Previous Page Table of Contents Next Page

Tropical rain forest management

R. Schmidt

R. Schmidt is Forestry Officer, Forest Resources Division. FAO Rome.

In recent years, FAO's Forest Resources Division has undertaken a series of studies on the management of humid tropical forests. Forestry Papers 53 and 55 presented case-studies of management systems in India (Kerala), Ghana, Honduras and Trinidad. Two forestry papers in preparation will review management systems for tropical humid forests in Asia and Africa, and a subsequent review of systems in Latin America is planned.

This article reviews management systems in selected areas of closed broad-leaved tropical forests. It draws heavily on Malaysian studies by Thang Hooi Chiew and Abdul Rashid him Mat Amin of the Forestry Department of Malaysia, on those in the Philippines by the Philippines Bureau of Forest Development (Leslie, 1985) and on two recent papers on Sarawak (Hutchinson, 1986a, b). Comprehensive studies of humid tropical forest management for French- and English-speaking countries in Africa have also been prepared (Catinot, 1965, 1986; Philip, 1986).

TIMBER FELLING IN MALAYSIA (left) clearing lowland forests for agriculture / R. PRADO

A LOGGING ROAD IN KALIMANTAN, INDONESIA (opposite) shifting cultivation often follows logging / C.COSSALTER

The potential for sustained management of natural forests in the humid tropics continues to be a subject of concern and uncertainty among tropical foresters. In this article, adapted from a paper presented at the Unesco-lVIC International Workshop on Rain Forest Regeneration and Management, Guri, Venezuela, in November 1986, the author examines a series of recent FAO studies and project reports that indicate that the productive management of many humid lowland forests is both technically feasible and economically viable.

A review of these studies and other literature (Mengin-Lecreuix and Maître, 1986) constitutes the section on Africa. Finally, many FAO field project reports concerning the management of humid tropical forests were reviewed.

The article considers "natural" forest management in a restricted sense: controlled and regulated harvesting, combined with silvicultural and protective measures, to sustain or increase the commercial value of subsequent stands, all relying on natural regeneration of native species.

In a model management programme, the negative ecological impacts of harvesting or alternative land-uses can be minimized; thus overall operation is productive and profitable while the essential ecological character of the forest is maintained.

· The destruction of approximately 7.5 million ha annually of tropical rain forest, plus about 4 million ha of open and savannah woodlands, has emphasized the importance of managing tropical rain forests (FAO/UNEP, 1981 a, b, c). Such destruction continues unabated, amounting to more than 19000 ha every day. The virtual disappearance of commercially productive tropical rain forests is imminent in some countries; in others, the process is developing at a slower pace, usually because large areas of forest are inaccessible.

Managing tropical forests for economic production is a key element toward their conservation. Otherwise, four main options for land-use remain: tree plantations, protected natural forests, degraded and depleted forests, and non-forest uses. Plantations yield great benefits but cannot replace the functions of current natural forest areas. Thus, plantation and natural forest management are not competitive but complementary. They provide different types of products and are suited to different terrains.

Economically unproductive forests are often regarded as reserves of unutilized land. It is clearly appropriate to protect them in many areas, but this will become increasingly difficult and expensive where short-term productive potential for agriculture exists and where the rural poor have no alternatives for food production.

Rain forest management in Asia

MALAYSIA Significant efforts to manage natural tropical forests have been made in Malaysia, where it was found that when marketable trees were cut, either commercial species regenerated or existing regeneration developed rapidly. While this did not always happen, it did occur under sufficiently different circumstances to be encouraging. Although few studies on humid tropical forest outside Malaysia have clearly confirmed this finding (Leslie, 1985), there is little to indicate that it would not be true in other areas with appropriate harvesting, and silvicultural knowledge appears to be adequate to commence management operations in many humid forests throughout the tropics.

The Malayan Uniform System (MUS), developed after the Second World War, converts virgin tropical lowland rain forest (a rich, complex, multispecies, multi-aged forest) to a more or less even-aged forest containing a greater proportion of commercial species. This transformation is achieved by a clear-felling release of selected natural regeneration of varying age, aided by the systematic poisoning of unwanted species (Wyatt-Smith, 1963).

Five important factors in this system are relevant to any humid tropical forest natural management operation:

· the stocking of regeneration must be adequate;
· the original partially harvested canopy must be removed;
· there must be no tending until regrowth has passed the ephemeral climber stage;
· an adequate new canopy must be maintained to prevent the redevelopment of climbers;
· linear sampling must assess the regeneration status.

Although successful in lowland forest areas, the MUS was judged less successful in the dipterocarp hill forests. Contributing factors included rough and variable terrain, consequent uneven stocking and variable regeneration, heavier damage to residual stands during steep-slope logging, and irregular seeding of the principal commercial species.

The modified MUS as practiced in Peninsular Malaysia and Sabah introduced certain refinements to the original MUS for specific sites and circumstances, and emphasized post-harvest sampling to determine appropriate silvicultural treatment (see Table 1).

Recently, the Selective Management System (SMS) has been developed in Peninsular Malaysia (Salleh and Baharudin, 1985). This system advocates flexible management options based on a pre-harvesting inventory to determine diameter limits and species selection for harvesting. Climber cuttings prior to harvest and marking for directional harvest attempt to minimize logging damage to residual stands. The key step of post-harvest treatment is left undefined.

SMS is not a true selection system according to standard silvicultural terminology where single stems or very small groups of trees are removed as they reach maturity on a more or less constant (polycyclic) basis. Truly polycyclic systems have not functioned success fully on a wide scale anywhere in the humid tropics. It is therefore perhaps more accurate to describe this as a system which leaves the manager with wide discretionary powers to determine where silvicultural treatment will be most advantageous from a cost/benefit standpoint. Silvicultural treatment continues in Peninsular Malaysia - poison girdling of noncommercial species was carried out on more than 62 000 ha in 1982 (Salleh and Baharudin, 1985).

