Previous Page Table Of ContentsNext Page

Let them eat trees

It was found that many development activities that promote commercial tree crops to improve hill tribe incomes and nutrition often overlook food tree crops, reasoning that with increased income a farmer will purchase additional food. It was found, however, that cash income was more likely to be spent on consumer goods rather than food. In addition, due to limited household labor, commercial tree crop production could actually result in less food production and thus an overall reduction in nutrition. The conclusion is that if food tree crops are promoted as well as commercial crops, the health/nutrition problem can be at least partially attenuated.

Careful attention to the current health and nutritional condition of the target population is needed, however. For example, promotion of citrus trees to provide C in areas where the population gets sufficient vitamin C from chili peppers in their diet is not appropriate, whereas promotion of papaya in an area deficient in vitamin A can significantly improve health/ nutrition.

Excerpted from: Robert, G. Lamar. 1987. Let them eat trees: some observations on health and nutrition in development. In: Withington et al (eds). Multipurpose Tree Species for Small Farm Use. Bangkok. Winrock International and IDRC.

Figure 1. Potential functions of trees and tree products on small farms.

The complementary role of communal and state forests in supplying large-volume demand is also shown (Source: Raintree and Lundgren 1985).

Figure 1 shows the variety of uses that can be made of multipurpose trees on small farms. Of course, not all of these uses can be maximized at the same time. There are tradeoffs between different uses of tree biomass on a given farm in a given year, but knowledge of management options can broaden the choices and allow farmers the flexibility to vary the outputs from trees in response to changing needs. The mention of "exotic" technologies like pyrolysis, gasification and anaerobic digestion highlights the importance of infrastructural supports in the adoption of new technology (e.g.sucessful biogas plants in India). Without credit facilities and extension support the efficiencies achievable by such technologies will remain beyond the reach of many rural people.

Not only are there many different functions which trees can perform, but the needed functions come in different combinations for different users and different land use systems.

Table 16. Indicative results from a selection of ICRAF's diagnostic and design case studies showing the wide range of possible roles for agroforestry (AF) in addressing the problems and potentials of different land use systems (Source: D&D Database, ICRAF).





PROBLEMS: Constraints to intensification of food production: declining soil fertility due to shortening fallows; acid soils, aluminum toxicity, vulnerability to erosion, leaching, structural degradation once the vegetative cover is removed; weed problems increased with shortened fallows. Labour shortage for clearing (a cause of shortened fallows where land is not scarce) and weeding (a consequence of shortened fallows). Competition for male labour for clearing from labour-intensive cacao. Underdeveloped homegardens (few plants, lack of diversity), hazard from free-ranging smallstock; labour constraint on gathering of distant forest products argues for domestication in homegarden. Cacao production suffers from poor shade, competition for land with food crops, labour shortage for more intensive management.

AF POTENTIALS: Selective clearing of forest (preserving useful species and shade trees for cacao) Enriched fallows (both biologically accelerated and economically enriched). Rotational hedgerow intercropping going eventually to permanent hedgerow intercropping. Mixed intercropping. Live staking for ngon climbing crop. Labour-saving tools. Fodder banks (cut-and-carry) and live fences for smallstock in homestead area multistorey homegardens. Enrichment of cacao plantations with leguminous shade trees, fruit trees and other MPTs, shade tolerant understorey crops, fertility enhancing ground covers; introduction of labour-saving tools for tree management.





PROBLEMS: Lack of capital for hiring of labour for land clearance, purchase of cattle for rapid herd buildup, fencing for improved grazing management. Low pasture productivity due to low native soil fertility after clearance; high erosion risk in steep pasture land.

AF POTENTIALS: Low input grass/legume mixtures following clearing or first cropping (as initial step) Silvopastoral system of alternate strips of timber trees and pastures along contour lines with living fences of vegetatively propagated fodder trees to protect young timber trees and compensate for reduced grazing--for erosion control, more efficient rotational grazing management in paddocks created by the strips, and cash income from sale of timber trees. Ley (shrub legume/grass) Living fences for rotational grazing.





PROBLEMS: Projected national timber deficit around year 2000. Budgetary constraints on plantation forestry. Illegal encroachment on forest reserve by individual farmers in search of land for expanded cash crop production (most of the encroaching farmers are not landless). Inadequate income of encroaching farmers from farm enterprises on own land.

