Previous Page Table of Contents Next Page

Silviculture for sustainable management of tropical moist forest

J.G. Bertault, B. Dupuy and H.F. Maître

Jean Guy Bertault, Bernard Dupuy and Henri Felix Maître are associated with the Forest Management Programme of the International Cooperation Centre on Agrarian Research for Development (CIRAD-Forêt) formerly the Centre technique forestier tropical - Nogent-sur-Marne, France.
Note: This article is adapted, with permission, from a paper prepared for the Tropical Silviculture Workshop at the Centennial Conference of the International Union of Forestry Research Organizations (IUFRO), 1-3 September 1992, Berlin.

This article examines the silvicultural options available to the forest manager by reviewing the results of research on plantation and natural forest dynamics. It examines existing knowledge on the behaviour, growth, regeneration and mortality of natural tree stands in relation to human impacts: logging for timber, thinning for improvement, accidental fires, etc. These major components must be thoroughly understood before an appropriate silviculture can be prescribed for any stand. This, in turn, is essential for forest management leading to the sustainable production of forest products and services, while maintaining biodiversity, and thus to the achievement of technical requirements for the conservation of tropical forest ecosystems.

The silvicultural research programme conducted by the former Centre technique forestier tropical (CTFT) in tropical moist forests over the past 20 years has attempted to develop and synthesize knowledge on the establishment and management of plantations for timber production, notably in French-speaking Africa. It has also focused on the dynamics of natural forests, with and without human intervention. For 15 years, this work has been based on a network of large-scale experimental projects located not only in Africa (Côte d'Ivoire, Central African Republic and Gabon), but also in America (Brazilian Amazonia and French Guiana) and recently in Asia (Indonesia).

Historical background

It used to be generally assumed that tropical moist forest ecosystems perpetually renewed themselves and thus represented an inexhaustible wood potential. However, with the expansion of harvesting, this was quickly found not to be the case. The valuable species that were harvested did not necessarily regenerate and it became apparent that logged-over forests would become impoverished unless silvicultural intervention could maintain or increase their timber potential.

Thus the first silvicultural practices were aimed at pragmatic procedures to sustain the wood yields of a few economically desirable species. Techniques varied considerably within and between countries as well as over time, depending on the great diversity and complexity of the forest ecosystem and changing attitudes to natural and artificial regeneration.

CIRAD-Forêt has a network of plantation projects in Africa, Latin America and Asia

With important exceptions (e.g. the INEAC Forest Division in Zaire, created in 1934 and in operation in Yangambi and Luki until 1960), organized silvicultural research did not really mature until after the Second World War, notably in the regions under British and French colonial rule.

During the past 50 years, tropical research and silviculture has oscillated between two seemingly contradictory concepts: artificial regeneration (e.g. enrichment and plantations) and natural regeneration and improvement of existing stands.

Initially, forest services undertook to enrich the forest through planting. The main idea was to replace valuable species that had been logged, thereby providing material for future harvesting. Additional benefits were expected from pre-existing trees and natural regeneration. The first operations were conservative and concentrated on species with a demonstrated market value, such as Tarrietia utilis (niangon), Entandrophragma utile (sipo), E. cylindricum (sapelli), Khaya spp. (African mahogany), Aucoumea klaineana (Gaboon mahogany) and Chlorophora excelsa (iroko).

The first enrichment techniques consisted of minimal intervention planting along narrow alleys cut into forests with wide spacing and had little effect on the environment. This method was intended to provide about 50 first-class trees per hectare and depended on frequent and vigorous intervention. It was used almost everywhere in the tropics and gradually evolved from partial to total clearing of initial stands so as to give full light to dense plantings.

Enrichment methods required ongoing maintenance to give seedlings the best chance of survival. The technical nature of the requirements and the need for careful planning and significant labour inputs were rarely satisfied and the results were not very convincing.

The techniques of natural REGENERATION prevailed for about ten years, in particular between 1950 and 1960. Initial silvicultural attempts in natural forests were conducted at the same time as work on enrichment. There was a multiplicity of methods but the common principle was to boost the regeneration of commercial tree species by one of two approaches.