In Sabah, a long history of exploitative logging was countered after 1971 by the application of the Modified Malayan Uniform System to 140 995 ha of logged-over dipterocarp hill forests (see Table 1). By 1978, poison girdling was halted because increased logging had opened up the stands to such an extent that additional opening was considered counterproductive. The cost-effectiveness of such treatment was uncertain, and market acceptability was changing so quickly that poisoning undesirables became a questionable operation (Chai and Udarbe, 1977).

Table 1 Sequence of operations of the Modified Malayan Uniform System, Sabah



n - 2 to

Allocation of coupe

n - 1

First silvicultural treatment - protective tree marking and climber cutting


Felling operation

n + 0-1 month

Clearance inspection

n + 0-2 month

Assessment of regeneration through Linear Sampling Milliacre (LSM); (2 m x 2 m plots)

n + 36 month

Second silvicultural treatment - first poison girdling of unwanted and defective trees, climber cutting if necessary

n + 10 to

Assessment of regeneration through Linear Sampling

n + 15

Half-Chain Survey (LS1/2); (10 m x 10 m plots) - third silvicultural treatment, liberation treatment where necessary

Source: Fox and Hepburn. 1972 Chai and Udarbe, 1977

The silvicultural system developed in Sarawak is currently one of the most consistently applied and successful. From 1974 to 1980, FAO projects assisted the Forest Department in developing and instituting silvicultural practices for the mixed dipterocarp hill forest (UNDP/FAO, 1982a; Hutchinson, 1981, 1986a, b). The conceptual and operational point of departure for this system is not undisturbed forest but the increasingly large areas of forest that have been selectively logged for valuable species.

In this selectively logged forest, liberation thinnings are performed. The concept is not new or even exclusively tropical: liberation thinnings are logically indicated in any situation where a young crop of potentially good trees is overtopped by older, distinctly less desirable trees. If the overtopped trees respond vigorously and speedily to form a new good-quality stand, quick and cheap silvicultural transformation to a productive stand is possible (Smith, 1962). To recover the cost of rain-forest treatment, the maximum rate of increment must be concentrated on what will prove to be the final crop trees (Baur, 1964).

Because a low proportion of commercially valuable species exists (a frequently cited constraint to tropical forest management), selective logging is often light. From 1974 to 1980, selective logging in Sarawak extracted 5-15 trees per ha, representing a volume of 10-50 m3, compared to total commercial volumes of 150 of 250 m3.

At these levels of extraction, 60 percent of Sarawak's logged tracts retained residual forest consisting of 20 percent undisturbed and 40 percent disturbed in some way by extraction. Both overstorey removal and liberation thinning were applied. Overstorey removal is cheap, but the increase in dbh increment is correspondingly modest. The distinguishing characteristic of liberation thinning is that it opens wells around trees selected individually to be potential final-crop trees. Thus trees are poisoned in localized patches throughout the forest, their numbers being inversely related to the minimum dbh specified for the selection of final-crop trees (Hutchinson, 1986b).

The detailed procedures of the system have been described by Wadsworth (1969). The essential principle is that trees are not systematically eliminated according to species and size across forest stands. There are very simple set rules, which are essentially distance tables with a few concrete modifications for poison girdling those trees that are judged to compete directly with final-crop trees. This conserves the forest and the foresters both ecologically and economically.

At the very least, the Asian experience shows that integrated management of the tropical mixed forest is technically feasible.

Table 2. Sustained yield management and residual forest activity data, the Philippines


Aren logged


Adequately stoked residual forests

Silviculturally treated






52 660


45 100

14 000


73 840


53 165

16 000


67 260


50 445

33 000


64 090


48 700

47 000


57 275


45 820

44 000


66 932


54 215

53 000


58 416


47 310

49 200


52 596


43 130

56 300


51 993


43 570

33 900

Source: L Leslie, 1985

Vines and climbers constitute a major component of tropical rain forests and have significant effects on tree growth. The systematic elimination of "undesirables" has often led to proliferation of climbers, inhibiting crop-tree development and forest operations. Equally frustrating to attempts at natural tropical rain forest silviculture has been the incredibly rapid growth of noncommercial pioneers or light-demanding species. Tropical forests have many more gradations of shade tolerance within different species and more species within broad categories of shade tolerance than temperate forests.

The choice of the type of improvement thinning needed should therefore be guided by the ecological groups into which the majority of the commercially valuable species fall. Extensive canopy opening favours light-demanding species, moderate opening favours the gap-opportunist species, while shade-tolerant species will be encouraged by minimum canopy disturbance, such as that provided by overstorey removal (Hutchinson, 1986a).

Countries attempting to manage tropical rain forests may be dealing with hundreds of tree species that reach millable size. It is necessary to group these species for commercial purposes and to simplify the real botanical complexity. In this effort, local people and their knowledge of the forest are often very useful.

The criteria for what constitutes a commercially valuable tree are changing constantly. The general trend throughout the tropics is that more and more species are being "discovered" as valuable. Some countries have regretted poison girdling of species that subsequently proved marketable.

However, relative values have seldom been reversed; more and more species have simply become economically acceptable. A tree should therefore be eliminated only if it is directly competing with a more valuable one.

Although there is much debate about what constitutes an adequately stocked residual stand (Salleh and Baharudin, 1985), it should never impede the implementation of management programmes. The average dbh increment in natural tropical forests varies with many factors, but is seldom greater than 1 cm per year and is often less. Thus, if crop trees were 10 cm average dbh, a minimum of 40 years might be expected to maturity. If larger trees can be successfully released, logging might occur every 25 or 30 years.

The complexity and variability of ecological dynamics and species composition in tropical rain forests means one thing for silviculture: flexibility, guided by commonsense acquired through field experience. Hutchinson (1986a) suggests examining seedlings and saplings present before and after logging; the impact of logging upon the forest, particularly regarding damage to surviving stems; and the pro vision of open space for regeneration. Findings will suggest the possibilities for a future crop and an appropriate silvicultural treatment.