AF POTENTIALS: Mixed intercropping of timber trees with agricultural crops followed by silvopastoral sheep grazing on improved pasture under trees (agrosilvopastoral). Zonal intercropping of strips of timber trees, fodder trees and agricultural cash crops followed by strips of timber trees and sheep pasture.





PROBLEMS: Inadequate cash income. Shortages of fuelwood, poles and sawn wood. Erosion and declining fertility in cropland and coffee plots. Lack of mulch material for coffee. Shortage and poor quality of fodder. Labour shortage. Soil moisture deficit in dry season. Hail damage. Lack of planting material for multipurpose trees.

AF POTENTIALS: For the coffee plot: hedgerow intercropping, mixed intercropping with high value MPTs, mixed intercropping with food crops. For the food crops plot: hedgerow intercropping, mixed intercropping with high value trees. Multistrata homegarden. Intensive fodder banks (high input with dairy, coffee/tree intercrop system). Boundary planting, high density woodlots. Trees on grass strips to stabilize river banks and for fodder and wood products.





PROBLEMS: Serious and growing fuelwood shortage for flue curing of tobacco. Deforestation associated with immigration to this frontier area. Declining soil fertility due to shortening fallow periods associated with increasing population density. Nematode problems in tobacco associated with lack of land for adequate fallow to break nematode cycle.

AF POTENTIALS: Enriched fallow for fertility, fuelwood, poles. Woodlots for fuelwood (mainly for tobacco curing), poles. Selective clearing for conservation of miombo woodland resources. Mixed intercropping for fertility, fuelwood, poles, fodder. Hedgerow intercropping for fertility, fuelwood, poles, fodder (rotational alley cropping if combined with enriched fallow intervention). Interstitial planting of domestic & industrial tree crops for subsistence, cash and local processing industries.





PROBLEMS: Resource poor land users (RP): inadequate food, shelter and cash due to low income from wage earning (irregular employment, under-employment, low wages). Medium to "rich" land users (RR): lack of development potential due to lack of cash for savings, lack of investment opportunities, insecure crop production due to erratic rainfall. Causal factors common to both groups: poor livestock production (few animals due to shortages of land, labour and fodder, low production per animal due to shortage and poor quality of fodder especially in dry season; poor marketing); low yield of cropping systems due to low soil moisture (low/erratic rainfall, high run off, poor infiltration on deep black soils, low water-holding capacity on shallow black soils, excessive evapotranspiration due to wind, late planting and weeding due

to lack of bullocks and labour); low and declining soil fertility (low inherent fertility, continuous cropping, erosion, inadequate nutrient input, weed competition); physical damage to plants due to erosion and poor drainage. Shortage of poles and timber.

AF POTENTIALS: Trees on bunds of rainfed fields for timber, fruit, fodder and soil fertility for cash and subsistence uses. Rotational woodlots (fallows) for fertility, fuel, fodder and timber. Hedgerow intercropping for fodder, fuelwood, and fertility. Agrosilvicultural systems. Sylvopastoral systems for tree products with grazing. MPTs on common land for community gathering of fruits, etc. Large-scale shelterbelts.

B.4 Locations for tree planting

As we have seen from the eucalyptus controversy, locational factors can be a key determinant of the success or failure of tree planting efforts, quite apart from the appropriateness of the functions addressed or the species chosen for the purpose. If a would-be user does not have secure tenure over the intended planting location, adoption of the tree planting innovation may be quite out of the question (e.g., as was eucalyptus farm forestry for the landless farmers of Karnataka). In such cases, either land must be made available through innovations in social infrastructure (e.g., land redistribution), or technologies must be chosen which are compatible with the existing spatial opportunities of the user; e.g., many "landless" families possess at least a house site on which they might have a small homegarden, or they may have usufruct rights on communal land. In order to ensure that the benefits of trees planted on public lands actually go to the poorest and most powerless members of a community, the concurrence of the entire community in the allocation of special tenure rights to the planters may be required, even when it is their labour that planted them.

Table 17. Landscape niches for tree planting. (Adapted from: Rocheleau gW 1989, Raintree 1989, Scherr 1987, Young 1989, Hardcastle 1987.)




Private, communal, state

1 Natural forests and woodlands


2 Forest plantations


3 Forest boundaries

Buffer zones on forest perimeters


Private, communal, state; irrigated/rainfed

4 Fallowed cropland

Near/far fields; currently under crops/ in fallow

5 Permanent arable fields

Irrigated/rainfed annuals/perennials

6 Conservation structures in fields

Terraces, bunds, drainage ditches, etc.

