Previous PageTable Of ContentsNext Page




All of the following case studies are fully referenced in the bibliography.


Tran, D. & Nguyen, N. Declining Productivity Gains and the Yield Gap in Rice.
Le, H. The Potential of Agricultural Biotechnology.


Brinkman, R. Development in the Moist Savanna Zone of West Africa.
Mitti, G. Community-based Seed Supply Systems, Zambia.
Mascaretti, A. Tassas - Improved Traditional Planting Pits in Niger.
Tanner, C. Community Based Land Tenure Reform in Mozambique.


Batello, C. Farming Systems in Arid Rangelands of Syria and Jordan.
Bazza, M. Improved On-farm Participatory Water Management to Reduce Mining of Groundwater in Yemen.
Fe'D'Ostiani, L. Participatory and Integrated Watershed Management in Upland Areas of Tunisia.


Meng, E. Advisory Services for Restructured and Co-operative Farms in Extensive Crop/Livestock Regions.
Kopeva, D. Transfer of Ownership and Land Fragmentation in the Transition to Family Farms in Bulgaria.
Martinenko, A. Natural Resource Use and Economic Viability as Affected by the Transformation of Farming Systems in Southern Ukraine.


Agarwal, A. Water Harvesting and Soil Rehabilitation in India: Potential and Practice.
Hoque, F. Integrated Intensified Rice Farming Systems in Bangladesh.
Kiff, E. & Pound, B. Integrating Crop Livestock Interactions into Hill Mixed Farming Systems, Nepal.
Dugdill, B.T & Bennet, A. Income Diversification in an Intensive Rice-Based System - Milk Vita in Bangladesh.


Ishihara, Y. & Bachmann, T. Stablizing High Altitude Swidden Cultivation in Laos.
Wang, Z., Innovative Rice-Based Farming Systems Development: Zhejiang Province, China.


De Grandi, J. Improved Income Generation among Small Farmers in the Peruvian Sierra.
Gulliver, A. Private Sector-Led Diversification Among Indigenous Producers in Guatemala.
Spehar, C. Improvement of Farming Systems in the Savannah Lands of Brazil.


Those case studies shown above in bold have been selected for inclusion in a condensed form in this volume, on the basis that they deal with key issues of more than regional importance to farming systems development and poverty reduction. The two global studies examine issues related to intensification of production (declining productivity and rice yield gaps), and the possible contribution of science and technology (biotechnology). The selected regional studies consider expansion of farmed area (opening up of moist savannah areas in West Africa), improved natural resource management (water harvesting in India and tassas in West Africa in a combined presentation) and the process of market development and diversification (private sector-led diversification in Central America).

All 22 case studies can be found in full, in electronic form, on the FAO web-site devoted to Farming Systems at:




The Green Revolution enabled rice production to satisfy the rising demands of many growing populations. Since 1990, however, the growth in rice production has slowed down causing concern in terms of food security, particularly among the rural poor. Studies have been conducted to identify the causes of this slowdown and the potential interventions that might support a sustainable increased rice production and may contribute to the alleviation of poverty in rural areas.

Land and water resources for rice production, especially in Asia, have been under increasing pressure from urban and industrial expansion. Improvements in productivity and efficiency would thus appear to be essential for increasing rice production in the future. Low national rice yields and large yield gaps are found in many countries but some countries have succeeded in closing most of the gap between their national yields and the potential yields. This suggests that closing the yield gap might be one possible means of increasing rice productivity. However, it is accepted that the reasons for low rice yields are due to a complex set of biophysical, technical, socio-economic and policy issues which operate in different ways in each major rice-producing country. The relative success of support measures for rice production in many countries over the past 30 years has led to surpluses of rice traded internationally followed by unfavourable terms of trade and consequent reductions in the prices received by farmers. This factor alone could be having a major impact on farmers' willingness to invest in yield-enhancing technologies.

At present, there are about 50 million hectares or more under intensive, irrigated rice-rice, rice-wheat or rice-other crop production systems. Yield and productivity declines which have been observed in these intensive rice production systems threaten, not only sustainable rice production, but also, indirectly, the incomes of small farmers.