THE PHILIPPINES The Philippines have extensive areas of productive tropical forests, and forestry institutions and educational systems are well established. The approach to management has been that property implemented selective logging leaves a residual stand that develops so that another commercial cut becomes possible in 30-45 years (UNDP/FAO, 1970a). An extensive and comprehensive inventory system has been developed which is quite similar in scope and approach to the national forest inventory in the United States. Permanently marked plots are selected at random within forested areas and periodically remeasured. Regeneration, growth and volume are monitored, and aerial photos are thoroughly evaluated to contribute data (Nillson, Marsch and Singh, 1978). Table 2 shows areas harvested and silviculturally treated.

From 1975 to 1981, a Philippine-German Timber Stand Improvement Project developed methods of post-logging treatment. The treatment prescribed was basically a selection of the best potential crop trees (Leslie, 1985). This confirms work done in Sarawak.

MARKING A TREE FOR HARVEST IN MALAYSIA management system relying on natural regeneration has been in use for decades/ R. PRADO

There are currently no large-scale sustained yield management programmes being implemented in the vast closed broad-leaved forests of tropical America. This is certainly not for lack of resources: in 1985, tropical America had an estimated 491.8 million ha of productive closed broad-leaved forests, of which 54.7 million ha have been logged over. The current absence of this type of management is certainly not because of lack of experimentation, recommendations and attempts at pilot demonstration programmes, even though more effective efforts need to be conceived and designed.

Rain forest management in tropical America

BRAZIL In 1978, the Brazilian Government and the FAO Forestry Department initiated a long-term, large-scale pilot project to manage the Tapajós National Forest in the Amazon basin for multiple uses, including timber production through natural silvicultural systems (UNDP/FAO, 1983b). The project was preceded by a series of inventories and experimental treatments conducted cooperatively by the Brazilian Government and FAO during the 1950s and 1960s. A 1976 review mission concluded that commercial operations might commence in 1978. The pre-feasibility study, including a sensitivity analysis, had shown that a 53-percent internal rate of return could be maintained at 20 percent even with increased operating costs and sharply reduced prices.

A complete management plan was prepared giving careful consideration to the recommendations of silvicultural and utilization specialists. The Forest Inventory in 1978 had found that the forest contained 54 m3 of roundwood per ha in stems greater than 45 cm dbh. Of this, there were 36 m3 of the 28 species comprising the valuable commercial trees. The experimental extraction of 64 ha in 1979 produced a total volume of 72 m3 roundwood per ha of which 64 m3 was commercial. The gross volume of all trees greater than 55 cm dbh was 132 m3/ha (UNDP/FAO, 1980 and 1983a).

The Project's Terminal Report (UNDP/FAO, 1983a) presents a comprehensive and intensive study in which 15 international specialists and 30 Brazilian scientists participated; the results indicated that it would be technically possible and economically viable to develop forest-based industries in selected areas of the Amazon simliar to the Tapajós National Forest while maintaining a permanent tree cover. The studies included substantial operational components. An evaluation consultancy reported that 27 000 m3 were felled in 1980, 25 000 m3 in 1981 and 17 000 m3 in 1982 (UNDP/FAO, 1983c).

The 1982 UNDP/FAO review mission noted the optimism and entrepreneurial opportunity that existed, but observed that no real forest management had yet commenced. It found that in order to process Tapajós material it was crucial to control outside procurement of logs (UNDP/FAO, 1982b). Existing mills in the area were being supplied at operational capacity, typically from land-clearing operations sponsored by the Instituto Nacional de Colonizacão e Reforma Agrária where logs were obtained virtually free at the roadside in return for clearing the land. The review mission concluded that it would still be possible to manage Tapajós profitably and that failure to initiate commercial harvesting by the middle of 1983 should be viewed with alarm.

By 1985, the project was concentrating its efforts on the dry northeastern part of Brazil. The Brazilian Government continues to maintain a study project in Tapajós, but management for industrial production is stalled for the moment.

PERU In 1971, the Peruvian Government requested UNDP funding for a demonstration project for forest management in the Alexander von Humboldt National Forest in the lowland Amazon basin. The project operated from 1974 to 1978. The long-term goals were to increase sustained yield of national forests through the demonstration of management, improvement, protection and utilization techniques; and obtain the maximum social and economic benefits through the development of an organized timber industry. The short-term objectives were to determine the technical and economic feasibility of developing a wood processing complex; study the regeneration of commercially valuable species; formulate a management plan for the von Humboldt National Forest; conduct feasibility studies; and train people at all levels (UNDP/FAO, 1979).

In a trial inventory carried out on 200 000 ha the forest was found to be very heterogeneous in floristic composition but quite homogeneous in volume actually commercialized (VAC). Although 300 tree species were included in the inventory, 28 of these made up 70 percent of the total. Three-quarters of the commercial volume was in 21 species, and the regeneration of 15 species comprised 85 percent of the trees. The VAC rose from 15 to 30 m3/ha during the term of the project, as 20 new species became commercially valuable. The economic analysis indicated that a total investment of US$26 million was necessary to operate the project; the internal rate of return was calculated at 12-17 percent and cash flow was always positive. A comprehensive management plan allocated different areas for agrosilviculture, plantations, natural production forest and protection forest.


Silviculturally, the production forest was to be managed on a 60 year rotation with a 30-year cutting cycle. However, when the project terminated, momentum was lost, and management activities were not implemented as foreseen. Landless poor occupied many cutover sites, including some research plots, and initiated largely unsustainable agricultural practices.

In the Peruvian Amazon, the Japanese Government is at present involved in a cooperative project of plantations and natural forest regeneration, and the Universidad Nacional Agraria La Molina is carrying out research in tropical silviculture (Romero Mejía, 1983). Meanwhile, official requirements for post-logging management activities are not currently enforced, and reported timber extractions in Peru rose from 474 205 m3 in 1977 to 643343 m3 in 1981.