7 Farm woodlots

Monocultural or mixed/multipurpose

8 Orchards, groves


9 Home compounds

Sites of homegardens (incipient to intensive)

10 Farm boundaries

Including internal field borders, existing

fencerows, hedgerows, etc.



11 Rangeland

Rough grazing lands

12 Pasture

Planted or substantially improved grazing lands

13 Conservation structures on grazing land

Bunds, cutoff drains, micro-catchments, etc.



14 Riverbanks, floodplains, lakeshores

Natually occurring waterways

15 Gullies

Naturally occurring & man-made

16 Irrigation & drainage channels


17 Irrigation tanks

Tank foreshores

18 Wells and waterholes

Surroundings may experience heavy grazing pressure

19 Fishponds

Needing shade and/or feed

20 Seaside environments

Numerous subniches, including mangrove swamps



21 Off-farm gardens

Often in specially productive sites like floodplains

22 Public or shared spaces

General category, including marketplaces, etc.

23 Commons

Commonly used for grazing, gathering, hunting, etc.

24 Roadsides, pathways

A low opportunity cost site

25 Shelterbelts

Large shelterbelts may cut across many parcels

26 Periurban environments

Numerous subniches

27 Dunes

Sites for stabilization plantings

28 Oases

Favoured by high water table

One of the factors which turned a potentially successful farm forestry innovation into a social issue in the eucalyptus controversy of India--namely, the use of private land for publicly subsidized tree planting--has reared its head in other locations, as shown by the following excerpt from a paper on social forestry in Nepal.

Trees on private land in Nepal to plant or not to plant: that is the question

Interestingly enough, the two most accessible landscape niches for poor and disadvantaged members of a community are often the ones that are both closest and farthest from home, i.e. the homegarden and the commons. This was pointed out by Rocheleau (1987) in her analysis of the tree planting opportunities of African women, but if we think of what we've learned from our analysis of the eucalyptus controversy it may well apply to many other disadvantaged groups.

Homegardens as a special agroforestry niche for women

The cultivation and management of homegardens by women is a widespread phenomenon among settled groups the world over (Buch 1980, Niñez 1985). This is particularly pronounced in Latin America in areas where women do not traditionally till the land, since it provides an agricultural production niche that is seen as an extension of the home. The homegarden is often a way around taboos against tilling the main cropland, and is usually considered an extension of the home as the women's domain. Moreover, by definition such plots are location-specific to the home area, and as such are accessible to women whose mobility may be limited by custom, or by the complex logistics of mixing travel with child care, food processing and food preparation. Homegardens provide an opportunity to intensify labor inputs to increase production, without adding time away from home and within a flexible schedule shaped around other household responsibilities (Chaney and Lewis 1980).

The homegarden is uniquely suited for agroforestry projects with women. The limited plot size encourages multistoried systems, while the woman's de facto control and the permanence (or relative permanence) of the site encourage investment in tree crops and site improvement (terraces, manuring, fencing). The small plot size also implies a high ratio of peripheral to enclosed area (Rocheleau and Hoek 1984) and hence a relatively high proportion of the site production potential can be allocated to multipurpose living fencerows. The site can also be an ideal place for small livestock, such as chickens or caged rabbits, and may provide residues useful for feeding hogs or goats confined nearby, or supplementary fodder for a larger milk animal. Likewise, such an intensive small plot is often a cost-effective site for application of manure from livestock confined nearby.

While homegardens may occur within systems ranging from shifting cultivation to intensive multiple-cropping in permanent plots, they seem to be the domain of women wherever such a plot is one among many other plots available to the household, or in cases where men are almost exclusively engaged in off-farm labour. In intensively cultivated areas of acute land scarcity, the whole household may work the homegarden under the management of the head of the household, as in parts of Southeast Asia (Sommers 1978, Hunink and Scoffers 1984). In such case, the rationale for the homegarden shifts more toward labour intensification on scarce land rather than efficient multiple use of women's scarce time. Even so, these plots may have greater relative importance for women than for men, based on the distribution of labor input and on the fact that men may have alternative sources of cash income (Stoler 1978). This is also reflected in the tendency for women in Java to inherit homegardens, while their brothers inherit the rice croplands (Palmer 1978, FAO 1979).

