 million cubic metres of industrial roundwood production or about 22 percent of total global industrial roundwood production.
Assuming that non-industrial forest plantations are used mainly for wood fuel production, the model also suggested that non-industrial forest plantations could have produced about 86 million cubic metres of wood fuel or just over 4 percent of total global wood fuel production.
Combining these two figures, the estimated total roundwood production potential from all types of forest plantations is 417 million cubic metres or just over 12 percent of total global roundwood production in 1995.
Figure 15 Estimated potential roundwood production from forest plantations as a percentage of actual production in 1995
Sources: FAO (1997b); and author.
Figure 15 shows the estimated potential roundwood production from forest plantations as a percentage of actual production in 1995. This information is presented by geographical region and by type of production (i.e. industrial roundwood - assumed to come from industrial forest plantations - and wood fuel production - assumed to come from non-industrial forest plantations). The figure shows a number of interesting features.
Firstly, the proportion of industrial roundwood that might come from industrial forest plantations is much higher than the proportion of wood fuel production that might come from non-industrial forest plantations. This difference appears at the global level and in all of the regions and demonstrates the far greater importance of forest plantations to global industrial roundwood supply than to wood fuel supply.
The second interesting feature is the difference in the importance of industrial forest plantations to industrial roundwood supply in different regions. Industrial forest plantations appear to have the greatest importance in Oceania, where up to 80 percent of industrial roundwood may be coming from industrial forest plantations. They may also contribute up to 35 percent of industrial roundwood supply in Africa, 27 percent in South America and 23 percent in Asia. The high level of potential production from forest plantations in each of these regions is due to a small number of countries in each case (i.e. Australia and New Zealand in Oceania, Chile and Brazil in South America, China and Japan in Asia and South Africa in Africa).
The final point of interest concerns the relative unimportance, at the global level, of forest plantations for wood fuel production. In all of the regions, potential wood fuel production from non-industrial forest plantations is estimated to be less than 7 percent of total wood fuel production. This is due to the relatively small area of non-industrial forest plantations. Indeed, even this estimate may overstate the contribution of non-industrial forest plantations, because many non-industrial forest plantations are actually planted for non-commercial purposes other than wood fuel production and may not, therefore, be available for wood fuel production. On the other hand, however, it is likely that a significant volume of wood residues from industrial roundwood grown in industrial forest plantations will be burned as fuel. Furthermore, it must be remembered that the production forecasting model estimates main stem yield volume while, in reality, a much greater proportion of a tree's biomass is often used as fuelwood (notably branches, twigs and tops). Thus, considering that these factors work in opposite directions, the estimated potential production of wood fuel from non-industrial forest plantations may not be all that inaccurate.
Generally, very little information is available about the mixture of roundwood sizes and qualities that might be produced in industrial forest plantations and, consequently, the products that might be produced from such wood. In only a few countries, where forest plantations account for almost all industrial roundwood production, can this information be easily obtained. For other countries, estimates can be made based on the forest plantation species mixture, but these are likely to be highly speculative and potentially misleading.
Only five countries (Denmark; Ireland; New Zealand; Chile; and South Africa) have a proportion of industrial roundwood production from forest plantations that is sufficiently high enough to make a reasonably reliable assessment of the product mixture coming from their forest plantations. Table 8 shows the mixture of products that is produced from industrial roundwood in these countries.
Table 8 Product mix from forest plantations in selected countries in 1995
Country |
Total |
Proportion |
Industrial roundwood utilisation (%) | ||||
industrial roundwood harvest (m3) |
coming from forest plantations (%) |
Round-wood exports |
Sawn-wood |
Wood-based panels |
Wood pulp |
Mining & other uses | |
Denmark |
1,797,000 |
_100.0 |
23.0 |
46.4 |
5.0 |
25.6 |
n.a. |
Ireland |
2,140,000 |
100.0 |
25.0 |
55.0 |
20.0 |
0 |
n.a. |
New Zealand |
17,627,000 |
99.0 |
31.4 |
31.5 |
7.4 |
29.7 |
n.a. |
Chile |
21,387,000 |
>85.0 |
40.0 |
31.4 |
4.0 |
24.6 |
n.a. |
South Africa |
17,600,000 |
100.0 |
13.3 |
29.6 |
n.a. |
35.5 |
21.6 |
Source: derived from FAO (1997b).
Forest plantations in all five countries are predominantly planted with temperate zone species21 and, consequently, the product mixtures coming from forest plantations in these countries can not be used as a guide to what might be produced in tropical forest plantations. Furthermore, it is quite likely that a very different product mix might be produced in temperate forest plantations with very different species mixtures to these countries. However, these figures illustrate some features that are worth noting.
Between 30 percent and 50 percent of industrial roundwood in New Zealand, South Africa and Chile is used for the production of wood-based panels, wood pulp, mining and other uses. Assuming that the majority of roundwood exports are exports of sawlogs, this would suggest that pulpwood production from industrial forest plantations is, therefore, between 30 percent and 50 percent, while sawlog production is between 50 percent and 70 percent. Denmark and, in particular, Ireland appear to produce a slightly higher proportions of sawlogs (70 percent to 80 percent). Given that growth rates in the first three (Southern) countries are somewhat higher than in the latter two, this would suggest that, where forest plantation yields are higher, a greater proportion of the roundwood produced from forest plantations might be in the pulpwood category.
As noted above, economic factors tend to favour using short rotations in forest plantations which, in turn, tends to favour the production of pulpwood rather than sawlogs. However, as the above differences demonstrate, where forest plantation yields are lower, it appears that forest plantation owners tend to pursue a strategy of producing relatively more higher value sawlogs and veneer logs. This may have something to do with the better structural qualities of roundwood coming from slower growing forest plantations.
19 Profit maximising strategies will vary according to the size and age-class structure of the forest plantations managed by a single owner and their long-run management objectives. For example, an owner with many forest plantations of varying age-classes will have a different cash-flow requirement and roundwood production strategy to a smaller owner with a small forest plantation all of one age-class.
20 The modelling in this paper allows for harvesting to be distributed around an average rotation length and thus mimics some of the uncertainty between planned and actual rotation lengths (see Appendix 2).
21 In addition, forest plantations in the latter three countries are predominantly planted with Pinus radiata.