section 2
PERSPECTIVES ON AGROFORESTRY

Chapter 5
Agroforestry: a development-bank perspective

John Spears
Forestry Advisor, World Bank Washington, D.C., USA

Introduction

The period 1977-1986 has seen a sharp escalation in development banking support for forestry. The approximately US$ 2 billion invested in forestry by four of the multilateral development banks represents a 13-fold increase in the volume of lending compared with the previous decade. More important than the volume" of lending, however, is the fact that the banks have revised their earlier forest policies to give much greater support to agroforestry, fuelwood, watershed protection, forest conservation, education, training, research and extension, in addition to the more traditional industrial plantation-oriented forestry projects that accounted for 80 percent of bank forestry investments in the 1960s and early 1970s. Agroforestry lending, for example, during this last decade totalled US$ 750 million, representing a rise from 6 percent to 37 percent of total bank forestry investments.

Editors' note__________

The author has used the World Bank perspective and classification for projects, many of which were initiated before the term agroforestry was widely used. He has also included projects which are peripheral to many readers' perception of agroforestry.

Economic rates of return of agroforestry projects

In most cases the economic rates of return of the agroforestry projects financed by development banks have been significantly higher than those yielded by the industrial plantation forestry projects -that characterized earlier bank lending programmes, as illustrated in Table 1.

Table 1 Comparative rates of return for forestry projects financed by development banks during the decade 1977-1986

As has been the Bank's experience in agriculture, the message is that investing in poverty-oriented agroforestry projects is a bankable activity that requires no special justification on humanitarian or environmental grounds. Agroforestry projects can ensure increased farm productivity and income for rural people on the one hand, as well as protection of the farming environment (particularly soil and water resources) on the other.

There are several clearly identifiable reasons why agroforestry projects have demonstrated fairly high financial rates of return:

1. On-farm tree planting or local woodland protection and conservation by village communities usually requires a lower level of investment in supporting infrastructure (e.g., access roads, firebreaks and administrative staff) than would normally be required for government plantation or forest protection programmes. Particularly in small-farm situations much of this necessary infrastructure is already in place;

2. In general, farmers have been quick to adopt fast-growing short-rotation genera such as Leucaena, Sesbania, Grevillea, Calliandra, Eucalyptus and Albizia with prospects of early financial return (in more favourable ecological situations 3-7 years) or which they recognize can benefit the farming system (e.g., by nitrogen fixation, provision of leaf mulch, fodder, fruit and other products);

3. The use of low-cost tree-establishment techniques such as direct seeding, or use of the "basket" system for distributing seedlings as is widely practised in India, together with decentralization of planting among many thousands of smaller farmers has helped keep establishment costs down;

4. Prices of poles, fuelwood and timber have, in general, been rising in real terms at a faster rate than those of other commodities. A combination of increasing scarcity and rising prices for building poles, fuelwood, fodder and timber make tree farming in favourable ecological conditions a profitable venture.

The underlying ecological framework for agroforestry is important. In general, opportunities for bank support for agroforestry schemes based on farm tree planting have been more obvious in tropical countries with reasonable rainfall or surplus irrigation water and reasonably fertile soil that favours fast tree growth and early returns. Developing investment programmes for agroforestry in semi-arid or arid-zone environments is proving a much slower business because of the length of time it takes to produce results and because of land-tenure and socio-economic problems associated with ensuring sustainable common-property resource management. This is especially difficult in situations where the same savanna woodlands are under pressure for production of charcoal as well as providing a critical dry weather grazing reserve for nomadic livestock (as is the case, for example, in many of the African countries situated in the Sudano-Sahelian zone).

Table 2 summarizes the cash flow and financial rate of return for a typical cash-crop-tree-farm operation in the Philippines where several thousand small farmers have taken up cash-crop-tree farming for production of pulpwood, poles, timber, charcoal or fuelwood.

Financial rates of return to the farmer from such investments have typically been in the range of 20-30 percent. Small farm woodlots of Acacia mearnsii and Eucalyptus spp. in Kenya covering about 60,000 ha are currently yielding surplus income (i.e., over and above the value of domestically consumed poles or fuelwood) of about US$ 0.5 million a year. For many rural families this revenue represents their only significant source of surplus farm income. Most of the food produced on the farm is needed for subsistence.