Among pastoralists

Although the homegarden is primarily a phenomenon of settled agriculturalists, there is an analogous niche among some seminomadic pastoralists. Milk animals, young animals or sick animals may be kept by women in special small corrals close to the home, as among the Maasai in southern Kenya (Barbara Grandin personal communication, Penny Nestel personal communication, Nambombe 1984). As in the case of the homegarden, the location of these special corrals minimizes time away from home and allows for a more than usually intensive investment of time, attention and protection. Such a land use unit constitutes a reasonable basis for introduction of a fodder-based agroforestry technology around or adjacent to it. By providing high quality fodder on-site, such a technology would reinforce the ideas of intensive care, convenience to the home, and milk production for sale or distribution by women.

Excerpted from: Dianne Rocheleau. 1987. Women, trees and tenure: implications for agroforestry research and development. In: J.B. Raintree (ed). Land, Trees and Tenure: Proceedings of an International Workshop on Tenure Issues in Agroforestry. ICRAF and the Land Tenure Center. Nairobi, Kenya, and Madison, Wisconsin, USA.


Without minimizing the desperate straits in which landless and near landless families may find themselves, their situation with respect to tree planting may not be as hopeless as one might at first suppose. It certainly does them no good for project planners to assume that they are beyond help when in fact they may be the clients that have the most to gain from even a modest involvement in tree planting. What better example to illustrate this point than traditional homestead tree husbandry in Bangladesh?

Tree growing among the landless and near landless in Bangladesh

Category of Landlessness

      Description of landholding

Percentage of households


Absolutely landless with no claims to either farmland



or homestead land

Owns no farmland but claims ownership of homestead



Claims ownership to farmland of < 0.5 acre and may



own homestead

Total landless and near landless


In spite of this people grow trees

The majority of rural families have a very limited homestead and thus very limited possibilities for tree husbandry. 97.7% have a homestead less than 0.40 acres. Nevertheless, tree husbandry appears to be a universal practice in Bangladesh. A farmer visited in Nipharmari Thaña with 0.15 acre of padiland and around 0.15 acre of homestead land had 3 Mango, 5 Jackfruit, 3 Neem, 5 Chassom and 1 Guava, a few lemon trees and three bamboo groves. He is very typical in the sense that whatever space is available is used for planting trees which give fruits, biomass for mulching and energy, other products and eventually wood, which he uses or sells.

Tree husbandry in Bangladesh is an old, well-established system for sustained yield forestry production, eminently suited to local needs and resources. The farmers have a good knowledge about techniques, management and species and represent a major resource in forestry. The Forest Department's knowledge and understanding of farm forestry is presently limited. Neither is the organization of the Forest Department well-suited for extension and support activities towards the farm forestry.

Excerpted from Bo Ohlsson. 1984a. Towards a system of promoting and supporting tree husbandry and rural forestry in Bangladesh: report from a consultancy in social forestry in Bangladesh. Stockholm and Ohlsson 1984b. Composite planting in Bangladesh. ORGUT, Stockholm, Sweden.


Homestead tree growing in Bangladesh becomes even more impressive when we compare it with state forestry.

Tree husbandry compared with State forestry

The magnitude and structure of the Tree Husbandry production system - numerous small units with a large aggregate production - is based upon variety, flexibility, multipurpose objectives, decentralized skills, knowledge and control and is integrated with the farming operations. The extensive State Forest is based upon known, centralized technology and administration, large units and rigid exercise of uniform standards and norms. The two systems are geared for different types of production. Any support to the farmers must be catalytic and supportive. The Forest Department ... is not organized to handle hundreds of thousands of small "plantations" . . . this is best handled by the farmers. And the needs are where the farmers are.

In places of extremely high population density like Bangladesh there is little common land left for the landless to use, except for interstitial public lands like roadsides and canal banks. In areas where the commons is still a more substantial estate, marginal farmers and landless families often depend upon it for their livelihood. Even in places where land is not yet scarce in an absolute sense, restrictions of access to land may put women producers at a disadvantage, even though it is upon the women that the community may depend for most of its subsistence food, fuelwood, fodder and gathered plant needs. This is particularly the pattern in Africa. In such cases the commons may assume a strategically important role as a still accessible source for all these things. Many of Rocheleau's observations in the following box regarding women's use of the commons may apply ceteris paribus to other disadvantaged groups, e.g. minority ethnic groups, the landless, poor people in general.