The first was to improve stands by thinning certain categories of trees, without trying to favour directly natural regeneration because the latter was more easily obtained with more homogeneous, treated stands. This approach was used in Zaire (where it was known as Uniforming from Above and Standardization) and Gabon (for Improvement of Gaboon Mahoganyrich Stands).

The second was to promote natural regeneration through cutting, clearing and maintenance to open up the overstorey and allow light to reach the ground to trigger the germination of seeds as well as to stimulate the growth of established seedlings suppressed by the forest canopy. These techniques included the Improvement of Natural Stands (in Côte d'Ivoire) and, particularly, the Tropical Shelterwood System developed in Southeast Asia (e.g. Malaysia) and applied in Africa (e.g. Ghana and Nigeria) and America (e.g. Trinidad and Tobago).

These natural regeneration practices were eventually abandoned because of problems with the proliferation of sun-seeking creepers, the growth of vigorous pioneers which blocked the development of valuable species and the difficulty of finding and maintaining the right balance of light. These difficulties were compounded by the many interventions widely spread through time and were difficult to justify technically and economically. This last aspect put a stop to stand improvement. Not only was it impossible to see the results immediately but also the gains in yield obtained were not quantified by substantial prior or accompanying research. By the beginning of the 1960s, the scales tipped in favour of artificial methods.

Indeed, advocates of artificial regeneration could claim higher efficiency and a better use of funds by concentrating field operations over less time and allowing systematic - thus stricter and easier-control. The deliberate choice of the field of action and the species was also a positive factor. The intention was to create a new forest, replacing the existing stand with one of a more even structure by planting one or two dominant species. The focus on dense plantings did not lower the unit cost but did make it possible to concentrate work over time and space and to mechanize much of the work.

As a rule, however, enrichment and natural regeneration as initially conceived did not fulfil the hopes of technicians and managers. Impoverished natural stands could not be naturally regenerated but had to be converted into plantations, and dense plantings were popular. Thus, in the 1960s, single-species plantations on bare ground became the main objective of forest managers and researchers in the tropics. At the same time, there was an increasing mechanization of clearing and maintenance work while the use of fast-growing species, such as pine or eucalyptus, for the production of pulpwood became popular after an experimental phase.

It is noteworthy that the past ten years have seen a renewed interest in natural forests management. Indeed, plantations have been questioned because of their prohibitive cost for many countries, their lower than expected yields in terms of both quality and quantity and on the grounds of biodiversity conservation. "Artificial" and "natural" techniques should not be seen as conflicting but rather as complementary methods. They must both be used and adapted to a new context; that is, the urgent need to protect and manage the tropical forest ecosystem sustainably. To demonstrate the potentially complementary nature of these techniques, the main results of silvicultural research by the former CTFT and its partners in tropical moist forests, both in natural forests and plantations, are presented below.

Plantation silviculture

The technical and economic difficulties of enrichment work led silviculturists to focus on plantation techniques based on the complete removal of the existing forest. This choice was directly inspired by experience gained with light-demanding species such as Tectona grandis (teak), Aucoumea klaineana (Gaboon mahogany) and Terminalia superba (limba). Most techniques currently recommended by CIRAD-Forêt for establishing and managing artificial stands are based on work done by the CTFT in Gabon and, above all, in Côte d'Ivoire.

Choice of species

Species suitable for tropical moist forest plantations may be classified into three groups:

· Species with a slow initial growth, including Entandrophragma utile (sipo), Khaya spp. (African mahogany) and Tarrietia utilis (niangon). Today, these are rarely used because of many silvicultural, phytopathological and economic constraints.
· Species that have a rapid initial growth and are used for industrial biomass production (e.g. pulp, construction wood, firewood). These include pines, eucalypts and acacias, which can be harvested early (between seven and ten years) and can be planted in degraded forest and savannahs.
· Species that grow well in monospecific (or mixed) stands destined for timber production. These species form the main subject of research on the management of tropical moist forests. They include the previously mentioned teak, limba and Gaboon mahogany as well as Cedrela odorata, Gmelina arborea, Terminalia ivorensis (framiré) and Triplochiton scleroxylon (samba).