Current FAO projects in the country are focused on the problems of deforestation, watershed management and fuelwood supply in the Andes.

COLOMBIA A similar project was carried out in Colombia between 1965 and 1970 in the Serranía San Lucas, a moist tropical lowland forest between the Magdalena and Cauca rivers. This area consisted of 1.2 million ha in 1965 and 1.0 million in 1970. Inventories indicated 114 m3/ha of total stem volume and 33m3/ha of commercial volume. Noting that only long-term studies could serve as the basis for effective silvicultural practice in these forests, the project staff were nevertheless convinced that judicious cutting would produce secondary forests that could be managed for equal or greater commercial volumes than the undisturbed forest. Stocking and regeneration were adequate, and the forest began to develop satisfactorily after sample cuts (UNDP/FAO, 1970b). However, the project site was abandoned in 1970 for security reasons.

The extremely wet Pacific coast of Colombia provides several lessons on tropical forest management. Silviculture and management of the Bosque die Guandal was analysed during a UNDP/FAO project (Neyra-Román, 1979). In Narino, it was reported that 500 00.0 ha of this forest exist, dominated by two species (Dialyanthera gracilipes and Campnosperma panamensis). A sound management programme has not developed in these forests, although the species structure would technically make this relatively easy. The enterprise Carton de Colombia utilized the natural wet forests on the Colombian coast for pulp. Regeneration was vigorous, approaching the original biomass within 15 years, but because local farmers utilized the developing secondary forests for a number of products, Carton de Colombia has not attempted subsequent management of the resources.

The complexity and variability of ecological dynamics and species composition in tropical rain forests means one thing for silviculture: flexibility, guided by common-sense acquired through field experience.

ECUADOR The impact of colonización, the occupation and clearing of forest lands for family farms, has been a determining factor in forest management in Ecuador. Fifteen years ago there were an estimated 30 million ha of productive and well-stocked natural tropical forests in the country (Jankovic, 1971). Three million ha of undisturbed, easily accessible forest in the northwest could have been the basis of an increasingly integrated forest industry. A later project (UNDP/FAO, 1977) for strengthening the Ecuadorian Forest Service observed that 418600 ha had been occupied as farmland in the northeast (Amazon basin) from 1971 to 1975, with only about 5 percent receiving legal title. The farmers had utilized the petroleum industry roads. The wood extracted (42 000 56 000 m3 in 1976) supplied 32 sawmills but represented no more than 2 percent of the timber cut for land clearing. A UNDP/FAO (1982c) project publication concluded that the problem of colonization had endangered the entire forest industry, and recognized a wide range of additional problems, including the necessity of reforestation in zones where commercial timber cannot be grown. Current government requests for international forestry expertise have shifted to watershed management to extend the life of reservoirs.

SURINAME Natural silviculture in tropical rain forests in Suriname is discussed by Jonkers and Schmidt (1984) and Boxman et al. (1985). Ninety percent of the country is still forested, and 9 000 ha of Pinus caribaea plantations have not fulfilied expectations for economical wood production. It was found that while current felling and extraction were haphazard, rationally planned skid trails and felling techniques reduced damage to remaining trees and also reduced extraction costs. Poisoning commercial trees over 20 cm dbh and cutting lianas increased annual dbh increment from 0.4 cm to 1.0 cm. Refinement of 200 ha required 2.8 worker-days/he end 17 l of 2.5 percent 2,4, 5-Tin diesel oil. The studies predict that this treatment would result in the production of 40 m3/ha of harvestable volume in 20 years, with 13.5 trees/ha becoming commercial.

The experimental work in Suriname has been well conceived in that it has been carried out on a fairly large scale in conjunction with forest industry. The basic conclusion, which has been reached in many countries, is that the silvicultural problems are manageable. It is important to note that no established forest estate is supervised by the Suriname Forest Service, which has only a minimal influence over forest concessions (Wood, 1982).

COSTA RICA Costa Rica is facing a possible loss of all legally and physically accessible productive forest by the twenty-first century. A project ending in 1986 (UNDP/FAO, 1985b) identified a study area of 14 000 ha containing 8 000 ha of undisturbed forest where a sawmill with a 5000 m3/year capacity will be installed. In a 70-ha pilot operation area where inventories before and after harvesting were conducted, 10-12 ha produced 828 m3 in 34 species in 1985 before rain closed down operations. The latest information indicates that follow-up operations did not occur in 1986, but it is hoped that the study will be continued.

FRENCH GUIANA A silvicultural research project was carried out in the 50 100 km-wide strip of coastal forest (Maître, 1982). Commercial volumes and harvesting costs were calculated, and the economic implications of scenarios with varying levels of harvesting for timber or fuel-wood were presented. The highest internal rate of return (7.78 percent) resulted from a combined strategy of removing 1.3 m3/ha/yr in timber and 3.0 m3/ha/yr in fuelwood. This forest does not at present experience population pressure nor is it subject to pressing needs for new agricultural land. Timber extraction is increasing from a modest 120000 m3 in 1980. Sustained yield management could preclude the necessity and cost of building new roads into the interior undisturbed forests.

Other programmes No other largescale, commercial, sustained-yield management programmes are known to be in progress in tropical rain forests in America. Venezuela, for example, has not implemented programmes based on natural regeneration and true programmes of forest utilization 'aprovechamiento' in which natural regeneration is implicit, have never existed in Mexico (Gómez-Pompa, 1985).

Tropical forest management systems in Africa

Silviculturists in Nigeria and Côte d'Ivoire experimented with natural I regeneration and line planting during most of the first half of the twentieth century. Many other African forestry departments tried to take up the challenge of silviculture in moist forests beginning in the 1950s. Some of the methods relied on natural regeneration, others utilized techniques for improving the dynamics of the - stands and others used artificial regeneration.