The commons: a special landscape niche for women's production activities

While the commons is rarely the exclusive domain of women, it is often a major source of subsistence and commercial products for women. In some stages of land use intensification and landscape development it may become the major source of women's livelihood or of their contributions to the household and community (Roy 1980, Dahlberg 1981). Where land is plentiful, as among some forest dwellers, shifting cultivators and pastoralists, the forest or range land is a shared domain of men and women, with division more likely to occur on the basis of labour, expertise and security. Women are usually, but not exclusively, responsible for gathering food and fuelwood, as well as fibre, some medicinal plants and other "minor" forest and range products. They may also manage grazing or browsing animals, but are generally more responsible for the other products and activities listed above.

While men may replace their foraging activities with wage labour or intensified agricultural and livestock production, the group may continue to rely heavily on forest and range products gathered by women. Safeguarding or expanding women's tree ownership and rights of usufruct in surrounding forests and rangeland may help to prevent environmental degradation and to maintain important sources of food and other products, as well as maintaining women's status and customary rights to use and protect forest and range lands of adequate extent and quality.

As village development cycles and land use conversion and intensification proceed, the commons may become a residual domain, left to women by default. In the process of land use intensification from bush fallow to multiple cropping (Raintree 1987) womens access to and dependence on gathering areas may change substantially, and the commons itself may shift from one spatial niche to another over time. This transition is also marked by constant adjustments between use of the commons and use of interstitial niches on-farm.

In bush fallow systems, women gather from forests at the village periphery, often in combination with work trips to outlying agricultural fields of the lineage or household (Ruthenberg 1976). Many of the same products or their equivalents are also gathered from scattered fallows in various stages of development. As the length of fallow is shortened in response to land shortage/population pressure (Boserup 1970, Lagemann 1977), the diversity, quality and quantity of the products [may become scarce and]... women may intensify their gathering activities in the peripheral forest or in croplands and short fallows, depending on the distance to the forest, the type of products sought, and the terms of access to forest and farm lands.

Women may eventually find their access to household or lineage fallow lands preempted by permanent cropping, and even their access to seasonal fallows may be curtailed by multiple cropping and/or irrigation. Changes in crops and reduction of crop and weed diversity can remove valuable sources of cropping system by-products.... Unless gathering remains an activity of the landowners themselves, its products are liable to be relegated to the status of by-products with no niches of their own, unless the plants or the niche are re-defined as gatherers' property. Perhaps this is what accounts for the integration of forest form and function into traditional multistoried agroforestry systems in areas where gathering was shared by men and women, and the intensification process was guided by the traditional values of both.

Agroforestry technologies can either reinforce or ameliorate this problem (Hoskins 1983). For example, a new alley-cropping system could displace shrubs or weeds that provide leafy vegetables (Fortmann, this volume). On the other hand, enrichment planting and more intensive management of selected (exotic or indigenous) woody species in cropland and grazing lands could provide forest or range products for women from a system compatible with men's crops or pasture. This would, however, require a clear statement of women's tree ownership or rights of usufruct.

As farmland use tends increasingly toward monoculture or more uniform and intensive cropping, women may rely increasingly on distant forest or bushland as sources of gathered products. They may also opt to settle for the basic essentials (fuel and fodder) that can be gleaned from a degraded, overgrazed commons, sacrificing quality, quantity and diversity of products for proximity. Some women modify their schedules to fit the spatial distribution of resources, with frequent visits to nearby commons and occasional visits to more distant forests for special products or for goods seasonally scarce in the farmland and commons (such as tree fodder).

Commons consists also of unclaimed interstitial niches such as roadsides, gullies and boundaries. In areas where a commons does not exist as such, women often rely on the "borrowing" of stickwood, grazing privileges, and other commodity goods and services from private woodland or pasture lands of a neighbor, relative or patron. This practice is widespread in Kenya among neighbors and clan members. In Latin America the patron-client relationship prevalent in rural areas often includes such arrangements.

Summary of advantages of commons plantings for women

Locational decisions--the answer give to the question of where to plant the trees--can have a dramatic impact on the success of tree planting efforts. An equally important decision, but one that is rarely accorded due consideration, is when not to plant trees. The example in the following box illustrates the same issue raised in India regarding the planting of eucalyptus monocultures on village commons and wastelands: independent of how the benefits are distributed, the substitution of native vegetation by exotic plantations can sometimes result in a net reduction of benefits to traditional land users. This is a locational problem because it all depends on where the trees are to be planted and what is already there.

