Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Agro-forestry research for the humid tropics

A.G. Seif el Din
International Development Research Centre, Nairobi, Kenya


If the principal aim of agro-forestry is to meet the needs of human communities, then research should be directed towards preservation or modification of the undisturbed habitats, rehabilitation of the disturbed ones, and rational use of the improved ecosystems. The goals of agro-forestry should be food production, forest production, and environmental conservation. Within the forest ecosystem, taungya offers particular scope for research, especially the determination of how long the forest land can sustain food crops before complete canopy closure, and the identification of optimal spacings that maximize food production without interfering with the form and the rate of tree growth. Within the agricultural production system, research should be devoted to the introduction of trees that rejuvenate soil fertility by nutrient recycling within the shortest possible time and that produce forest products to increase the farmers' income or satisfy their needs.


This presentation does not undertake the definition of agro-forestry practices and systems, because much has already been written and said in this regard. The term agro-forestry is used here as defined by King and Chandler (1978), i.e., a sustainable land management system that increases the overall yield of land. There seem to be two main reasons for the growing interest in a sustainable land management system: the first is that there is a serious degradation of the ecosystem as a result of deforestation; the second, which is a consequence of the first, is that the global forest resources and arable land areas are diminishing at an alarming rate. The environmental aspect of agro-forestry necessitates planting trees in such a way as to bring about conservation and improvement of the ecological factors that influence the production systems. In addition, the trees should provide one or more of the forest products, such as wood, fodder, and food, that are needed for the improvement of the living conditions in the rural areas.

The term agro-forestry should thus imply land management systems in which forest management, food production, and conservation form integrated components, i.e., a multidisciplinary exercise.

Some of the recognized agro-forestry practices have been in use for centuries in many parts of the world (King 1968). They were evolved in the rural areas when demands for specific tree products were not easily met from the natural forest or when a tree species proved useful and important to food crops. Gum arabic production in the Sudan and the cultivation of crops under Acacia a/bide in West Africa are two of the known examples of agro-forestry practices today two of the better-known examples of agro-forestry practices today (Self el Din 1980). The continuation of these practices is threatened by the problems caused by the growing human populations, for example, over-cultivation of the land and, hence, the gradual elimination of the tree component from the system. The task of agro-forestry research is, therefore, to find the best system within physical and biological limitations and communal or individual socio-economic requirements.

One of the important tasks of agro-forestry research workers should be to find ways to maintain stability in ecosystems that have not yet been seriously interfered with, to render them more productive of human needs. The painful reality is, unfortunately, that there is very little left of the natural forest to maintain. Consequently, agro-forestry research programmes should be directed towards tackling the immensely difficult task of rehabilitating degraded lands and devising land management systems that will fulfil the objectives of agro-forestry.

The gravity of the problem is apparent in the rapid disappearance of the tropical moist forests. Myers (1980) has given the global rate of conversion of these forests into other forms of land use as being 40 ha/minute. He states that up to the mid-1950s, about 1 million kmē of tropical forests were lost in Africa alone, and that 40,000 kmē of forest are annually disappearing on this continent. The Ivory Coast is estimated to have lost nearly 40 per cent of its forests between 1966 and 1974, and in Nigeria it is reported that only about 25,000 kmē of forest exist today, representing less than 3 per cent of the total area of the country. Addo-Ashong (1980) states that of the original 82,000 kmē of tropical rainforest in Ghana, only 20,500 kmē or one-quarter, remain today.

Research Proposals

If the principal objective of agro-forestry is to meet the immediate and future requirements of the human communities living in the region concerned, then the approach should be directed towards the preservation or modification of undisturbed habitats; the rehabilitation of disturbed ones; and rational utilization of the improved ecosystems.

In other words, agro-forestry should be viewed in the context of food production, forest production, and environmental conservation, which is similar to what has been suggested by Combe (1979) as being economic, ecological, and silvicultural. The result of an agro-forestry approach should be a new environment, dissimilar but strongly related to the natural one in its basic features, so that it is reasonably stable and productive enough to satisfy human requirements. It should offer, as stated by Wassink (1977), "a reasonable and acceptable permanent way of life to the people that dwell therein."