Natural regeneration The three main methods based on natural regeneration were the Tropical Shelterwood System (TSS) in Nigeria; the Amelioration des peuplements of naturels (APN), or improving natural populations, in Côte d'Ivoire and the Selection System in Ghana. The TSS was designed in Nigeria in 1944 on the basis of 20 years of tests. Its objective was to enhance are the natural regeneration of valuable species before exploitation by, gradually opening up the canopy (poisoning undesirable trees, cutting climbers) to obtain at least 100 of 1-m high seedlings per ha over five years. The forest thus worked was logged over in the sixth year and cleaning and thinning operations were then carried out over 15 years. Two hundred thousand ha of forest were treated this way by the Nigerian Forestry Department between 1944 and 1966, after which the method was dropped. The main problems encountered were the exuberant spreading of climbers once the canopy had been opened up and the failure of the seedlings of valuable species to grow adequately. Moreover, some poisoned eliminated trees later turned out to be commercially valuable, e.g. Pycnanthus angolensis.

In 1950, the Forestry Department of Côte d'Ivoire found the initial results in Nigeria appealing and gave up line planting in favour of APN, a technique linked to TSS. Apart from technical considerations, there was an economic motive for this drastic change: the productive capacity of sawmills for domestic consumption was increasing. The geographic dispersal, as well as the widening of the range of species used, called for widespread operations and more species to be regenerated. The APN method was applied to forests that had been logged over and were well stocked in valuable trees of average size. The aim was to favour the growth of these average stems and also to ensure regeneration through natural seeding of the valuable species by removing climbers and opening up the canopy. The method was applied by the Forestry Department from 1950 to 1960 on large areas, but was then abandoned after results were disappointing.

The Selection System has been applied in Ghana since 1960. Its objective is to assure the regeneration of forests well stocked with valuable species. Harvesting occurs about every 15 years, after the Forestry Department has marked the stand to retain some well-distributed seed trees, followed by thinning operations. The method has been found to cause considerable felling damage because of the relatively short rotation. Regeneration has been poor, and less valuable shade-tolerant species dominate because of insufficient opening up of the canopy.

Improvement of stand dynamics This technique was utilized in the 1950s in Gabon's Aucoumea klaineana forest to accelerate the growth of all-sized stems of valuable species in naturally well-stocked stands, without trying specifically to pro- yoke regeneration through natural seeding. The species grew in patches or clumps, presumably as a result of natural seeding of forest trees in the clearings or gaps. The objective was to let these stands attain commercial diameters as quickly as possible through thinning operations, but the production gain was never measured. After treating about 1 000 000 ha in this way, the Forestry Department gave up the technique in 1962 to switch to Aucoumea klaineana plantations.

Belgian foresters also tried stand improvement techniques in the 1950s: uniformisation par le haut and normalisation. Several thousand hectares were managed through these techniques, mainly in the Mayumbe region (Low Zaire). After independence in 1960, these trials were not continued by the national Forestry Department.

Natural regeneration techniques are thus not practiced on a commercial scale in the francophone countries, nor in Nigeria. However, Ghana and Uganda have' in principle, continued both moist forest planting and management.

Forest harvesting Forest harvesting in Africa has evolved considerably since 1950. The forest lands logged over before 1930 are estimated to cover about 42 million ha of the 162 million ha of productive closed broad-leaved forest. The most accessible forests, such as those remaining in Côte d'Ivoire and Nigeria, have already been logged over several times. Approximately 90 percent of the undisturbed forests are found in Zaire, Gabon, the Congo and Cameroon.

Harvesting is still as selective as it was originally. The net volume of logs extracted from the forest averages from 5 to 35 m3/ha. The number of species utilized is still fairly limited, in densely populated countries such as Ghana (55 inhabitants per km2) and particularly Nigeria (more than 100 inhabitants per km2), where domestic markets have supplanted exports, this should normally lead to a diversification of utilized species. In other major wood-producing countries, however, such as Côte d'Ivoire Cameroon, Gabon and the Congo, 75 percent of the timber trade volume consists of exports, and the domestic market is not large enough to absorb the lesser utilized species.

A MATURE TERMINALIA IN COTE D'IVOIRE research indicates that management is justified / C. COSSALTER

Harvesting rules that existed before 1960 in anglophone countries (particularly in Ghana, Nigeria and Uganda) and in Zaire have been gradually abandoned in most countries because of insufficient staff and funds. When the efficiency of silvicultural techniques was questioned, more spectacular planting operations backed by external funding were chosen.

In the 1960s, an original system was set up in the Central African Republic: silvicultural operations were carried out within concessions, financed mostly by concessionaires. They consisted of marking saplings for protection during harvesting and post-harvesting thinning operations. Unfortunately, this procedure was discontinued after ten years because of institutional problems.

Forest management in francophone Africa has sometimes consisted of harvesting regulations that fix the terms and areas for concessions and a minimum exploitable diameter for economic species. Stumpage rates paid in harvesting contracts are theoretically earmarked in order to finance management and regeneration costs, but in practice the funds revert to the general state budget.

Although such measures are not complete management systems, they are a positive start. Les Unités forestières d'aménagement (UFAs) were outlined in the Congo in the 1960s with the assistance and advice of FAO. Areas for concessions were delineated, and inventories required to provide information on stocking levels. When an area is judged to be depleted, it is closed to harvesting for recuperation. Progress is being made toward the control of product flow and the development of the institutional structures and responsibilities that are prerequisites to further evolution of management programmes.

Experimental pilot management projects have been set up by FAO in Cameroon and Gabon but unfortunately project recommendations have never been put into general practice. Although silviculturally sound, management restrictions are sometimes resisted because they may specify the utilization of too many second category species that are judged unprofitable in areas of poor access; and they may specify progressive harvesting per block, requiring the utilization of all species on the annual allowable cutting area simultaneously and disallowing repeated harvesting so as to let regeneration occur. This may create insoluble marketing problems because of the fluctuation in the tropical timber market.

