An e-publication by the World Agroforestry Centre


section I : general

Agroforestry systems in major ecological zones of the tropics and subtropics

P.K.R. Nair
International Council for Research in Agroforestry (ICRAF)
P. O. Box 30677, Nairobi, Kenya


During the past five years, ICRAF, through its Agroforestry Systems Inventory (AFSI), has collected, evaluated, synthesized and disseminated information on existing agroforestry (AF) systems and practices from different parts of the tropics and subtropics. The systems have been classified according to various criteria such as their structure, function, socio-economic aspects, ecological spread, and so on. But the primary step in classifying them is on the basis of their structure; thus there are three basic types of 'AF systems

  • agrosilvicultural (crops land trees)

  • silvipastoral (trees and pasture/animals), and

  • agrosilvipastoral (crops + pasture/animals + trees).

Under each of these major AF types, several systems have been recorded in the AFSI data base. But the distinct AF practices that constitute these systems are only a few; these are discussed in the paper.

An ecological analysis of the spread of various AF systems shows that the existence / adoption of an AF system in a given area is determined primarily by the ecological potential of the area, but socio-cultural and economic factors determine the complexity of the system and the degree of intensity of its management. From a purely ecological point, certain types of systems can be identified as characteristic of each major ecological region. Examples are:


There are certain types of systems that are present in almost all ecological regions; however, the nature of components in each will vary from region to region depending on the ecological potential of the regions .


ICRAF has just completed a global inventory of existing agroforestry systems and practices in different parts of the tropics and subtropics. Initiated in September 1982 with partial financial support from the United States Agency for International Development (USAID), the project had the general objective of increasing the understanding about the extent of existing agroforestry systems, and their various productive and protective roles in land use systems in the developing countries. The exercise consisted of collection, collation, synthesis, evaluation and storage of data, and publication of various results and outputs concerning the existing land use systems of relevance to agroforestry, and it involved a large number of institutions and individuals interested in and knowledgeable about agroforestry from all over the world. This project has gathered information on a few hundreds of agroforestry systems, and this information is available in the form of:

  1. Agroforestry System Descriptions, being serialized in Agroforestry Systems (published by Martinus Nijhoff, The Hague). A list of the system descriptions is attached (Appendix page 93-94);

  2. A computerized Agroforestry Systems Register at ICRAF containing data entries on about 150 systems and/or practices;

  3. Various miscellaneous publications from the project (Appendix page 94-95);

  4. Unpublished records at ICRAF.

This paper is based on this vast information base on agroforestry systems. Additional details of any aspect of the project can be obtained from the author.

Types of agroforestry systems

In order to evaluate the existing agroforestry systems and develop action plans for their improvement, it is necessary to classify these systems and thus provide a practical and realistic framework. Depending on the main purpose to be served by a classification scheme, various criteria can be employed to classify the systems. The most common among these are the system's structure, functions, socio-economic scale and level of management and ecological (environmental) spread:

  1. Structural basis — refers to the composition of the components, including spatial admixture of the woody component, vertical stratification of the component mix and temporal arrangement of the different components;

  2. Functional basis — refers to the major function or role of the system, mainly of the woody components (these can be productive, e.g. production of food, fodder, fuelwood and so on), or protective (e.g. windbreak, shelterbelt, soil conservation and so on);

  3. Socio-economic basis — refers to the level of inputs of management (low-input) or intensity or scale of management and commercial goals (subsistence, commercial, intermediate);

  4. Ecological basis — refers to the environmental condition and ecological suitability of the systems, on the assumption that certain types of systems are more common in/appropriate for certain ecological conditions

A classification scheme of agroforestry systems based on these approaches is summarized in Table 1. These underlying principles of classification are by no means independent or mutually exclusive. Indeed it is obvious that they have to be interrelated because the structural and functional bases relate to the woody components in the systems whereas the socio-economic and ecological stratifications refer to the organization of systems according to certain defined conditions (socio-economic or ecological). And in any defined socio- economic or ecological situation, the systems should still have specific structure (composition and arrangement or components) and function / role.

Nevertheless, in agroforestry land-use systems, there are only three basic sets of components that are managed by man, viz., woody perennials, herbaceous plants and animals. Therefore, a simple and logical first step in classifying agroforestry systems is to use this component composition as the basis, which will give three basic types of agroforestry systems (Nair 1985):

  1. Agrisilvicultural       (crops — including tree and shrub crops — and trees);

  2. Silvopastoral          (pasture with animals and trees); and

  3. Agrosilvopastoral   (crops + pasture with animals + trees).

Other specialized agroforestry systems such as apiculture with trees, aquaculture in mangrove areas, multipurpose tree lots, and so on, can also be specified. A generalized picture of the component composition of these basic types of agroforestry systems is given in Figure 1.