Several workers have suggested the essential elements of an agro-forestry system (Greenland 1977; Grinnell 1975; Combe and Budowski 1979; and Reategui 1979). Most people agree that an agro-forestry system should cater to:

The next step is to consider the type of land and the combinations of plants and animals that constitute an agro-forestry system. There are considerable differences of opinion in this regard, for the simple reason that most of the ideas advocated in agro-forestry have not yet been subjected to systematic investigation.

In the Philippines, Kuo (1977) proposed three classes of land: forest land, agro-forestry land, and agricultural land. He places the forests on the steepest slopes with the shallowest soils, agriculture on the deepest soils with gentle slopes, and agro-forestry in the middle occupying a relatively small portion of the area. He points out, however, that the latter can be practiced on either side of the classification depending on land-use policy and soil characteristics.

In the example cited by Kuo, some of the forest lands are so steep that any cropping or grazing would endanger the entire landscape, whereas some areas are suitable for continuous cultivation of agricultural crops that can only be economically produced when cultivated in monocultures. The implication is that not all crops and not all lands can be included in agro-forestry systems. Consequently, there should be areas solely managed for crops, pastures, or pure forest stands. For the agro-forestry combinations, the broad classification of the systems proposed by Budowski (1977) and also by Torres (1979) are thought to be adequate, i.e., trees combined with farm crops; trees combined with farm crops and animals; and trees combined with animals. In order to arrive at the desired agro-forestry system, one must tackle research from both the agricultural and the forestry perspective. These two disciplines have, for a long time, had conflicting interests over land use, whereas livestock was, in most cases, completely neglected. Forestry and crop production systems should therefore be gradually modified independently toward a common end where both meet at a stage called agro forestry. Livestock production should be accommodated in both systems in such a way that optimum yields are obtained without adversely affecting the system, and even improving it. This, simply stated, means production of crops and livestock within forest land and production of forest products and livestock on agricultural land.

Some agricultural research scientists view the subject of agro-forestry as merely involving the use of short-lived, fast growing tree species to improve soil fertility for food production and not as a source for forestry production. Still others view it as an introduction of new tree crops into farming systems to increase and diversify the food supplies of the rural people. The last point of view is similar to that of domestication of new animals as a form of agrosilvopastoral system. The danger of this approach to agro forestry is that very little is known about the biology and the environmental implications of the new crops and animals.

Some of the forest research scientists look upon agro-forestry as another name for the taungya system, in which a certain amount of food is produced at the treeplanting stages of forestry plantations. This group does not take into consideration the need for the use of the space available for food production at the time of tree establishment in even-aged plantations, at the stage of opening up the canopy for natural regeneration in what is known in West Africa as the Tropical Shelterwood System (TSS), and, finally, at the various stages of stand treatments such as thinning. With this understanding of agro forestry, the taungya system then forms only one step of the comprehensive system that calls for the use of appropriate combinations of plants and animals and the best spatial and temporal distribution for food and wood production as well as for the preservation of the ecosystem. Agro-forestry research on the taungya system should be devoted to the determination of how long the forest land can sustain cropping before complete canopy closure or the decline of soil fertility, and to the identification of optimal spacings that maximize food production without interfering with the form and the rate of tree growth as required in traditional forest management.

Food Production within the Forest Ecosystem

Various silvicultural systems were developed to regenerate the natural forests in the humid tropics of West Africa, but the most popular one at present is the TSS. This involves gradually opening up the forest canopy by killing and clearing the trees so that sufficient light is available for the growth of the natural seedlings of the desirable species. The degree of clearance depends on the forest type (density and composition by species), the available resources, and the status of the naturally occurring seedlings.

The agro-forestry research approach here should aim at food production by increasing the intensity of clearance and by the use of shade-tolerant food crops that do not endanger the tree species. Even though little experience is available to show the interactions between food crops and trees, it is possible that the latter would benefit from cropping because most of the undergrowth (bush) would be removed to provide space for the food crops and, hence, competition from weeds would be reduced.

It is desirable to test different crops at different stages of tree growth that entail changes in the environment, especially the amount of leaf litter and the sunlight reaching the ground surface.