The SODEFOR-CTFT study on moist forest silviculture It has been difficult to measure accurately the productivity gains obtained by natural regeneration and stand improvement techniques. Uncertainty concerning the efficiency of the methods used has undoubtedly contributed to their progressive abandonment by most African countries. By contrast, the productivity and cost of forest plantations are relatively well known.

In 1976, an important plan of action for the study of moist forest development relative to different silvicultural interventions was set up in Côte d'Ivoire by the Société ivoirienne de développement des plantations forestières (SODEFOR) with the technical support of the Centre technique forestier tropical (CTFT). The advantage of this project, compared to previous experiments, was that it allowed accurate measurement of the impact of the silvicultural operations. The mensuration methods, size of trials and spatial replication of the treatments were designed to obtain reliable results.

The project covers 1200 ha and three field stations characteristic of three ecological areas of the moist forest of Côte d'Ivoire: semi-deciduous forest, evergreen forest and transition forest. The silvicultural practices used are the traditional exploitation of economic species and thinning by poison girdling. Two thinning regimes were tested (30 and 45 percent of the total basal area) by beginning systematically with the tallest trees in the residual forest until the desired percentage of basal area was reached. The objective of the thinnings was to favour valuable trees larger than 10 cm. No particular operation has been planned for regeneration through natural seeding.

After four years of observation, volume increment was 3-3.5 m3/ha/yr against 2 m3/ha/yr in the control stands, i.e. a gain in growth of 50-75 percent in stems of 73 main species larger than 10 cm dbh. Measurements taken every year showed that the volume increment increases with time and that the impact of the thinning operations will probably be felt for at least ten years. The largest yearly diameter increments, averaging 1 cm/yr, were found in species such as Triplochiton scleroxyla, Terminalia superba, Tarrietia utilis and Swietenia macrophylla.

A management system based on this practice would seem justified since the ratio of commercial volume to cost is greater than that of forest plantations set up in the same area. Over 30 years, the ratios for both systems may be estimated as follows:

- management of the natural forest:

about 1 m3 produced for every US$5.6 invested. (This assumes that untreated natural forest would produce 60 m3 and treated forest 85 m3 of commercial volume.)

- plantations:

about 1 m3 produced for every US$7.4 invested.

It must be remembered that natural forest investments are usually compounded for a longer period, but the figures are indicative of the quantities involved. If a wider range of species from the natural forest can be introduced in local markets, the economics of the operation improves.

The 10 000 ha Yapo forest is now managed experimentally on the basis of the results of the silvicultural research project set up by SODEFOR-CTFT. Other management programmes will soon be undertaken in progressively larger areas eventually to cover all the forest lands of Côte d'Ivoire.

A comparison of Asia, America and Africa

It is often dangerous to extrapolate findings beyond the context from which they were derived, and this could be the case in applying the experience from Asia too rigidly to other tropical forest regions. The ecological and socio-economic differences between the regions are quite marked, and it is most unlikely that specific practices will have much direct application from one region to the other. However, to assume that nothing can be transferred from one region to another would be equally unconstructive. The Asian experience shows that sustainable, integrated management of the tropical mixed forest is technically feasible.

A common feature of discontinued natural forest management programmes in tropical America is that technical feasibility is never cited as the reason for the discontinuation. In Colombia, the project ceased operations because security in the area broke down. In Peru and Ecuador, the familiar pattern of landless farmers spontaneously occupying forest lands where roads had penetrated brought the management scheme to an end. In the Brazilian Amazon, the abundance of old growth timber resources, combined with publicly subsidized land clearing, saturated the existing conversion capacity, resulting in negative stumpage prices which eliminated the possibility of any economic management.

EXPORTING LOGS IN PAPUA NEW GUINEA forest resources are an important source of revenue / R. PRADO

In Africa, political developments have often meant that institutions dealing with forest management have been given new directions, and new techniques, promising more rapid results than those achieved through the relatively slow growth of natural forests, have been preferred. As these promises have failed to materialize there is some renewed interest in managing natural forests. However, many institutions and national budgets do not have the strength to initiate new programmes.

These are not problems of species diversity, lack of understanding of ecosystem dynamics, inability to retain adequate regeneration or lack of response to silvicultural treatment. The problems concern land-use policy, socio-economic conditions and political realities. A large part of the problem is that the productive potential of abundant resources is undervalued. Certainly this article is not the first to point out that viable forest industries based on sustained yield could provide long-term economic well-being for more people than at present, by opening up new areas of forest for unsustainable, short-term agricultural exploitation. The problem is to expand the time horizons of policy-makers. The tropical world already offers many examples of the consequences of continuing to neglect sustainable resource management policies. Once productive capacity has been reduced, efforts to restore it become very expensive.

Problems and opportunities

Although more information would be useful in areas such as growth and yield statistics, annual increment and sustainable removals (FAO, 1984), there is now enough to implement a sustainable natural system of management. Any claim that the silvicultural and yield regulatory elements, at least in Asia, are the limiting factors to the advancement of forest management would be hard to sustain.

National leadership at the highest level must be committed to current investment for economic benefits that will be realized 20 to 30 years in the future. National policy must ensure that population development is in harmony with the optimum productive capacity of land available. Institutions responsible for management must have long-term programme stability, with stable leadership in key positions to provide continuity. Effective training institutions must prepare workers from the technician to the doctoral research level, and institutional mechanisms must place knowledgeable and competent personnel in the field where management activities occur. Profitable management must be integrated with the national economy and the world timber market, and present plans must assess the demand for products 20 years or more in the future. Effective legislation and national land-use planning that identify the forested areas to be managed are indispensable.

Table 3. Largest national areas of productive closed forests 1985 forests,1 1985

The slivicultural system developed in Sarawak is currently one of the most consistently applied and successful.