The average annual growth rate (AGR) of the world's rice yield was about 2.2 percent during the 1962-1970 period, but decreased to about 1.6 percent between 1971 and 1980 and then increased to about 2.3 percent during 1981-1990. The AGR of the world's rice yield during 1991-1998, however, was only one percent. Rice is grown under a wide range of agro-ecological conditions from irrigated lowland to freely drained upland; from temperate to tropical climates; and from land influenced by tides (mangrove rice) to land flooded with a water depth of several meters for a considerable period (deepwater rice). Yields vary with the agro-ecological conditions during growing seasons. In 1999, the national rice yields in 80 countries were lower than the world average of about 3.8t/ha. In the same year, the gap between the world's highest and lowest national rice yields was about 9 t/ha (Table 1).

More than four-fifths of the world's rice is produced and consumed by small farmers in low-income and developing countries. In 1996, nearly three billion people depended on rice as the major source for daily calories and protein.

Table 1. World Rice Yield, Highest National Rice Yield and Lowest National Rice Yield



World rice yield

3.84 t/ha

World's highest national rice yield

10.07 t/ha

World's lowest national rice yield

0.57 t/ha

Number of countries with national rice yield below the world average rice yield


Source: FAOSTAT.

Rice and rice-based production systems provide not only food, but also the main source of income and employment opportunities for about a billion poor people in rural areas of Asia, and for smaller members in Africa and Latin America. Small-scale and resource-poor are in the majority in developing countries. They often do not have access to adequate amounts of inputs, especially fertilisers, at the right time for obtaining high yields, as input supply has not been adequately decentralised to village markets. Also, small farmers are generally unable to buy sufficient quantities of fertilisers and to cover other expenses for field operations due to lack of credit. Research and extension support, which is essential to ensure the effective reduction of the gap and improvement rice yield and productivity, is also not always available either.

The intensive important rice-rice systems are common in tropical climate areas, whereas intensive rice-wheat systems are dominant in the sub-tropical zone. Most of the irrigated rice farms in Asia and Africa are small, and farm households are generally poor.

In non-irrigated areas, rice is planted under rainfed conditions during the wet season and other subsidiary crops are planted during the dry season. Rainfed wetland production systems are dominant in Asia, while in Sub-Saharan Africa and Brazil, rainfed upland rice production systems are dominant. These upland rice production systems, however, are not stable and farmers in Sub-Saharan Africa and in Brazil are increasingly developing wetland rice production systems.

Rainfed rice farmers, both wetland and upland, are generally poorer than the farmers in irrigated areas due to the lower and less stable yields under these production systems. Improvement in the productivity of the rainfed rice production systems will require new varieties and technologies. Recently the West African Rice Development Association (WARDA) developed new upland rice varieties, called NERICA, for use in West Africa and elsewhere.

Yield and productivity levels in rice production systems, especially irrigated ones, could be substantially improved using current technologies, but with appropriate crop management. A UNDP/FAO project in Bangladesh (BGD/89/045 Thana Cereal Technology Transfer and Identification), for example, found that a specific combination of research and extension support to farmers has increased rice yield and reduced production costs. Recommended technological packages, with emphasis on Integrated Crop Management using currently available technologies, were systematically transferred through effective training and demonstrations with the participation of farmers.


There are about 50 million hectares or more under intensive rice-based cropping patterns (rice-rice, rice-wheat and/or rice-other crop), producing two to three crops per year. Average yields are about 4-6 t/ha per crop or about 10-15 t/ha per annum. In several Asian countries such as Indonesia, China, Viet Nam and Bangladesh, farmers grow two to three rice crops, using early-maturing varieties of 90-100 days. In China, farmers grew three rice crops in the early years, but now they are changing to two rice crops and a short-cycle winter crop such as maize or beans. This is due to the fact that the three-rice cropping systems are time-consuming and labour-intensive, but their returns are insufficient to justify the inputs and efforts.

Declining yields and productivity have been observed in several intensive rice production systems. In intensive cultivated areas in the Chiang Mai valley, Thailand, yields declined from 7 to 4 t/ha under normal crop management for unknown reasons (Gypmantasiri et al., 1980). In India, yield declined over 10 years from nearly 6 t/ha to 2-3 t/ha in the Rabi dry season crop (Nambiar and Ghosh 1984).

At the International Rice Research Institute (IRRI) farm at Los Baños Luzan, in the Phillipines, rice yields declined from 8 t/ha in 1968 to 6 t/ha in 1990, while the fertiliser rate remained constant (Greenland 1997). In three other locations in the Philippines, rice yields have been declining by 0.1 to 0.3 t/ha per year over a 20 year period (Cassman et al., 1997). Ladha et al., (2000) reported a declining rice yield in four out of eight experimental sites on long-term rice-wheat systems in the Indo-Gangetic region. Wheat yields remained more or less unchanged or increased slightly in the same experimental sites. In Africa, sharp declines in rice yields in irrigated schemes after a few years of intensive cultivation have been reported in Tanzania (Duwayri et al., 1999), and Burkina Faso, Cameroon and Nigeria (Fagade and Nguyen 2000). Dawe and Dobermann (2000) also observed yield and productivity declines in intensive rice-rice systems, but indicated that these occurred only in limited locations.