The Philippines and Kenya experiences cited above relate to situations where fanners are practising cash-crop-tree fanning in discrete woodlots which are maintained under tree cover and operated on a rotational basis. The financial rate of return to the farmers' investment can be readily measured by quantifying the cash value of pulpwood, poles, fuelwood or other forest output.

High economic rates of return are also being demonstrated for agroforestry projects where trees are being interplanted with crops or included as an integral component of livestock-farming systems. In such situations the benefits from investment are based on measurement of the impact of the agroforestry tree/ crop combination on both agricultural and forestry output. Evaluation of benefits must take into account both positive and negative effects of trees on crop yield or livestock output, as well as the cash value of forest products. An encouraging feature of development banks'experience of project appraisal in the last decade has been the improved perception of techniques of economic analysis that can help to take these sometimes less obvious benefits into account.

For example, Table 3 summarizes the results of a recent development bank appraisal of the economic viability of a shelterbelt/ agroforestry programme in Northern Nigeria. If the analysis in this particular case had been confined only to consideration of wood benefits, the internal rate of return would have been of the order of 4.7 percent (which is below the opportunity cost of capital and would probably have led to the project being rejected for bank financing).

However, when the benefits of improved fodder availability, the positive impact of shelterbelts on crop yields, prevention of soil erosion and enhancement of soil fertility were taken into account, rates of return jumped to between 16 and 21 percent.

Table 2 Smallholder tree farming project in the Philippines: Summarized cash flow and financial rate of return for a 10 ha Albizia fakataria tree farmer (US$, 1980)*

Table 3 Appraisal of economic benefits of an agroforestry / shelterbelt/ soil conservation project in Northern Nigeria*

Criteria for bank appraisal of agroforestry projects

Criteria for banks' appraisal of agroforestry projects must clearly go far beyond consideration of a project's economic rate of return. Some of the other key factors taken into account include:

Sociological aspects

Putting people first is a recent World Bank publication (Cernea, 1985) that focuses on the increasing attention now being given by the banks to sociological aspects of project design in agriculture, forestry and other types of development projects. It has become a standard practice for the banks to incorporate, as a part of agroforestry project preparation, a sampling of farmer or community attitudes to tree planting and forest conservation prior to implementation. For example, farmers interviewed prior to bank involvement in a rural development project in Yemen Arab Republic had a very clear picture of the reasons why they felt trees were important for incorporation on their farmlands (see Table 4). The same survey also clearly identified farmers' preferences over choice of species in addition to potential end-use of different trees.

Clearly such preliminary surveys cannot substitute for the more in-depth diagnostic methodology such as that developed by ICRAF as a basis for designing agroforestry research programmes (Raintree, 1987). Nevertheless, sufficient information can be gained in a relatively sh&rt period to identify people's preferences in tree species, the end management objectives of tree planting and likely availability of land for it. In other words, a start can be made by investing in expansion of existing and socially acceptable agroforestry technologies pending the results of longer term" agroforestry research which, in time, will help to ensure further gains in productivity. A key point is the need for a flexible approach to project targets that permits funds to be rapidly shifted from one category of lending to another according to people's receptivity.

Table 4 Farmers' perception of the usefulness of trees in two regions of the Yemen Arab Republic

Land-tenure issues

Experience from a wide range of agroforestry projects financed by development banks in India, Kenya, Haiti, the Philippines, Thailand and Niger has clearly demonstrated that security of land tenure is a major incentive to investment in agroforestry and to protection of trees or woodlands. World Bank support for a recent project in Thailand, for example, is intended to assist the government in speeding up land titling programmes in which planting of trees will frequently be carried out by farmers as a first step in land consolidation and demarcation of their new farm boundaries.

Whilst the incentive of private land ownership can act as a powerful stimulus to farmers' interest in tree planting, it is not in any way a mandatory precondition for bank involvement. In situations of communal ownership of woodlands it is not inconceivable that various usufruct and harvesting rights could be subdivided among various claimants and satisfactory arrangements made for management of common property resources. Encouraging small-scale experiences along these lines have been reported for development bank projects financed in Nepal, Niger and Sudan. Nevertheless, as noted earlier, problems of ensuring sustained and effective management of common-property resources management are still a constraint to larger scale bank support of agroforestry in many arid-zone countries. How to overcome this problem is an issue that is receiving high priority in the Bank's research agenda.