Once forest stands pass the seedling stage, they are treated at different intervals by processes known as cleaning and thinning to create the best possible stand of managed forest. The intensity of these operations, especially thinning, depends again on the tree species concerned and the objectives set out for each plantation. Because the need for more food producing space is universally recognized, thinning itself can be modified to provide room for crops and livestock without jeopardizing the objectives of forest management. Appropriate avenues for research are the planting of food crops at different intensities of thinning, and fodder production at the various stages of thinning from both the planted trees and the natural undergrowth.

The main points to be considered when planning the research experiments are: ( 1 ) the types of food crops that will favourably respond to the newly created environment in the forest without endangering the trees; and (2) the potential for the introduction of browse shrubs and grasses to enrich the natural pastures for optimum livestock production.

Forestry Production on Farm Plots

Forestry production on farm plots is feasible only if there is demand for the forest products grown on the farm plots; otherwise there will be little incentive to the farmers to plant trees. However, if the gains in soil fertility are clearly demonstrated, farmers may be willing to introduce and maintain trees. This possibility is the basis for present efforts by researchers to adapt and optimize the traditional system of land rotation or shifting cultivation, where soil fertility is regained through the regrowth of the natural bush during the fallow period. The suggestion is that trees be planted on abandoned farm plots to rejuvenate soil fertility by nutrient recycling within the shortest possible time and also produce forest products to increase the farmers' income or satisfy their own requirements.

The tree species used in these efforts must be fast-growing to produce marketable products in a short time, i.e., the farmer does not lose time when he again requires the land for food production. At the same time, the tree plantation should be at least as good as - and preferably better than-the natural bush in restoring soil fertility.

Tree planting in conjunction with food crops entails producing forest products and food crops simultaneously on the same piece of land. In addition to their principal role, the trees should improve the soil. In this way the farming system can be prolonged beyond the current one to three years and perhaps lead to the development of a combination that will sustain permanent cropping. The combination aimed at is an extension of that occurring in certain dry tropical areas where food crops are cultivated under Acacia albida, but, in addition, the trees will be managed in such a way as to produce marketable products continuously or at regular intervals. The correct tree species and their spatial arrangements should be adequately investigated in research trials.

Some crops, like cocoa, coffee, etc., are known to tolerate a certain amount of overhead shading. A number of forest tree species have been successfully used to provide shade as well as produce wood. One example in South America cited by Fuentes Flores (1979) is what he calls "stratified tropical agriculture" to produce coffee, citrus fruits, plantain, beehives, and vegetables under scattered trees of Cedrela sp. In India, coconut plantations are intensively intercropped in ways that maximize the use of both aerial and subterranean space (Nair 1979). Okigbo (1977b) and Getahun (1980b) have compiled comprehensive lists of woody plants, including forest trees, which are of nutritional importance in the traditional farming systems of Africa.

It should be pointed out that neither the use of shrubs solely for soil improvement nor the introduction of new tree crops on the farm as additional food sources serves the purpose of agro-forestry. instead, the research should be directed towards finding the tree species that create favourable environmental conditions for optimal production of crops and livestock in addition to wood production.

The research topics outlined are by no means exhaustive. Instead, they should be considered as a basis for discus signs leading to a comprehensive agro forestry programme with long- and short-term objectives. Some of these proposals are likely to apply to the drier tropics where agro-forestry is just as important, in fact even more so, especially when livestock production becomes a more prominent landuse factor.

Before an agro-forestry research programme is worked out, sufficient basic information must be obtained. There should be a survey of the existing land-use practices in the region in a manner similar to that carried out by Getahun (1979a) in southern Nigeria. This will enable researchers to understand the way in which the rural people operate so that they can aim at improving the existing land-use systems rather than designing entirely new ones that would not be easy to implement. The survey should include evaluation of the soil characteristics, the existing vegetation and its effect on soil protection! and the needs of the local communities in terms of food and forest products. The next step is to identify the crops that grow best under trees and the forest trees that will provide protection and soil improvement on farm plots. The research activities should be designed in such a way to find the optimum combinations of trees, crops, and animals to ensure maximum productivity and improvement of the environmental conditions. Finally, a strategy for agro-forestry research should be outlined and methods for its implementation be developed. One such strategy has been proposed by Steppler and Raintree (1981 ) and provides the basis from which to start (see paper by Steppler in this volume, pp. 1 - 5).