This categorization of agroforestry systems into three major types is so fundamental that anyone of these can conveniently be used as a prefix to other terms emanating from other efforts to classify or group agroforestry systems in order to explicitly express the basic composition of constituents of any such system. For example, there can be an agrisilvicultural system for food production, a commercial silvopastoral system for fodder and food production in lowland sub-humid (or dry) tropics, an agrosilvopastoral system for food production and soil conservation in highland humid tropics, and so on. Therefore it seems logical, compatible and pragmatic to accept the nature of components as the basic criterion in the classification of agroforestry systems.

Agroforestry practices and technologies

A further evaluation of the organization of these various agroforestry systems shows that an agroforestry system involves extensive application of agroforestry practices, thus forming a significant form of land use in a locality or area, described according to the system's biological composition and arrangement, level of technical management, or socio-economic features. An agroforestry practice, on the other hand, denotes a specific agroforestry combination in a land-management unit, and consists usually of arrangements of agroforestry components in space and/or time. Several such practices are involved in the constitution and maintenance of an agroforestry system.

Although several agroforestry systems have been recorded from around the world, the distinctive agroforestry practices that constitute these systems in various ecoregions and locations are few in number. These most commonly found agroforestry practices and their essential characteristics are listed in Table 2 (Appendix page 72-77).

The scale of management and extent of adoption of these different practices in any particular system will vary considerably depending upon a number of factors. Any of these practices can become an agroforestry system when it is developed or spread to such an extent in a specific local area so as to form a distinct land utilization type in that area. One essential point to note here is that an agroforestry practice can be found to exist even in a non-agroforestry land-use system. An example is the practice of growing rows of Sesbania grandiflora on the bunds of rice paddies in Java, Indonesia. This woody species is harvested for its leaves as green manure, its flowers are eaten as a vegetable, it provides firewood and it adds to the fertility of the soil underneath through biological fixation of atmospheric nitrogen. Thus the woody species interacts ecologically and economically with the crop (rice) production system; in other words, it is an agroforestry practice in a crop production system.

Table 1 Major approches in classification of agroforestry systems and practices (Nair 1985)




Another term that is commonly used in all land use systems including agroforestry is 'technology'; it is usually used in combination with a particular production system or land-use system as the prefix; thus, agricultural technology, wood production technology, agroforestry technology, and so on. The word or suffix 'technology' in this sense indicates an improvement or innovation, usually through scientific intervention, that can be applied with advantage in the management of the system or practice concerned. The inputs that are used for developing such technologies can sometimes be the most significant aspects, and these are called input technologies; e.g., improved varieties of plants, agrochemicals, and so on.

Most, if not all, of the practices mentioned in Table 2 (Appendix page 72-77) need to be improved scientifically and developed into modern agroforestry technologies. The success potential of an agroforestry practice (and therefore an agroforestry system) depends on the extent to which such technologies have been perfected, and made use of in the management of the practice (or the system).

Ecological spread of major agroforestry systems

Ecological zones of agroforestry significance in the tropics and subtropics

The major physical parameters that determine the ecology of a location in the tropics are rainfall (quantity and distribution) and temperature regimes. Altitude also is important because of its influence not only on temperature, but also on land relief characteristics. The FAO Agro-Ecological Zones inventory divides tropics into 'warm', 'moderately cool', 'cool' and 'cold' climates depending on the climatic (mainly temperature) regime during the 'growing period'. These growing periods refer to a few selected agricultural crops (that are seasonal) and therefore do not necessarily reflect the growing periods of perennial species (although it can be argued that the agricultural growing period represents the peak growing seasons of the perennial species also in most cases).

From the agroforestry point of view, the major ecological regions recognized in the FAO State of Food and Agriculture Reports (SOFA) are more relevant: these are temperate, mediterranean, arid and semi-arid, sub-humid tropical (lowland), humid tropical (lowland) and highland. These classes, excepting the first (and possibly the second), represent the tropical and subtropical lands of agroforestry significance where agroforestry systems exist or have a potential. These major ecological regions of the tropics and subtropics are presented in Figure 2, and the main characteristics of these major agro-ecological regions (adapted mainly from Young (1986)) are summarized in Table 3.