Tree planting on 'useless bushland' in Burkina Faso

Local government officials and forestry advisors selected a tract of land described as "useless bushland." They designed a plan to clear off the brush, scrub and gnarled trees, and to plant fine straight rows of fast growing exotic fuelwood species. However, neither the project designers nor the foresters had realized that this useless looking brushland fallow was actually a part of a delicately balanced indigenous agroforestry system.

Local women helped their husbands in the grain fields and raised gardens. But beyond these more visible activities, they collected shea nuts (Bul}mspermum paradoxum) from which they made cooking oil, gathered leaves and seeds essential for nutritional sauces for their starchy staple grains, searched for grasses and bark for weaving and dyeing mats and baskets, concocted home remedies from leaves, pods and roots, and let their goats browse on the shrubs and bushes in this unused looking area. Women also piled their heads high with dead branches and sticks to carry home for cooking fuel. Their children ate the nutritious monkey bread of baobab (Adansoni4 dieitata) fruit or hunted small animals in the undergrowth. Their husbands cut chew sticks (the local substitute for toothbrushes) and stripped and twisted bark into ropes. The whole family picked and ate "desert raisins" (Cissus fruits) and other fruits and nuts, and various family members earned small sums selling fuelwood or other surplus items which the bushland provided.

The land was not - as it had seemed - useless: its use was essential to fulfilling subsistence needs of local populations. With the coming of the project the land was cleared of the natural growth, and what had been everyone's land was planted for fuel for the urban market, becoming off-limits to local residents. The project plantation was burnt and residents believed the fire started because local traditional land-use rights were overridden by leaders and project managers. Residents had lost access to needed forest products essential in their indigenous agroforestry system.

Excerpted from: M. W. Hoskins. 1982. Observations on indigenous and modem agroforestry activities in West Africa. United Nations University Workshop on Problems of Agro-Forestry. June. University of Freiburg, Germany.


Sometimes the zero-option is the best option. In most cases, however, the options are not limited to a choice between either a) completely replacing the existing vegetation, or b) doing nothing. There are a whole range of intermediate options, from improved management of existing vegetation to selective clearing and encouragement of useful species, to enrichment planting with selected economic species - all of which may be used to bring about a richer mix of useful, semi-domesticated vegetation by working with rather than against the natural diversity of existing vegetational forms. Management of natural woodlands is an increasingly important topic for research in support of community forestry (Carlson and Shea 1986, Heermans 1987, Shepherd 1987).

B.5 Tree planting arrangements

Decisions on the location within the landscape of tree planting are quite distinct from those on the spatial arrangements in which trees are to be planted (geometry, spacing, rectangularity, vertical architecture), whether they are to be grown as a monocrop or in combination with other trees or herbaceous crops, whether management is to be applied to maintain the desired spatial arrangements, how the time dimension is handled, etc.

In plantation forestry, interest in spatial arrangement has generally reduced to questions of tree espacement in block plantations or linear spacing of roadside plantings. In agroforestry one is concerned with trees within a much expanded range of geometrical possibilities. Table 18 represents an attempt to arrive at a practical summary of the most noteworthy possibilities.

Table 18. Possible tree planting arrangements in space and time, including trees grown both as monocultures and in association with other components of agroforestry systems.

In Space

Zonal arrangements

Mixed arrangements

In Time

Successional systems

Rotational systems

Rotational successions

Under the heading of spatial arrangements, the basic division between zonal and mixed arrangements is a function of the degree of interaction between trees and other cropping system components within agroforestry systems. Zonal arrangements are chosen whenever the desire is to minimize the interaction between trees and other crops. The extreme example is a monospecies woodlot or forest plantation. Classical forestry trees like pinus, cupressus and eucalyptus do well in such systems and are examples of the "crop" ideotype, i.e. trees that grow well in dense monospecies block plantings.

Mixed arrangements are chosen whenever the desire is to maximize the interface and promote a more intimate association between trees and other crops. Success for the tree components in such combinations is most likely to be found with one or the other of two possible ideotypes: 1) an "associative ideotype" with exceptional capabilities for resource sharing in multispecies mixtures, and/or 2) a tree which conforms to a tall, free-standing "isolation" ideotype with a light or seasonal canopy, or possibly one which casts a denser shade but which is grown only in combination with relatively shade t-lerant understorey species (Huxley 1985, Huxley and Raintree 1983).

Classification by component combination and arrangement is one of the most common and useful ways of classifying agroforestry systems:

Table 19. Agroforestry practices classified by component combination and arrangement (Young 1989).