Establishment and management of stands

The planting of any species, whether exotic or indigenous, must take into consideration specific, limiting ecological factors (e.g. soil and climate). Some requirements are specific to each species but a few general silvicultural principles can be stated.

Young seedlings are very quickly threatened by self-propagating, sun-seeking plants. Maintenance (manual, mechanical or chemical) must be prompt and regular to stop the canopy closing again. Artificial pruning improves the quality of the final product (ensuring small and healthy knots).

The increasing areas established annually as plantations and the growing shortage of labour have led to a greater use of mechanical site preparation, especially for clearing (chainsaws and bulldozers equipped with land-clearing rakes) and maintenance. However, mechanization requires level sites that are free of tree stumps and debris. Heavy equipment also considerably increases the cost of plantation establishment, but it has proved to be indispensable in large-scale projects (i.e. those establishing more than 500 ha per year), especially where climatic conditions restrict the planting season. It also makes it possible to mechanize maintenance and thus improve the survival and initial growth of young stands.

Planting techniques, from nursery preparation of seedlings to field maintenance during the first years, have been the subject of many trials and experiments. Suitable work programmes, including the scheduling and conduct of operations, have been established for several species. The initial plantation density must be high enough for early canopy closure and depends on the architecture and growth pattern of each species. Suitable stocking numbers may range from 1500 to 2000 stems per hectare for teak, 1100 stems for Gaboon mahogany, Cedrela sp. and Gmelina sp. and as low as 700 stems per hectare for framiré and limba. High initial stocking enables rapid occupancy of the site, but a thinning schedule must be maintained to achieve maximum growth per unit area.

The extensive investments impose a judicious choice of the vegetative material to be propagated. The genetic improvement of forest species starts with an analysis of genetic variability, choice of the best provenances, phenotypic selection and the creation of seed nurseries for the propagation of quality vegetative material. Genetic improvement can be accelerated through vegetative multiplication by taking cuttings from high-performance clones. This technique has been mastered in Triplochiton scleroxylon (samba) and Gmelina arborea.

TABLE 1. Silvicultural regime and expected yields for three plantation species


Aucoumea klaineana

Tectona grandis

Terminalia ivorensis

Final density (stems/ha)




Diameter of logs (cm)




Rotation age (years)




Basal area (m²/ha)




Bole volume (m³/ha)




Productivity (m³/ha/year)




Bole volume production (which approximates timber volume) depends on stand characteristics that can be controlled by the forest manager. The choice of cutting diameter has a strong influence on the nature and yield of forest stands and should be determined by technical, silvicultural and financial constraints. Present knowledge allows forecasts of production under prescribed silvicultural regimes, average fertility and well-managed plantations (Table 1).

Role and suitability of plantations

To conclude the subject of artificial regeneration, it can be stated that suitable plantation establishment techniques have been documented for many species. Enrichment methods provided useful results, both as underplanting and planting on bare ground, but socio-economic (e.g. available labour) and technical constraints (difficulties of scaling up to large areas) have pushed them aside in favour of intensive mechanized methods. These were justified by increasingly impoverished natural forests, by the technical possibilities of large-scale activities and by the insufficient knowledge of silviculture and natural forest management. The question is whether reforestation is a financially viable option for indebted countries with limited resources.

The high cost of plantations and their low profitability (according to traditional, short-term economics) are disadvantages, but they only apply in the case of purely speculative economic options. The forest, its management, its protection and its reconstitution by plantation are long-term concepts which are financially unappealing, at least given the current market prices of final products. The only way to make the forest financially viable is to integrate market-driven price increases with unquantified sociological and economic advantages: creation of employment, conservation of biodiversity, protection of the environment and the climate (water production, buffering CO2 levels, etc.). Moreover, because of the extent of irresponsible deforestation in tropical areas, a plantation often constitutes the only forest alternative, even if the new forest forms a simplified ecosystem.