Ecological sustainability of the proposed agroforestry farming system

The success of agroforestry projects will ultimately be judged by their long-term sustainability well beyond the life of any development-bank support. A basic principle of bank support for any agriculture or forestry investment is that it should build on an existing and well-proven, socially acceptable, technically and economically viable farming system. As has been stated elsewhere, agroforestry is not a new science invented by the development establishment. Agroforestry farming systems have been practised in Asia, Africa and Latin America for hundreds of years.

Long established, traditional agroforestry farming systems, such as preservation of Acacia albida trees in millet, sorghum or groundnut fields (a farming system practised in many West African countries; von Maydell, this volume), have provided a solid basis for expansion of development-bank support for agroforestry in several Sahelian-zone countries. In such situations, provided bank appraisal makes realistic judgements about, for example, the take-up rate of agroforestry practices, the area of woodland that can be brought under effective protection within a given time-frame, farmers' likely receptivity to credit, extension advice and other inputs, etc., the prospect that the farming system will be ecologically sustainable beyond the life of the development bank's support will be reasonably assured.

Of a more controversial nature has been the development banks' widespread support for agroforestry systems incorporating introduced species such as Eucalyptus and Albizia which have raised environmental concerns. The issue has been clouded by emotivecriticism and the debate is at this time unresolved. The main arguments advanced by the banks for their support of Eucalyptus planting, for example, include, firstly, the fact that the farmers themselves frequently specify the use of Eucalyptus species because they are well aware of their fast growth and potential high yield of building poles, fuelwood and income.¹ The farmers interviewed about their perception of the usefulness of Eucalyptus species usually also cited their unpalatability to grazing animals and the fact that the species coppices very vigorously.

Eucalyptus woodlots established by farmers have already proven themselves sustainable over long periods of time. For example, as noted earlier, spontaneously planted Eucalyptus woodlots are a common feature of the landscape throughout Western Kenya, even in the most densely populated regions. Many of these woodlots have been in existence for 50 years or more and are now into their fourth or fifth coppice rotations.

Problems have arisen with Eucalyptus in situations where larger block plantations have been established, usually by government-sponsored agencies, on the slopes of catchment areas situated in arid environments. In such situations the transpiration by the trees can cause a reduction in water yield, and if planted too close together they can accelerate soil erosion. A conscious attempt is being made to avoid bank support for such programmes. Eucalyptus is not a good tree for intercropping because of its potential allelopathic effects on crops, but those effects can be largely avoided by widely spacing trees along farm boundaries. For boundaries and small woodlot planting on poor land it will probably continue to be a preferred species for farm planting in many developing countries. Clearly, great care needs to be taken to select appropriate situations for on-farm planting of Eucalyptus and to take into account the sensitivity of agricultural crops to the possible toxic effects of some Eucalyptus species. The scientific evidence relating to the positive and negative effects of Eucalyptus planting have been spelled out in a recent FAO publication (Poore and Fries, 1985).

What are development bank funds used for?

Of the approximately US$ 750 million invested by development banks in agroforestry during the last decade, about 70 percent has been for farm forestry planting or protection and upgrading of existing woodlands, and 30 percent for a range of education, training, extension, research and institution-building activities.

In areas with favourable ecological conditions, the bank-supported agroforestry programmes have helped accelerate on-going, spontaneous cash-crop-tree farming and agroforestry by financing, for example:

1. Sociological surveys of farmers' attitudes to tree planting and of potential cash markets carried out prior to or as an integral part of the project-implementation process;

2. Decentralization of seedling production to involve more small farmers and schools in raising seedlings, and provision of seeds, pesticides, water pumps, etc.;

3. Financial support for mass-media promotion of agroforestry programmes in rural areas, e.g., by ensuring a supply of audio-visual material for encouraging agroforestry in schools and villages and by financing film strips and radio programmes;