Logged over

(million ha)

(million ha)





































Papua New Guinea


















French Guiana











0 5

Viet Nam



Central African Rep.
























Côte d'Ivoire












Sri Lanka



1Defined as forests that cover... "a high Portion of the ground and do not have a continuous dense grass layer...; their characteristics... allow (or might allow) for the production of wood for industry." "Logged over once or more times during the last 60 to 80 years." (FAO/UNEP, 1981)

It is difficult to commit scarce financial and human resources to unproven technical packages. There is therefore an urgent need to demonstrate feasibility through successful pilot management programmes. However, these cannot operate in isolation, for they depend on the institutional strength and policy commitment referred to above.

Future action

It would be foolish to maintain that the current management picture in moist tropical forests is an encouraging one. In the three great centres - the Amazon basin, central Africa and the islands of Southeast Asia - substantial areas of forest are being cut over or converted to other uses; significant programmes of silvicultural treatment are occurring only in Malaysia. There are several countries where the reduction of productive moist forest areas has severely limited the former potential for extensive natural forest management (Costa Rica, Nigeria, and Viet Nam are representative and diverse examples). However, Table 3 gives a list of 36 countries with at least 1 000 000 ha of tropical forest suitable for productive management. They represent more than 90 percent of the world total.

Achieving sound management will not be easy, but failure will result in the loss of the great majority of tropical rain forests. There is every indication that economically unproductive areas in tropical countries will continue to be highly vulnerable to development or conversion, even if such development is unsustainable. There is no economically profitable alternative use for large areas of biologically highly productive tropical forests.

Effective management programmes may draw on international expertise, cooperation and financial assistance, but they will be carried out by national governments and institutions. Action must be coordinated and comprehensive both at a national and, if international assistance is to be used, at international level.

The Tropical Forestry Action Plan, fully described elsewhere (FAO, 1985a, 1986b), has a focus well suited to natural forest management in that it seeks to coordinate international assistance for priority areas based on a comprehensive analysis of each particular national situation. These are exactly the measures that are needed. The productive management of natural tropical forests, with the social and ecological benefits these forests can produce, is not the only necessary activity, but it is the key to combating irrational deforestation.


BAUR, G. 1964 The ecological basis for rain forest management. Sydney Forestry Commission NSW. 499 p.

BOXMAN, O. et al. 1985 Towards sustained timber production from tropical rain forests in Suriname. Neth. J. Agric. Sci., 33: 125-132.

CATINOT, R. 1965 Sylviculture tropicale en forêt dense africaine. Bois et forets des tropiques, 100-104.

CATINOT, R. 1986 Etudes sur les systèmes d'aménagement dans les forêts tropicales mixtes d'Afrique francophone. Rome, FAO. (Sous presse)

CHAI, D.N.P. & UDARBE, M.P. 1977 The effectiveness of current silvicultural practice in Sabah. Mal. For., 40 (1): 27-35.

FAO. 1984 Intensive multiple-use forest management in Kerala. Etude FAO: Forêts n° 53. Rome.

FAO. 1985a Plan d'action forestier tropical. Rome.

FAO. 1985b Aménagement polyvalent intensif des forêts sous les tropiques. Etudes de cas en Inde, en Afrique, en Amérique latine et aux Antilles. Etude FAO: Forêts n° 55, Rome.

FAO. 1986a Sylviculture tropicale. Réimpression de l'Etude des forêts et des produits forestiers n° 13 (1957). New Delhi, Periodical Experts Book Agency.

FAO. 1986b Plan d'action forestier tropical. Unasylva, 152 (38): 37-64.

FAO/PNUE. 1981a Proyecto de evaluación de los recursos forestales tropicales. Los recursos forestales de la América tropical. Première partie: synthèse régionale. Deuxième partie: résumés par pays.

FAO/PNUE. 1981b Tropical Forest Resources Assessment Project. Forest Resources of Tropical Asia. Rapport technique n° 3. UN 32/6. 1301-78-04

FAO/PNUE. 1981c Les ressources forestières de l'Afrique tropicale. Première partie: synthèse régionale. Deuxième partie: résumés par pays (rédigés en français, anglais ou espagnol selon la langue officielle de chaque pays).

FOX, J.E.D. & HEPBURN, A.J. 1972 Code of Silvicultural Practice. Sabah Forest Dept.

GÓMEZ-POMPA, A. 1985 La función protectora y los servicios ambientales de los bosques: el problema de la reforestación en el trópico. 9e Congrès forestier mondial.

HADLEY, M. & LANLY, J.P.1983 Tropical forest ecosystems: identifying differences, seeking similarities Nature and Resources, XIX. 1.

HUTCHINSON, I. 1981 Sarawak liberation thinning: background and an initial analysis of performance. A practical guide. Document de terrain n° 15. U N DP/ FAO/ MA L/ 7 6/ 008. Sarawak Forest Dept., Kuching, Malaisie. 121 p.

HUTCHINSON, I. 1986a Improvement thinning in natural tropical forests, aspects and institutionalization. In Symposium on Natural Management of Tropical Moist Forests. Yale School of Forestry and Environmental Studies. (In press)

HUTCHINSON, I. 1986b The management of humid tropical forests to produce wood. In US Forest Service Inst. Tropical Forestry Conf. on Management of tropical American forests: prospects and technologies. (In press)

JABIL, M. 1983 Problems and prospects in tropical rainforest management for sustained yield. Workshop on tropical rainforest management. Queensland.

JANKOVIC, C. 1971 Management of natural tropical forests in the northwest of Ecuador. FOL/CF (866). Quito.

JONKERS, W.B.J. & SCHMIDT, P. 1984 Ecology and timber production in tropical rainforest in Surinam. Interciencia, 9 (5): 290-297.