Silviculture in natural forests

The initial research undertaken in natural tropical moist forest was scattered. The means were never sufficient to achieve the goals, and research and application were often confused. Experimental plots (almost always too small) were established independently in most tropical forest regions, without common guidelines or a similar model, so it was not possible to interpret or compare the collected data. In addition, imaginative efforts often suffered from successive changes in financing and forest policy. Thus, when concern about management of tropical forests increased in the mid-1970s, much of the information on natural forest silviculture was in the form of incomplete and uninterpreted data.

TABLE 2. Overview of the CTFT experiment network

Date work began


Main technical partners


Total plots

Total area(ha)

Treatment levels


Côte d'Ivoire

SODEFOR(Société Développement des plantations forestières)

La Tene



2 thinnings
2 thinnings
1 thinning or logging


Central African Republic

National Forest Service - Bangui




Logging ± thinning


French Guiana

INRA (National Institute of Agricultural Research)




2 loggings ± thinnings



INPA(National Research Institute of Amazonia)




3 loggings



National Forestry Commission- Libreville




4 thinnings (selective)



SNR/OCF(Congolese Forest Service)

N'gouha II



1 thinning



AFRD (Forest Ministry) Pt Inhutani I




2 leggings + thinnings

In 1976, the CTFT began establishing a network of experimental plots, based on a series of simple principles:

· only use large plots (several hectares) with as many comparable ones as possible in other places;
· measure simple parameters (e.g. circumference, tree location);
· use statistical means to interpret data.

At the same time, the main objectives focused on the study of:

· stand dynamics and tree growth in relation to simple silvicultural activities (e.g. logging), with attention paid mainly to primary stands;
· regeneration and the influence of silvicultural treatments;
· silvicultural treatments to increase timber yields;
· the development and transfer of technical concepts from experiments to large-scale trials for demonstration and forest management projects.

The experiment network now spans three continents and features national and international exchanges (Table 2) which supply a considerable and steady flow of information. The following sections concentrate on conclusions that can be drawn from growth, productivity, vulnerability, mortality and regeneration data.

Impact of logging

In the Central African Republic, in a forest with a total standing volume of just over 300 m³ per hectare (counting trees of more than 10 cm in diameter), harvesting removed three to four trees of over 80 cm dbh, representing 50 to 65 m³ per hectare. The inclusion of logging damage increased removals to 68 to 95 m³ per hectare, or 20 to 30 percent of the initial volume. In Guyana and Brazil, where the standing volume may be 310 to 370 m³ per hectare, the logging of about ten trees of 50 cm in diameter means a harvest of about 50 m³ per hectare and a total removal of 75 m' per hectare or 20 to 25 percent of the initial volume. Thus, the direct impact of logging is considerable. Disturbance should be controlled so as not to exceed the threshold of irreversible deterioration (a priori, 40 percent of the standing volume). This is especially important, as mortality may be abnormally high (2 to 4 percent per year) during the two to three years after logging. Opening of the canopy during logging may stimulate diameter increment but this affects relatively few trees because of the heterogeneous nature of logging activities.

Effect of thinning

Thinning was accomplished by devitalizing (with or without arboricides) noncommercial or ´'secondary" species, which were left standing to reduce cost and damage. Only trees in the higher stratum were devitalized, since these exert the most competition for future crop trees. Logging removes valuable tall trees and thinning gets rid of tall trees that have no commercial value, so the result is to create "younger" stands and stimulate their growth while preserving biodiversity (only large trees of all species are affected).

The opening of stands by thinning (and harvesting) favours the growth of most tree species, especially for small and medium-sized trees. The first results obtained in Côte d'Ivoire were very promising and were confirmed not only by other experiments in a similar environment in Africa but also by those in America in a very different type of forest. In Guyana, for instance, the growth rate of commercial species is increased by 75 percent for stems of 10 to 25 cm in diameter and by 50 percent for 25 to 40 cm trees during the four years after logging. But taller trees subject to less competition may show no response to treatment.