4. Strengthening of government extension support for agroforestry planting. Special emphasis has been given to accelerated training of extension staff, with particular reference to encouraging closer co-operation between forestry and agricultural extension staff;

5. Financing of monitoring and evaluation programmes for assessing farmer or community response to agroforestry investments, measuring their positive and negative impacts on food and forest production and monitoring the distribution of project-financed inputs (such as seedlings) to try and ensure that smaller farmers or poor communities are the main beneficiaries (FAO, 1966; World Bank, 1984);

6. Strengthening support for both agroforestry education and training and forestry research with special reference to improved agroforestry research. The World Bank, for example, provides part of ICRAF's core budget and the Bank is deliberately incorporating into its project design various technologies identified by ICRAF as having potential for increasing the productivity of farming systems (e.g., introducing leguminous species in place of non-leguminous trees, selection of spacing systems that offer possibilities for improving both forestry and crop yield, and replacing slow-growing species with faster-growing multipurpose trees that can produce a wider range of products suitable for indigenous consumption).

Bank support has also been provided for strengthening of basic forestry research in national research institutions focused on high-priority topics that can lead to likely productivity gains and improved rural incopies. Jointly with the United Nations Development Programme (UNDP) and other agencies, Bank support has been made available to the International Union of Forestry Research Organizations (lUFRO)'s "Special Program for Developing Countries" which is playing a key role in strengthening regional research networks in Africa, Asia and Latin America. An example of the type of research being supported is multipurpose tree breeding and improvement programmes that can help to ensure access to improved planting stock for farmers practising agroforestry.

In arid and semi-arid countries where the conditions are much less favourable for new tree planting, the current emphasis of development bank support is shifting away from forest plantation and community woodlot approaches towards support for schemes that would encourage local and village community involvement in more effective protection and management of existing savanna woodlands. Such investments usually aim to improve technical packages for increasing the productivity of arid-zone savanna woodlands (e.g., via more effective fire protection, adjustment of timing of lopping and harvesting of trees to encourage more vigorous regrowth and encouragement of rotational grazing schemes which would permit adequate regeneration of trees in areas being rested). Bank support in countries such as Niger, Mali and Senegal has also been made available for planting trees in family compounds and around homesteads, with species for which local people have a high preference such as neem (Azadirachta indica) and gao (Acacia albida). Support has also been given to irrigated tree planting around agricultural perimeters, particularly in situations where effluent water is available and irrigated forests will not seriously compete with food crops (e.g., in the Sudan).

Examples of successful agroforestry projects

The word "success" should be viewed in this context with caution. As noted earlier, criteria for success in agroforestry go well beyond the disbursement period of a bank loan (typically five years). Nevertheless, it has been possible in several instances to record positive changes in a relatively short period of time. The three examples given below from the Philippines, India and the Sudan illustrate a range of alternative approaches to development-bank support of agroforestry projects and highlight their more successful aspects to date.

The project in the Philippines is a traditional cash-crop-tree farming operation where trees are being grown in discrete woodlots for a clearly identifiable pulpwood market. In India the National Social Forestry Project works mainly through the existing government Forest Service network in the different states. In the case of the Sudan Western Savanna Development Project the agroforestry package was financed through a combination of Ministry of Agriculture Forestry Department and local tribal institutions as a part of a wider integrated rural development approach incorporating agriculture, livestock and forestry components.

The Philippines PICOP Pulpwood Project

This programme gained significant momentum in 1972 when the Paper Industries Corporation of the Philippines (PICOP), a pulp and paper company, entered into an arrangement with the Development Bank of the Philippines (DBP) to develop a loan programme for smallholder tree farming. Loans were offered to smallholders to finance 75 percent of the costs of plantation development and maintenance. Farmers with titled property (typically ranging from 2 to 10 hectares) could receive loans at a 12 percent rate of interest, and farmers with unsecured property could receive loans at a 14 percent rate of interest. Provision was made in the tree-farm plan for part of the farm area to be maintained under food crops. PICOP continued to guarantee a minimum purchase price for smallholder production, but allowed farmers to sell wood to other outlets if they could get a better price. The scheme proved quite popular with farmers. By 1981, the programme supported 3,800 participants and covered 22,000 ha. About 30 percent of these farmers had taken advantage of the credit programme.