Agroforestry systems in lowland humid and sub-humid tropics

Characterized by hot, humid climates for all or most of the year, and an evergreen or semi-evergreen vegetation, the lowland humid and sub- humid tropics (but referred to as humid tropics) is by far the most important ecological region in terms of the total human population it supports, extent of area and diversity of agroforestry and other land-use systems. Because of the climatic conditions that favour rapid growth of a large number of plant species, various types of agroforestry plant associations can be found in areas with high human population. Various forms of homegardens, plantation crop combinations, and multilayer tree gardens are common in such regions. In areas with low population density, such as the low selvas of Latin America, trees on rangelands and pastures, other silvopastoral systems, improved fallow in shifting cultivation areas, multipurpose tree woodlots, etc., are the major agroforestry systems. Examples of agroforestry systems in humid tropical lowlands about which ICRAF has data entries in the computerized AFSI database are included in Table 4 (Appendix page 78-92).




Table 3 Main characteristics of the major ecological regions of agroforestry importance in the tropics and subtropics. (Adapted from A. Young 1986; in consultation with A. Young.)

agroforestry_systems_major_5 agroforestry_systems_major_7

The lowland humid tropics also represents areas under natural rain forests. In such areas, cutting of rain forest at rates much faster than natural or managed regeneration is a common problem. This causes shortening of fallow periods in shifting cultivation cycles and resultant loss of soil productivity and accelerated soil erosion. The potential of appropriate agroforestry systems to combat these problems needs to be exploited in future land-use strategies in this zone.

Agroforestry systems in semi-arid and arid tropics

Extending over the savanna and Sudano-Sahelian zones of Africa, the cerrado of South America and large areas of the Indian subcontinent, the semi-arid and arid tropics are characterized by one or two wet seasons (Köppen Aw or Aw", respectively) and at least one long dry season. Drought is a hazard in the drier parts of the zone.

The main agroforestry systems in this zone also depend on the human population pressure; homegardens and multilayer tree gardens are common in the wetter areas with high population pressure. But generally speaking, the predominant agroforestry systems in this zone are:

  1. Various forms of silvopastoral systems;

  2. Windbreaks and shelterbelts; and

  3. Multipurpose trees on crop lands, notably Acacia albida-based systems in Africa and Prosopis-based agrisilvicultural systems in the Indian sub-continent.

Fuelwood shortage is a major problem in most parts of the semi- arid and arid tropics. Agroforestry potentials in fuelwood production are well documented (e.g., Nair, forthcoming). Similarly, desertification and fodder shortage — the other major ecological problems in this zone — could be tackled to some extent through the agroforestry approach.

Agroforestry systems in tropical highlands

Approximately 20% of the tropical lands are at elevations from 900 to 1800 m, in about half of the Andean highlands of Central and South America, parts of Venezuela and Brazil, the mountain regions of the Caribbean, many parts of East and Central Africa, the Cameroons, the Deccan Plateau of India and some parts of the southeast Asia mainland. The altitude exceeds 1800 m in about 3% of the tropical area in the Andes, the Ethiopian and Kenyan Highlands, northern Burma and parts of Papua New Guinea. In the subtropical regions, the most important highlands are in the Himalayan region. These highland tropics with significant agroforestry potential are humid or sub-humid, such areas with dry climates being of very low potential.

Land-use problems in the highlands are similar to those in humid or dry lowlands depending on the climate, with the addition that sloping lands and steep terrains make soil erosion an issue of major concern. Moreover, annual temperatures are low in the highlands ( for every 100 m increase in elevation in the tropics, there is a decline of 0.6°DC in the mean annual temperature), which affect the growth of certain lowland tropical species.