Natural mortality

Natural mortality is an important mechanism of forest regulation and regeneration. It is an extensive phenomenon but is difficult to quantify and study. It concerns only a few trees per hectare each year (1 to 2 percent of the total) but the overall loss in volume may be considerable when large stems are concerned. No connection could be established between natural mortality and the intensity of stand opening: trees die whatever their size and whatever their growth, in both treated and control plots. It is a discontinuous phenomenon (contrary to the growth in diameter and regenerative growth) with exceptional occurrences and it requires very long periods of observation.

Natural regeneration

Several experiments addressed stems of 2 to 10 cm in diameter. This research is far from conclusive but suggests that treatment stimulates the regeneration of most species (major and secondary) without causing major floristic modifications outside the areas affected by logging. Experiments were unable to establish a link between regeneration and the presence of the species (seed source) in the over-storey. Estimates in this field will remain provisional for a long time, since a long period is necessary to encompass the successive effects of a single action.


A fire occurred in 1983 in one of the three projects in Côte d'Ivoire. This accident was used to quantify the scope and the consequences of the damage. On the whole, the areas most damaged by fire were those which had been thinned and exploited and which, because of the abundance of dry standing and fallen trees, were more flammable than the untouched plots. This discovery is not very encouraging for forests that have been modified by humans, but it emphasizes the importance of forest protection as an "everyday matter". Good silviculture is not sufficient; attention must be paid daily to the preservation of investments and capital.


Annual volume production (integrating growth, mortality and regeneration) of all commercial species (over 10 cm in diameter) depends on the treatment used. Figures obtained in Africa, for example, show that production doubles during the four years after treatment. As the initial volume of standing trees was 100 to 150 m³ per hectare, this represents an annual productivity of about 2 to 3.5 percent. Studies in the Central African Republic show similar results despite differences in stand structure, species and treatment (Table 3).

The volume (or basal area) growth of trees above the cutting limit is more variable. In Côte d'Ivoire, for instance, the gain in production obtained by thinning or logging turns out to be relatively low, between 0.5 and 1.5 m³ per hectare per year. This may be explained by results stated previously:

· the growth of dominant trees is almost independent of treatment;
· mortality is very variable and may be unrelated to silvicultural activities;
· the gain in volume by recruiting to a commercial size is clearly influenced by the treatment and its intensity. Average-sized trees benefit most from the elimination of competitive trees.

The loss of commercial volume through natural mortality is about the same magnitude as the volume growth of commercial trees. It is the new trees reaching commercial size (60 cm dbh) that are affected by treatment and provide the gain in production quantified here. Volume growth is practically non-existent within untouched forests, so thinning and the mobilization of resources are extremely important in productive forest management.

Silviculture for sustainable management

Is the conservation of tropical forests a myth? Some 15.4 million ha are deforested every year! The "developmentalist" and "conservationist" approaches confront each other: each has limits and constraints. After harvesting, the silviculturist must use the various techniques made available through research to reconstitute the forest resource. The choice of technique will depend, on the one hand, on the management objectives and, on the other, on the constraints inherent in the stands: the remaining trees, the capacity for silvicultural treatment, susceptibility to fire, competing demands for land, etc.

Three conditions for success

If natural forests are to be conserved, three (non-silvicultural) conditions must be fulfilled as a point of departure before the technical activities of forest management are undertaken:

· at the governmental level, land management plans (supported by legislation that is applied and respected) must prescribe management of stable, demarcated forested areas;
· at the local level, the land allotted to forest should integrate the needs of local people and offer assistance for the development of farmland in the areas surrounding the forest;
· at the resource level, the logger and the silviculturist must work as partners with the understanding that harvesting is part of silvicultural treatment but, at the same time, ensuring that immediate operations and long-term goals are commercially realistic.

Silvicultural procedure

Assuming that the three above-mentioned conditions are achieved (an admittedly significant assumption), the first necessary condition from a silvicultural point of view is that harvesting should not exceed production. Inventory and growth modelling techniques provide reliable estimates of productivity and enable informed management. Ideally, individual trees could be identified, quantified and auctioned as standing trees. This approach is used in temperate areas but is far from operational in the tropics, even though it corresponds to the basic concept of the intrinsic value of the standing tree.