There were several weak areas in the project design (particularly lack of adequate support for harvesting of the wood). Nevertheless, the overall lesson that emerged was that, given adequate market incentives and security of land tenure, development-bank assistance aimed at improving availability of seedlings, extension advice and access to credit could help significantly to accelerate this type of agroforestry scheme. The key factor which triggered farmers' response was the prospect of making handsome profits from tree growing. As Table 2 dearly illustrates, financial rates of return for participating farmers in that project typically exceeded 20 percent.

India National Social Forestry Project

This project is an expansion of the previous Bank-supported social forestry programmes covering several different states. In terms of tree planting area agroforestry (mainly farm forestry tree planting) is by far the largest component (nearly half a million equivalent hectares of planting spread over some 3 million farm holdings). The farmers themselves control the choice of species and use of the product.

Most plantings are taking place around individually owned farm boundaries, bunds, around the homestead and along water channels. The main direct cost to the local forest departments is for seedling production. A substantial part of the investment is providing support for extension and monitoring and evaluation of project progress.

The project also included a component for tree and fodder-grass planting on eroded agricultural wasteland covering a total of some 40,000 hectares. The state forest departments assist in the establishment of wasteland plantations, and provide incentive payments during initial months to the farmers or communities owning the land in order to compensate them for income foregone while they work on the agroforestry plantations. Improved (grafted) orchards consisting of fruit-bearing bushes (e.g., Zizyphus mauritania) are being planted on farmers' land. Farmers sell the fruit and use loppings for fuel. As a requirement for support under the wasteland reclamation category, the Forest Department would have to ascertain that the land was seriously eroded or in imminent danger of erosion, and hence of concern for conservation, which would justify higher government investment than under farm forestry. To give planting and other technical advice to farmers, the state forestry departments make extensive use of the existing agricultural extension service.

One of the more obviously successful aspects of these experiences in India has been the very positive response to schemes aimed at decentralization of seedling production and attempts to involve farmers, schools and villagers in raising nursery stock under guaranteed buy-back contracts. Table 5, extracted from a recent Bank supervision report on an agroforestry project in Gujarat State in India, illustrates the point. As can be noted from this table the number of non-Forest Department nurseries increased 15 times during the project period and the share of seedlings produced by farmers and schools from 2 percent to 25 percent.

Table 5 Evolution of school and farmer involvement in seedling production programmes in Gujarat state, India

Sudan's Western Savanna Development Project

This project was based on a more integrated approach, including involvement of local communities in the management of natural forest and savanna woodland resources, resettlement of people in the Sudanian-Guinean zone, continued and expanded research into more drought-resistant, high-yielding millet and sorghum varieties, and fast-growing tree species suitable for drought-prone environments.

This project focused on measures to improve the way of life of some 25,000 families in the area. A settlement programme was initiated with the dual purpose of opening up presently blocked stock-routes and increasing crop production. The opening up of stock-routes was vital for the annual migration of the transhumant families and their livestock. The project included settlement and crop-production programmes with major emphasis on extension services, programme-specific adaptive research, and pilot crop-marketing schemes. Range management encouraged local populations to set aside specific areas of woodland as dry-weather grazing reserves. Livestock-development components included veterinary services, livestock-related adaptive research, and non-formal education of livestock producers. The project also included improved maintenance and further development of water facilities, project monitoring and evaluation.

The original appraisal report identifies one of the main causes of the decline in per-unit animal production and the deterioration of the range in the project area as being an increase in the cattle population beyond the sustainable capacity of the range. The report considered that deferred grazing, the introduction of improved species and establishment of browse reserves, plus improved cattle management, would be ineffective without control of grazing pressure.