KARANI, P.K. 1986 Management of tropical rainforests of Uganda. Rome, FAO

KIO, P.R.O., EKWEBELAM, S:A., OGUNTALA, A.B. LADIPO, D.O. & NWONWU, F.O.C. 1985 Management systems in tropical mixed forests of anglophone Africa (with special examples from Nigeria). Rome, FAO.

LESLIE, A.J. 1985 Study of management systems in the tropical mixed forests of Asia. Rome, FAO.

MAÎTRE, H.F. 1982 Projet de recherches sylvicoles sur les peuplements naturels en forêt dense guyanaise. Ministère de la recherche et de la technologie, Secrétariat d'Etat aux DOM-TOM.

MAÎTRE, H.F. 1986 Recherches sur la dynamique et la production des peuplements naturels en forêt dense tropicale d'Afrique de l'Ouest. 18th World Congres of IUFRO, Ljubljana, Yougoslavie.

MASSON, J.L. 1983 Management of tropical mixed forests: preliminary assessment of present status. FO:MISC/83/17, Rome, FAO.

MATTSON MARN, H. & JONKERS, W.B.J. 1981 Logging damage in tropical high forest. Sarawak Forest Dept. Kuching, Malaisie.

MENGIN-LECREULX, P. & MAÎTRE, H.F. 1986 Les systèmes d'aménagement dans les forêts denses humides d'Afrique. Workshop on Regeneration of Tropical Rainforests, Venezuela. (In press)

NEYRA-ROMAN, M.G. 1979 El bosque de Guandal y sus posibilidades de manejo silvicultural. Proyecto de Investigaciones y Desarrollo Industrial Forestales. COL/74/005. PNUD/FAO.

NILLSON, H.E., MARSCH, H.E. & SINGH, H.W. 1978 Identification and planning of a National Forest Inventory for the Philippines. FAO Working Paper 15. PHI/72/006. Rome.

PHILIP, M.S. 1986 Management systems in the tropical moist forests of the anglophone countries of Africa. Rome, FAO. (In press)

ROMERO MEJÍA, R. 1983 La selva central: situación actual y perspectivas para su desarrollo forestal. Working document 11. PNUD/FAO/PER/ 81/002.

SALLEH, M.N. & BAHARUDIN, J. 1985 Silvicultural practices in Peninsular Malaysia. In The future of tropical rain forests in Southeast Asia. Commission on Ecology Papers, 10. UICN.

SCHIRMER, I. 1986 Les conditions de développement de la transformation locale de bois d'oeuvre tropical en Afrique de l'Ouest. Thèse de 3ecycle.

SMITH, D.M. 1962 The practice of silviculture, 7eéd. New York, John Wiley. 578 p.

SODEFOR/CTFT. 1985 Dispositifs d'étude de l'évolution de la forêt dense ivoirienne suivant différentes modalités d'intervention sylvicole. Présentation des principaux résultats après quatre années d'expérimentation.

SYNOTT, T.J. 1980 Tropical rainforest silviculture, a research project. CFI Ocasional Paper 10. Dept. For. Univ. Oxford. 45 p.

UNDP/FAO. 1970a Demonstration and training in forest, forest range and watershed management, the Philippines. Based on work by D.J. Nicholson. Technical Report technique n° 3. FO: SF/PHI 16. Rome. 52 pp.

UNDP/FAO. 1970b Estudio de preinversión para el desarrollo forestal en los valles del Magdalena y del Sinu. Technical Report. FO:SF/COL 14.

UNDP/FAO. 1977 Fortalecimiento del servicio forestal. FO:DP/ECU/71/527.

UNDP/FAO. 1979 Demostración de manejo y utilización integral de bosques tropicales. Perú. Resultados y recomendaciones del proyecto. Terminal Report. FO:DP/PER/71/551.

UNDP/FAO. 1980 Management Plan for the Tapajós National Forest. FO:BRA/78/003. Technical Report 2. Based on work by T.W.W. Wood. Brasilia.

UNDP/FAO. 1982a Forestry development project Sarawak terminal report. Project result, conclusions, recommendations. FO:MAL/76/008.

UNDP/FAO. 1982b Report of the Joint UNDP/FAO Review Mission for BRA/82/008. Brasilia.

UNDP/FAO. 1982b Desarrollo forestal en Ecuador. Resultudos y recomendaciones del proyecto. FO:DP/ECU/ 77/005. Rome.

UNDP/FAO. 1983a Roading, logging and transport in the Tapajós National Forest. Based on work by J. Overgaard by. Technical Report Draft. BRA/82/008.

UNDP/FAO. 1983b Forestry development Brazil: project findings and recommendations. Terminal Report. FO:DP/BRA/78/003.

UNDP/FAO. 1983c Forestry research in the Tapajós National Forest, Santarem, Brazil. Based on work by P.L. Weaver. FO:BRA/82/008.

UNDP/FAO. 1985a Forestry development in northeast Brazil; terminal statement. FO:TCP/BRA 2202.

UNDP/FAO. 1985b Investigación en silvicultura tropical. Based on work by J.T. Sterringa.

Unesco/UNEP/FAO. 1979 Tropical forest ecosystems. A state-ofknewledge report.

VADYA, A.P., COLFER, C.J.P. & BROTOKUSUMO, M. 1980 Interactions between people and forests in East Kalimantan. Impact, Sci. Soc., 30: 179-90

VANNIERE, B. 1974 Management possibilities of tropical high forest in Africa. Rome, FAO.

WADSWORTH, F.H. 1969 Posibilidades futuras de los bosques del Paraguay. Working document 2. FAO:SF/PAR 15.

WOOD, T.W.W. 1982 Suriname forest management. Operational aspects. Project working document 22. FO:DP/SUR/81/004. Paramaribo.

WYATT-SMITH J. 1963 Manual of Malaysian silviculture for inland forests. 2 vol. Malaysia Forest Record 23. Forestry Dept Malaysia.

Previous Page Top of Page Next Page