The second necessary condition is to plan how trees can be logged. It is obvious that forest harvesting affects the environment, but it also constitutes a silvicultural operation. Properly planned and controlled logging systems are preferable in every aspect: economic, ecological and silvicultural.

The third necessary condition is to enforce the appropriate post-logging measures to stimulate the growth of commercial species by selective thinning of dominant secondary species while being careful to maintain natural biodiversity and sufficient regeneration.

TABLE 3. Annual volume increase (over 10 cm diameter) of commercial species four years after treatment in two parts of Africa


Côte d'Ivoire

Central African Republic

m³ per hectare per year

Percentage of initial volume

m³ per hectare per year

Percentage of initial volume










2 2.1

Logged and thinned

2 2-3.6




These three technical conditions are justified in a management plan where the standing tree capital is sufficient to maintain commercial potential but, if the forest is too impoverished for natural regeneration, then plantation establishment becomes justified. The choice of plantation type must take into account the ecological vulnerability and, depending on production goals. the intensity of establishment methods may vary: low for extensive manual methods (enrichment); average for intensive manual methods (taungya, agroforestry systems); and high for mechanized methods.


This overview should give the reader an idea of the silvicultural knowledge and experimentation necessary for tropical moist forest management. One of the main conditions for silviculture's success is continuity-overtime: silvicultural trials should not be judged prematurely (e.g. an apparent failure of enrichment may be due to the premature interruption of maintenance) and should remain durable despite changes in the fashions of forest management.

Research results are sufficient to support large-scale plantation activities (even though there remains scope for further utilization research). Research on the dynamics and silviculture of natural forest stands has provided effective practical tools that can now be used by forest managers. However, the level of knowledge is still modest and concepts such as natural mortality and the behaviour of many species are still only partially known.

Finally, it must be emphasized that experimental results obtained from medium-sized areas raise the problem of extrapolation to larger areas, hence the necessity to associate a phase of financially well-established research to any forest management operation in full-sized forests.


Bertault, J.G. 1992. Etude de l'effet du feu en forêt semi-décidue de Côte d'Ivoire au sein d'un dispositif d'expérimentation sylvicole Thèse Université de Nancy 1.

Brunck, F., Grison, F. & Maître, H.F. 1990. L'Okoumé Aucoumea klaineana (Pierre). Côte d'Ivoire, CTFT. (monograph)

Coic, A., Viera, G. & Minette, L. 1990. Dégâts causes par l'exploitation forestière sur le dispositif ZF2. Proc. Cayenne Workshop. CTFT/INPA, Manaus, Brazil.

FAO. 1989. Management of tropical moist forests in Africa. FAO Forestry Paper No. 88. Rome.

FAO. 1993. Timber plantations in the humid tropics of Africa. FAO Forestry Paper No. 98. Rome.

FAO. 1994 Management and conservation of closed forests in tropical America. FAO Forestry Paper No. 101. Rome

Maître, H.F. 1990. Recherches sur la dynamique de peuplements arborés en vue de définir une sylviculture assurant la conservation et la protection durable de l'écosystème forestier tropical humide. In Proc. 19th Congr. of IUFRO, Vol. 2 Montreal, August 1990. Vienna, IUFRO.

Mengin-Lecreulx, P. 1990. Simulation de la croissance d'un peuplement de la forêt dense: le cas de la forêt de Yapo (Côte d'Ivoire). Côte d'Ivoire, CTFT/SODEFOR.

Reviere, L. 1992. Etude de l'évolution des peuplements naturels d'Okoumé, Aucoumea klaineana, dans le Sud-Estuaire du Gabon. These Université de Paris, 6.

Schmitt, L. & Bariteau, M. 1990. Gestion de l'écosystème forestier guyanais Etude de la croissance et de la regeneration naturelle (dispositif de Paracou). Bois et Forêts des Tropiques, 220: 3-23.

Tran-Hoang, A., Facrichon, V. & Maître, H.F. 1991. Dispositif d'étude de l'évolution de la forêt dense centrafricaine suivant différents types d'intervention sylvicole. Presentation des principaux résultats après huit années d'expérimentation. Bangui. Central African Republic, Projet ARRF.

Previous Page Top of Page Next Page