The range-demarcation schemes aimed at rectifying the overgrazing situation by granting exclusive tenure of dry-season grazing areas to small groups of pastoralists while maintaining the transhumant movements of cattle during the remainder of the year. This strategy was based on the assumption that, under such a scheme, there would be.an incentive to equate cattle numbers with the long-term sustainable carrying capacity of the range. The range demarcation schemes were tested with three groups of livestock owners:

1. A small group of transhumant pastoralists who made up the total population around a small well-field;

2. A similar group of transhumant pastoralists that graze- an area within a larger community area's well-field; and

3. A group of sedentary livestock owners.

A recent evaluation of the project noted that with the introduction of the Range Improvement Programme increased livestock production was attained through grazing land enclosures and rotational grazing at Umm Belut. Herd performance was monitored over the project period. Results indicated that the density of plants on the grazing land increased. Some improvement in the calving rate was achieved together with a reduction in calf and adult mortality.

Currently the United States Agency for International Development (USAID), and a number of other donors are examining the scope for replicability of these experiences on a larger scale.

Problem areas in agroforestry projects financed by development banks

Cost recovery and incentive policies

Currently there is a wide variation in the incentive and cost-recovery policies being applied by the development banks in different regions and countries, and a lack of clear policy guidelines. In the State of Gujarat in India, for example, the World Bank has endorsed a free seedling distribution programme for the past seven years — a programme which has been widely acclaimed as having had a very positive influence on farmers' willingness to plant trees. By contrast, under a Bank-financed project in the State of Uttar Pradesh, India, a charge is levied on all seedlings. Paradoxically, there has also been a significant increase in tree planting in that State although it is not clear whether this is taking place on the scale recorded in Gujarat or whether smaller farmers are benefiting to the same extent.

In the east African region, there is a similar variation in seedling sale, distribution and cost-recovery policies. In Kenya, for example, the third phase of a Bank-financed project will raise the price of seedlings to all fanners from US$ 1 per hundred to US$ 6 per hundred over the next five years. By contrast, in Burundi, a range of seedling charges is in force based on the premise that poorer farmers located in areas remote from cash-tree-crop markets may need to be induced to plant trees by issue of free seedlings, whereas those farmers sited close to attractive cash markets for poles or fuelwood can afford to pay a higher price.

This issue of the cost-effectiveness of alternative subsidy/incentive policies for encouraging reforestation is poorly understood. The range of different approaches being tried by the various development banks includes credit, guaranteed access to markets, and guaranteed prices (e.g., the Philippines PICOP project), free or subsidized seedlings (India, Kenya, Bangladesh, Rwanda, Burundi, Zimbabwe, Haiti), and improved security of tenure (Philippines and India's Gujarat and West Bengal States). Other types of incentive such as revolving funds, tax rebates or similar fiscal incentives have not yet been tried in World Bank-financed projects although they are widely used elsewhere. World Food Programme aid has been a major factor in stimulating reforestation initiatives in many countries and the value of food-aid support for such programmes currently exceeds US$ 150 million a year. The scope for more systematic integration of food aid into World Bank-supported agroforestry planting is under review. Evaluation of various incentive approaches and of their relative cost-effectivness is the subject of a planned research project being proposed for financing by Rockefeller Foundation and the World Bank.

Forestry extension

Another major area of institutional concern has been the weakness of bank-financed forestry extension programmes. As noted earlier, a considerable shift in emphasis of forestry has taken place over the last decade towards farm forestry and other types of social forestry programme.

Dealing directly with farmers and local communities has necessitated a considerable change in the nature, organization, and attitude of traditional forest departments which, in many developing tropical and subtropical countries, have been primarily concerned in the past with-protecting and "policing" government-owned forest reserves and keeping people and livestock out of the forests.

Some of the earlier attempts by the banks to address this issue failed to make any significant impact, partly because in project design too little attention was paid to the importance of ensuring a clear link between forestry and on-going agricultural extension programmes. In response to that issue, the World Bank in 1982 assisted the Government of India to carry out an in-depth review of forestry education, research and extension with special reference to alternative options for strengthening of forestry extension. The main extension approach recommended for wider adoption was the well-proven training and visits (T & V) system that has made a positive impact on agricultural crop yields in several South and South-East Asia region countries. Individual state forestry department staff are being retrained to provide specialist technical advice to existing agricultural extension staff, who, through more systematic T & V farm visits, concentrate on a few critical impact points such as choice of species, planting, espacement, depth of planting, and improved systems of farm-tree management (e.g., pollarding or coppicing).