The main agroforestry systems in tropical highlands are:

  1. Production systems involving plantation crops such as coffee and tea in commercial as well as smallholder systems;

  2. Use of woody perennials in soil conservation and soil fertility maintenance;

  3. Improved fallows; and

  4. Silvopastoral combinations.

Agro-ecological analysis of existing agroforestry systems

The existence of an agroforestry system or practice in a particular area is determined not only by the environmental and agro-ecological factors, but also by socio-economic considerations. In different areas with identical agroclimatic conditions, factors such as human population pressure, availability of labour and other production resources, proximity and accessibility to market sources, etc., are the main determinants of the types and forms of agroforestry systems. Even in the case of systems that are widespread in most ecological and geographical regions such as shifting cultivation and taungya, these systems have very many variants that are specific to certain defined situations. As a general rule, it can be said that while the ecological factors determine the major type of an agroforestry system in a given locality, the complexity of the system and the degree of intensity of its management increase in direct proportion to the population intensity and land productivity of the area. A typical example is the spread of the multi-species, multi-storied homegarden systems. Though found mostly in humid lowlands, the home-gardens are common in pockets of high population density in other ecological regions also. Analysing the structural and functional aspects 10 selected homegarden systems from different ecological regions, Fernandes and Nair (1986) found that although the average size of the homegarden units is less than 0.5 ha, the units are composed of a large number of woody and herbaceous species, carefully structured to form 3-5 vertical canopy strata, and each component having a specific place as well as function in the multispecies complex.

There are some functional aspects of specific agroforestry practices that are emphasized or necessitated in specific agroecological situations. For example, the functional emphasis of agroforestry systems in sloping lands is erosion control and soil conservation; in wind-prone areas, woody species are successfully used in agroforestry practices of shelterbelts and windbreaks; the emphasis is on fuelwood production in areas with fuelwood shortage; there are also specific agroforestry approaches for the reclamation of degraded lands and other waste lands (such as highly eroded lands, areas with high levels of salinity or alkalinity, etc.)

The ecological potential of an area is undoubtedly the prime factor that decides the adoption and spread of specific agroforestry systems in different parts. The preponderance of homegardens and other multi-species mixes in fertile lowlands and high potential areas at one end of the ecological scale and extensive silvopastoralism at the other end, with various systems in between gives a spectrum of agroecological adaptability and spread of various agroforestry systems. Figure 3 shows the general pattern of the major agroforestry systems in various ecological regions. It does not include all agroforestry systems, nor does it fully portray the relative spread (in terms of areas involved) or the level of management of the various systems. A further step in this direction is to indicate the major known areas of concentration of important agroforestry systems in the tropics and subtropics, i.e., something like a 'map of the agroforestry situation in the tropics and subtropics' (Figure 4).

This 'map' represents areas with relative abundance of specified agroforestry systems, but does not identify localized systems as well as systems whose presence is relatively less significant in an area. For example, as mentioned earlier, 'multipurpose trees on farmlands' can be found in almost all ecological and geographical regions. But there are only a few such systems that make them distinctly unique agroforestry systems. Examples include the arid zone systems involving Acacia albida (Miehe 1986; Poschen 1986) and Prosopis (Shankarnarayan and Harsh 1987). Thus, only some such 'multipurpose trees on farmlands' systems have been studied and/or described in some detail, and thus gone into ICRAF's Agroforestry Systems Inventory (AFSI) data records. The mention of the existence of any of these systems on Figure 4 therefore is mainly an indication that AFSI records are available. Indeed, innumerable location-specific agroforestry systems and practices do exist all over the tropics and subtropics and many of them are important in various ways. But most of them may not often be significant enough in terms of their individual importance in the overall economy and land-use pattern of the area where they exist, and thus may not merit a special mention on such a condensed and abstract global map.


Legend for system names in Figure 3 and Figure 4


This 'map' is an improvement over Figure 3 in that geographical spread of the systems is shown in relation to the ecological conditions in different places. One significant feature that emerges from this is that irrespective of the socio-cultural differences in different geographical regions, the major types of agroforestry systems are structurally similar in areas with identical ecological characteristics. The level of intensity at which these different systems are managed will, obviously, be different in various situations.



This paper is based on data gathered by ICRAFs Agroforestry Systems Inventory (AFSI). Useful discussions on ecological zonation were held with T. Darnhofer and A. Young of ICRAF.


Fernandes, E.C.M. and P.K.R. Nair. 1986. An evaluation of the structure and function of some tropical homegardens. Agric. Syst. 21: 279-310.

Miehe, S. 1986. Acacia albida and other multipurpose trees on the Fur farmlands in the Jebel Marra highlands, Western Darfur, Sudan. Agrofor. Syst. 4: 89-119.

Nair, P.K.R. 1985. Classification of agroforestry systems. Agrofor. Syst. 3: 97-128.

Nair, P.K.R. 1987. Agroforestry and firewood production. In D.O. Hall and R.P. Over-end (eds.), Biomass: renewable energy. Chichester: John Wiley and Sons.