In the long-term experiments where yield decline was observed, yield was restored with increased N fertiliser. This suggests that the indigenous N supply of the flooded soil-water system had declined over time (Cassman et al., 1997). However, the number of long-term monitoring sites is limited. Hence, the soils and soil types on which they are located may not fully represent all rice soils. Futhermore, the causes of the yield declines that have been documented vary from one location and ecosystem to another. Often, the causes are not completely understood due to the lack of appropriate measurements. Besides a declining soil nitrogen supply, several other factors - such as zinc deficiency, phosphorus and potassium deficiency, increased pressures of insects and diseases, erosion in yield potential of new genetic materials and changes in the chemistry of rice soils under prolonged submergence - were cited as responsible for yield declines in intensive rice-rice cultivation (Pulver and Nguyen 1999). Similarly, FAO/IAEA (2001) lists decline in organic matter, decreased nutrient supply capacity, declining soil quality or health, micro-nutrient deficiencies and nutrient imbalances as important soil-related problems in rice-wheat systems.

There is little doubt that intensified cropping has profound effects on the biological and chemical processes in rice soils. The availability of modern early-maturing varieties and low cost of nitrogen fertilizer have stimulated crop intensification, especially under irrigation. Insect and disease pressures build up under such cropping systems. Intensification and imbalanced fertilizer application may also lead to depletion of meso- and micro-elements in soils. In Indonesia, Bangladesh, and the Philippines, the drastic reduction in sulphur supply by the increased use of concentrated fertilizers (replacement of ammonium sulphate by urea) and intensive rice cropping has caused increasing S deficiency (Ponnamperuma and Deturck 1993). Also, intensive paddy cropping usually tends to prolong anaerobic conditions in the soils. At the moment, our understanding of organic matter management has not yet kept pace with the intensification of rice cropping systems.

There are also important economic and social factors which have affected smaller and poorer farmers' willingness to continue maximising yields of individual and sequences of crops. These may relate to the availability of ,and returns to, labour, land, water and draught power or mechanisation, access to irrigation and the management of irrigation structures and water, declining farm-gate prices and threats to social capital from modernising interventions.

The Expert Consultation on Yield Gap and Productivity Decline in Rice Production in Rome, September 2000, discussed yield declines2 in three distinct systems, each with distinct causes. In irrigated African systems, yield and productivity declines appear to be due primarily to deterioration of infrastructure and management problems; in Asia due degradation of the natural resource base. In upland production systems, falling yields over time are due to missing nutrient and shorter fallow periods because of population growth.

In tropical Asia and Africa, the intensive rice production systems provide income and employment opportunities not only to rice farmers but also to the landless in rural areas. The yield declines, which were observed in limited cases, may spread to wider areas under intensive cultivation in the next 30 years, substantially increasing the number of poor people in rural areas of Asia and Africa, unless these problems are addressed within an integrated national development programme that gives priority to the agricultural sector by improving access to inputs and new technologies.


Yield gaps of rice are considered to have at least two components:

Figure 1. Components of Yield Gaps (Adapted from de Datta, 1981)

It is widely recognised that, of the various strategies to achieve the production growth needed to raise and sustain yield levels in rice, the most practical short-term strategy is the realisation in farmers' fields of a large proportion of the presently available genetic yield potential. This requires assessment of the yield gap, identification of key technological, institutional, socio-economic and policy constraints, and determination of appropriate remedies.

Attempts to achieve a reduction of the rice yield gap have been advocated by many scientists and developers as a means of increasing rice production with existing technologies. However, despite this approach, such an analysis has not led to a significant of the gaps'. Social scientists have consistently required technical scientists to examine the wider picture, bring in other social and economic factors and develop a more systemic approach to the analysis. For example, it well documented that in any given rice growing area, yields vary greatly among farmers, suggesting considerable variation in farmers' knowledge, priorities, economic and social conditions which affect their decisions about allocation of resources, not only with respect to rice production.

It has been recognised that only a part of the total yield gap can be remedied by currently available technologies. Policy environment and interventions are considered vital components of the strategy to bridge these differences. Technology exchange and learning groups within research-extension-farmer partnerships play an equally important role.