Summary of discussion: Principles of agro-forestry

The presented papers stimulated extensive discussion, and this focused mainly on the institutional and socio-economic constraints to the dissemination and adoption of agro-forestry techniques. Several speakers emphasized that one must begin with the needs of the farmer, and that if proper attention had been paid in the past to their perspective, much misunderstanding and trouble could have been avoided. Scientists must be realistic and should not expect farmers to adopt agro-forestry practices just because of the value of trees in terms of soil and water conservation. There must be more direct benefits for the farmer, especially if they do not own the land. In short, conservation must be designed as a spin-off of agro-forestry and cannot be considered as a selling point on the farm level.

With reference to the planting of tree crops, it was recognized that the most successful method of persuading farmers was through financial inducements, with two basic elements being free seedlings and appropriate extension services. It was pointed out that the farmer may occasionally receive conflicting advice from different extension agents or institutions. In this sense agro-forestry should not consider setting up a third layer of extension services, but should be integrated into existing programmes.

Particular attention was paid to the fact that current legal systems have been primarily designed to protect forests and, if anything, serve to discourage farmers from growing longterm tree crops, especially for timber. Even if a compensation system does exist, these are generally for growing plantations (e.g., in a taungya system) and are not applicable to other agro-forestry situations. Land tenure plays a decisive role, for if the farmer is in a forest estate or forest reserve, or if he is only leasing the land, there is no incentive for him to grow or protect trees. In many cases the farmer is liable for any merchantable trees, so it may be easier for him simply not to allow any trees to become established.

With regard to the design of agro-forestry systems, the theme which ran throughout the workshop was that we must begin with existing systems. An urgent need was felt to catalogue existing systems-successes as well as failures-and then try to quantify them. It was pointed out that ICRAF's approach will follow this basic methodology of assessing structure and function of existing land-use systems, particularly those which show signs of deterioration, and will then try to formulate alternative systems. Both the conceptual and methodological problems of comparing farming systems were brought out. The sheer number of variables puts agro-forestry into a dilemma, for full quantification and exhaustive testing of agro-forestry systems is generally impossible, yet one cannot try to disseminate an unproved system to the farmers. Some balance must also be found between the in-depth investigation of single systems on the one hand, and the fact that there will be an infinite variety of systems due to both environmental factors (soil type, rainfall, etc.) and socioeconomic factors (access to market, cash crops vs. subsistence crops, etc.).

In evaluating agro-forestry systems, it was generally reaffirmed that the most valid standard would be a comparison with monocultures, and one should not automatically expect that agro-forestry systems would be more successful. For one thing, the management of farms with mixed crops is more difficult, and this impedes mechanization. It was also pointed out that trees may be able to bring more nutrients into the system by bringing up deep nutrients and lowering leaching rates, but they also may immobilize certain nutrients for a long period of time and eventually export large quantities of certain nutrients in the form of fruits, leaves, or wood. Performance must therefore be qualified in terms of factors such as time-span, economic return vs. subsistence, biological productivity, sustainability, etc. It was noted that since agro-forestry systems might be successful in marginal areas unfit for traditional agricultural systems, a comparison with monocultures would not always apply. A cautionary note sounded by several speakers was that we cannot regard agro-forestry as a panacea, but as an alternative land-use system which must be adequately tested and impartially evaluated.

Another theme which ran throughout the discussion was the need to approach agro-forestry in a multidisciplinary manner. Agronomists or foresters working alone would not be able to fully develop the potential of agro-forestry systems, and some participants felt that this meant a new institutional framework had to be created. Other participants stressed the need to revamp the heretofore narrow training of both scientists and extension workers to take account of the multidisciplinary approach demanded by agro-forestry; this was pragmatically described as taking food crops into the forest and trees into the fields. The idea of integrating farmers into the forest was recognized as anathema to most traditional foresters, but necessary to meet the needs of the local inhabitants. This integration could also be the basis for a more cordial relationship between farmers and foresters than has been the case in the past.

Finally, it was pointed out that the simple mixing of trees and crops was not sufficient, as the improvement of yields would probably require improved varieties, weed control and fertilizers. To be truly effective in improving living standards. agro-forestry should form part of an integrated rural development programme and thereby meet more of the farmer's basic needs.

Contents - Previous - Next