Spatial mixed

Spatial zoned

MAINLY OR PARTLY SYLVOPASTORAL (trees with pastures and livestock)

Spatial mixed

Spatial zoned



B.6 Management options

Aspects of crop management that relate to spatial arrangement of trees and other crops are dealt with under the heading "arrangement" in the design algorithm (see B.5). Here the focus is on management of the biomass of individual trees. Many trees and shrubs exhibit enormous flexibility of form and productivity in response to management treatments that alter their natural growth form. The following boxes summarize the most important principles and practices involved in the manipulation of tree biomass through pruning.

Principles of tree pruning

Pruning can alter tree shape, total dry matter production per tree and dry matter distribution within the tree.

Principles of shape control

Many woody perennials can be trained and pruned to a variety of shapes and sizes quite unlike those they assume in nature. The following principles apply to many trees: (a) pruning stimulates the outgrowth of dormant buds, (b) distal buds, near the cut ends, are stimulated to grow more than basal buds, (c) pruning near ground level produces longer shoots than pruning higher-up, (d) in a sense, pruning reverses tree aging: the new shoots are 'vigorous', less periodic in growth, have fewer short shoots and greater apical dominance, (e) on many species there is a tendency for new shoots, derived from resting buds, to grow vertically, (f) the highest upwardly directed shoots attain dominance, especially in nutrient deficient conditions, and left alone these shoots will reconstitute the tree in its juvenile form (reiteration), (g) bending shoots horizontally checks their elongation and stimulates buds to grow on their upper sides, especially those buds at the highest point on the arched shoots nearest the root system.

Principles regarding dry matter production

Pruning usually decreases total dry matter production per tree because leaves are removed, and root pruning has the same effect because new foliage development is checked. However, the decrease in productivity may be less than expected because there can be an increase in net photosynthesis by the remaining leaves. If only old, shaded foliage is pruned, there may be no decrease in dry matter production at all. Whenever foliage is to be removed from trees for fodder or mulch, there is an advantage in removing only old shoots, especially from closed stands.

Principles regarding dry matter distribution

There are three main principles to be observed in manipulating dry matter distribution in trees. First, the proportion of dry matter devoted to bole wood as opposed to leafy shoots depends on the size of the bole cambial sink relative to other sinks. If the trees are kept small, and branching is encouraged, a relatively small proportion of total dry matter increment will be used to produce bole wood, whereas if the trees are allowed to grow tall, and branching is discouraged (by pruning competition) a larger proportion of total dry matter increment will be used to produce bole wood. Thus, by manipulating the height and branchiness of trees, one can alter the proportion of dry matter devoted to leafy shoots as opposed to large diametered fuelwood or timber.

Second, it should be realized that the relative growth rates of the stems of trees are proportional to the relative growth rates of both the roots and crowns; that is, there are strong steam-leaf and stem-root allometric relationships, just as there are strong shoot-root allometric relationships. This means that (a) it is not possible to permanently change the growth rate of one part of the vegetative structure without simultaneously changing the growth rates of all other vegetative parts, and (b) the diameter increment and taper of the stems will depend on the size and disposition of the crowns.

Third, neither shoot nor root pruning can permanently alter shoot-cambial-root allometric relationships. Shoot pruning temporarily checks root growth, and root pruning temporarily check shoot growth, in both cases in proportion to the amount of shoot or root removed and for as long as it takes the original shoot-root allometry to be restored. Root pruning never diverts assimilates to the shoots, it diverts them to the roots. Severe lopping increases the proportion of dry matter taken by new leafy shoots, it does not alter the allometric relationships between roots, cambium and leaves.

Basic tree pruning practices

Pruning practices can be divided, for convenience, into four basic types: (a) operations that spread the trees laterally, namely, bending and coppicing, or multiple-heading, (b) operations that lop off the main stem and encourage branching [heading back], (c) operations that limit branching but leave the main stem intact, and (d) root pruning. Examples of the first three are shown below.

Both bending and coppicing increase the ground cover per tree so that fewer need to be planted. Also, the vertical shoots can be repeatedly harvested, they are rapidly replaced by new shoots, there is less need to weed than when replanting, and one has all the other advantages of ratoon cropping. Regular coppicing will also keep the leafy shoots accessible for harvesting or browsing, and by restricting the development of the boles and woody frames, coppicing will increase the proportion of dry matter going to new leafy shoots. New growth may rely heavily on root reserves. This is successful only when the individual stumps and root systems are large, implying a minimum tree size and spacing, combined with timely coppicing and a minimum coppice cycle. Trees vary greatly in their requirements and each new species has to be tested to determine the optimal pruning regime.