Attempts to transplant this experience to African countries (e.g., in Malawi, Uganda and Kenya) have so far produced mixed results and suggest the need for a more flexible approach to extension which would take into account the wide range of traditional agroforestry farming systems. Extension needs to build on farmer and local-community experiences and their perceptions of' the usefulness of trees in the farming system rather than exclusive dependence on a single formula such as the T & V system. Where the latter would appear to be of particular relevance is in cash-crop-tree farming situations where simple technological changes such as introduction of improved planting material, or changes in planting espacement and harvesting practices offer a reasonable prospect of early gains in productivity and income.

Ensuring that the benefits of bank support reach the smaller farmer and the landless

A significant proportion of the banks' agroforestry investment has benefited cash-crop-tree farmers who were well placed to take advantage of the availability of seedlings, extension advice, improved marketing opportunities, credit and other inputs. Through monitoring and evaluation and adjustments to government policies on seedling distribution, attempts have been made, with some success, to ensure that smaller fanners are assured a high proportion of the benefits from subsidized seedling programmes, but the track record is far from statisfactory.

A high priority issue for the future is how to ensure a more equitable sharing of the benefits from the banks' agroforestry investments. Experiments with "group" tree farming on government land, such as those being tried by the governments of Gujarat and West Bengal States of India, are one promising approach to involving the rural poor and landless in agroforestry.

In looking ahead to the 1990s, perhaps the greatest challenges that the development banks face are:

1. How to find more cost-effective ways of channelling funding directly to farmers, schools and local communities (i.e., with more emphasis on investment in nongovernmental local institutions and less emphasis on support for government forestry agencies); and

2. How to widen the support for agroforestry systems to give greater emphasis to intercropping situations aimed at increasing farm productivity and sustaining crop yields (i.e., in addition to providing support for cash-crop-tree farming).

A recently appraised project for Uganda, for example, will build on the success of a past non-government organization (NGO>supported agroforestry programme with part of the project funds being disbursed directly to NGOs and local communities with the technical advice and support of the government forest service. The project will cover a balanced range of agroforestry planting and forest protection activities focused both on improvement of productivity of the existing fanning system as well as generation of cash income.

Conditions in bank agroforestry lending

Perhaps the most widely misunderstood aspect of development bank involvement in agroforestry projects has been in the use of policy-based lending as a way of tackling fundamental issues such as inappropriate government fuelwood/charcoal pricing policies, the need for assurance that project benefits, will flow to poor segments of society, and for assurance that government forestry or agricultural agencies will receive adequate and timely local budget support for implementation of project activities. The incorporation of loan covenants dealing with such issues is a standard practice in all project agreements. Their formulation takes into account local sensitivities on such issues and the terms of the loan agreements are a matter for negotiation between the bank and government prior to loan effectiveness. A review of the range of policy-based interventions that have been applied in bank-funded projects over the last decade is beyond the scope of this chapter but it is important to re-emphasize that the underlying purpose behind these policy interventions is to assist in creating a more favourable socio-economic and investment climate for agroforestry. In the case of the Philippines PICOP project, for example, key loan covenants required, as a precondition for bank involvement, that satisfactory arrangements be concluded between PICOP and the farmers over price guarantees for the pulpwood produced and assurance of land title for farmers participating in that scheme.

REFERENCES

Cernea, M. (ed.). 1985. Putting people first. Washington, D.C.: World Bank.

FAO. 1966. Monitoring and evaluation of participatory forestry projects. FAO Paper 60. Rome: FAO

Poore, M. E. D. and C. Fries. 1985. The ecological effects of eucalyptus. FAO Forestry Paper 59.Rome: FAO.

Raintree, J. B. 1987. The state of the art of agroforestry diagnosis and design. Agroforestry Systems 5 (Special Issue) (in press).

World Bank. 1984. Operational guide to monitoring and evaluation of social forestry in India. Washington, D.C.: World Bank.

Footnote __________

¹ In a sample of recently completed surveys of people's attitudes to tree planting covering a range of countries with different ecological and socio-economic conditions (India, Yemen, Zimbabwe, Haiti and Thailand), notwithstanding the listing of a wide range of indigenous species which fanners wished to plant, Eucalyptus species were given a high priority by most of the fanners interviewed.