Poschen, P. 1986. An evaluation of the Acacia albida-based agroforestry practices in the Hararghe Highlands of eastern Ethopia.Agrofor.Syst.. Syst. 4:129-142.

Shankarnarayan, KA. and L.N. Harsh, L.N. 1987. Agroforestry in the arid zone of India. Agrofor. Syst. 5: 69-88.

Young, A. 1986. Evaluation of agroforestry potential in sloping areas. In Land evaluation for land-use planning and conservation in sloping areas. ILRI Publication 40:106-132. Wageningen: Institute for Land Reclamation and Improvement.

Table 2 Major agroforestry practices


Table 4 Major agroforestry systems and practices in different ecological zones of the tropics and subtropics (Source: AFSI Database, ICRAF)


Publications from the Agroforestry Systems Inventory

Project: agroforestry systems descriptions

This series of agroforestry system descriptions is edited by Dr. P.K.R. Nair and published by Martinus Nijhoff/Dr. W. Junk Publishers in the journal Agroforestry Systems.


Publications from the Agroforestry Systems Inventory

Project: miscellaneous publications.


Nair, P.K.R. (ed.). Agroforestry systems in the tropics. Martinus Nijhoff, The Hague (in preparation).


Nair, P.K.R. and E.C.M. Fernandes. 1984. Agroforestry as an alternative to shifting cultivation. In Improved Production Systems as an Alternative to Shifting Cultivation. FAO Soils Bulletin 53:169-182. Rome: United Nations Food and Agriculture Organization.

Nair, P.K.R., E.C.M. Fernandes and P.N. Wambugu. 1984. Multipurpose leguminous trees and shrubs for agroforestry. Agroforestry Systems 2:1 45-163. (Also published in Pesquisa Agropecuaria Brasileira 19: 295-313).

Nair, P.K.R. 1984. Fruit trees in agroforestry. Working Paper, East- West Center, Honolulu, Hawaii, (also published as ICRAF Working Paper No. 32,1985).

Nair, P.K.R. 1985. Classification of agroforestry systems. Agroforestry Systems 3: 97-128. (ICRAF Reprint No. 23.1985).

Nair, P.K.R. 1985. Agroforestry. Paper presented to the Conference of Agricultural and Rural Development Officers of the Asia Missions of US AID, Los Banos, PI, 22-26 April 1985. International Rice Research Institute.

Nair, P.K.R. 1985. Agroforestry in the context of land clearing and development in the tropics. Paper presented to the International Workshop on Land Clearing and Development, IBSRAM, Jakarta, 26 August - 03 September 1985. (ICRAF Working Paper No. 33,1985).

Nair, P.K.R. 1986. Alternative and improved land use systems to replace resource-depleting shifting cultivation. In Strategies, approaches and systems in integrated watershed management. FAO Conservation Guide 14:61-84. Rome: United Nations Food and Agriculture Organization.

Nair, P.K.R. 1987. Agroforestry and firewood production. In D.O. Hall and R.P. Over-end (eds.), Biomass: renewable energy. Chichester: John Wiley and Sons.

Nair, P.K.R. (in press). Tropical agroforestry systems and practices. In J.I. Furtado and K. Ruddle (eds.), Tropical resource ecology and development. Chapter 14. Chichester: John Wiley and Sons.

Fernandes, E.C.M. and P.K.R. Nair. 1986. An evaluation of the structure and function of some tropical homegardens. Agricultural Systems 21:179-210.

Nair, P.K.R. 1986. Agroforestry systems and practices: their application in multiple use of forests. Keynote speech to V Brazilian Forestry Congress, Recife, Brazil, November 1986.

Nair, P.K.R. 1987. Soil productivity under agroforestry. In H. Gholz (ed.), Agroforestry: realities, potentials and possibilities. The Hague: Martinus Nijhoff.

Nair, P.K.R. 1987. Agroforestry systems in major ecological zones of the tropics and sub-tropics. Paper for the ICRAF/WMO Workshop on the Application of Meteorology to Agroforestry; ICRAF, February 1987. (ICRAF Working Paper No. 47, April 1987).

Nair, P.K.R. 1987. Agroforestry systems inventory. Agroforestry Systems 5: 301-317.

Nair, P.K.R. 1987. Use of perennial legumes in Asian farming systems. Paper for the International Workshop on Sustainable Agriculture, 25-29 May 1987, International Rice Research Institute, Los Banos, The Philippines.