Pollarding and bushing are two basic lopping operations. Pollarding is no more than elevated coppicing: high pollarding will produce trees with a greater potential for wood production and the development of large individual root systems, while low level pollarding and lopping will produce spreading bushes with a greater potential for fodder production. The regrowth of new shoots will be less vigorous with increased distance from the roots. As with coppicing, the optimum cutting frequency to ensure survival and rapid recovery will depend on the species, environment and amount of foliage left. Bearing in mind that pollarded trees have more reserve capacity left after cutting than coppiced ones, trees which recover poorly after coppicing may recover better after pollarding. Pollarding itself may be preferred, for instance, to keep the new foliage out of the reach of animals, and to allow some timber and thick fuelwood production.

Branch pruning may involve simply the removal of lower branches, as in intensive forestry, thereby lessening stem taper and decreasing the number of knots in the bolewood without greatly decreasing bole volume increment. This is a highly desirable operation in forests grown for high-value timber, and yields some small diameter firewood. Alternately, branch pruning may be highly selective, as in tree fruit culture, restricting height growth, encouraging the development of spreading, shallow, well-illuminated canopies, enhancing fruit bud production and allowing some light to reach herbaceous crops beneath. In general, woody perennials require more complicated management when grown for fruits or seeds than for some vegetative product. Whenever trees are grown specifically for fruits or nuts they will generally benefit from intensive care.

Root pruning, although it checks shoot growth, apparently in direct proportion to the amount of roots removed, can have two beneficial effects: (a) enhanced flower bud production, and (b) the regrowth of a more fibrous root system, with greater access to immobile soil nutrients, and occasionally, in P deficient soils, temporarily enhanced tree growth. In the case of nursery seedlings, the more fibrous root system produced by undercutting or wrenching helps them to survive after transplanting.

Adapted from M.G.R. Cannell. 1983. Plant management in agroforestry: manipulation of trees, population densities and mixtures of trees and herbaceous crops. In: P.A. Huxley (ed). Plant Research and Agroforestry. ICRAF. Nairobi, Kenya.

In the case of trees grown in intimate associations with field crops, root pruning may also be desirable as a way of limiting tree competition with the crops, as research on eucalyptus boundary planting and leuceana alley cropping has begun to suggest (Drake Hocking, personal communication; Jonsson et al 1988; Szott 1989).

The issue of root competition and its management in agroforestry intercropping systems is extremely complicated and very poorly understood at present. For example, in opposition to the conventional wisdom on the desirability of reducing tree root penetration into the root zone of associated crops, recent work on intercropping with trees has suggested that high rooting densities of nitrogen-fixing trees may actually improve the growth of the associated crops by creating conditions which favour the uptake of solubilized phosphorus and other minerals by the non-nitrogen-fixing species (Gillespie 1989). Clearly, more research is needed before reliable extension recommendations can be formulated for many promising, but still little understood, agroforestry systems.

B.7 The technology repertoire

The following table presents a structured list of keywords, synonyms and related terms found in the "agroforestry potentials" field of ICRAF's D&D database, representing the results of diagnostic and design studies carried out in a sample of 40 land use systems thus far recorded in the database. As an empirical summary of the findings of independent D&D field teams at the planning stage of activities in agroforestry research projects, it illustrates the scope of agroforestry design considerations which arise from rapid appraisal forms of the D&D discovery procedures. This listing is indicative only and should not be regarded as a definitive list of agroforestry design considerations. Nevertheless, in combination with other reference materials in these appendices, it may serve as a preliminary indication of the potential technology repertoire of the new forester.

Table 20. Basic vocabulary of diagnostically-oriented agroforestry designers. Source: ICRAF D&D database (first 40 land use systems recorded).


building materials

by-products (various)


domestic tree crops


fertility (see soil management)


orchards of pod producing trees



industrial tree crops (for local processing)

ley (shrub legume/grass mixture in rotation with crops)

live staking

living fences

pesticide trees



soil management




boundary planting



interstitial plantings

mixed intercropping





trees on ...

upperstorey trees

zonal intercropping

windbreaks & shelterbelts (see shelter above)



domestication of gathered plants in homegardens

environmental rehabilitation

integration of other agroforestry components into forests

local processing of tree products (foods, fiber, industrial extracts, etc.) for value-adding local industry

Previous PageTop Of PageNext Page