Factors contributing to the yield gap

There are several groups of key constraints contributing to the yield gap, ranging from the biophysical to the institutional.

The work of Ramasamy (1996) indicated that yield gaps in Southern India are due to diverse factors: physical (problem soils, nutrient deficiency and toxicity, drought, flash floods, temperature stress); biophysical or management-related (varieties, weeds, lodging, imbalance in fertilizer application), and, particularly, socio-economic (labour shortage, cost-benefit, farmers' knowledge, skills and others).

During the next three decades, farmers' knowledge will also be continually improved and practical innovative technologies will become increasingly available, which should lead to a reduction in one component of the yield gap of rice. However, as the yield potential of rice plants will also increase through the above-mentioned emerging technologies, a yield gap will inevitably continue to exist, but may be smaller than at present, particularly in countries with strong national agricultural research systems which can engage farmers more effectively in the research process. Most importantly, scientists and extensionists need to understand the context of different rice-based systems and accept that rice, even though it may dominate farm production, is but one component in complex livelihood system.


The food security and income situation of small rice farmers in the world, especially in South Asia and in Sub-Sahara Africa, might be substantially improved by reversing the declining trend in productivity. Appropriate government policies aimed at improving the supply of inputs, credit and prices to farmers as well as infrastructures in irrigated rice systems will be a vital part of the future efforts. These measures also need to be complemented by improvement in farmers' knowledge of, and access to, more sustainable crop management practices.

To reverse the productivity decline

To reduce the rice yield/productivity gap

It is recognised that none of these measures will be effective without an enabling policy environment including support services, marketing and equitable terms of trade. In addition, Governments inevitably face the need to balance the demands of growing urban populations for adequate supplies of rice at a reasonable prices, with the need to earn foreign exchange from export markets, as well as responding to the needs of rice growers for adequate rewards and incomes from their efforts.


This paper suggests that rice production systems are facing two key problems;

The first relates to the differences that have been acknowledged for many years between the high rice yields that are known to be possible on research stations and the generally lower average yields found on many rice-based systems3. It is the thesis of this paper that this `gap' can be reduced through the application of a combination of technical, social, economic and policy measures and partnerships between many stakeholders - researchers, farmers, extensionists and planners.

The second relates to the evidence for the apparent stagnation and sometimes, declining, yields and productivity of rice in many areas where yields have been high for many years. The reasons for this are complex and need further investigation in many situations, but it is also suggested that these declines can be reversed though a better understanding of the social, economic, biophysical and technical context of rice-based systems and their interaction with their wider environments and by the combination of the measures referred to above.


The Potential of agricultural biotechnology4


FAO estimates that, over the next 30 years, more than three quarters of the growth in crop production that is needed to satisfy increasing food needs, will have to derive from increases in crop yield. This will only be possible if substantial technological innovation takes place. Modern biotechnology tools of recombinant DNA, including genetic engineering, offer some opportunities for generating such innovation.

Although the use of genetically engineered drugs and vaccines has not stirred much controversy, the deployment of genetically modified (GM) crops has met with fierce resistance, particularly in Europe on ethical grounds and on concerns of perceived negative impacts of GM crops on the environment and food safety. Ethical considerations revolve around topics such as the `unnatural' nature of gene transfers across species; possible socio-economic impacts of widening the gap between the rich and poor farmers and countries and the fear that agricultural biotechnology will increase the dependency of global food supply on few multinational corporations controlling the seed industry. With all the negative publicity of GM crops, there are concerns that the resistance to GM crops by consumers in Europe may have hindered the transfer of this new innovation to the developing countries where increasing crop productivity is most urgent.

Despite the resistance which has arisen among some groups, biotechnology is already changing the way in which the many necessities of life - food, feed, fibre, fuel and medical drugs, are being produced. In the agricultural arena, biotechnology tools have been used for animal and plant disease diagnostics, for production of recombinant vaccines against animal diseases and for the improvement of livestock and crops. Cultivation of GM crops has grown from two million ha in 1996 to 44 million hectares in 2000 (James 2000) with the bulk of transgenic acreage in three countries, namely, the USA, Canada and Argentina (UNDP 2001). Although land under transgenic crops in developing countries has increased steadily from 15 percent in 1999 to 23 percent in 2000, this increase was mainly in Argentina at 10 million ha. Had Argentina's acreage not been accounted for, this would have left the developing countries' transgenic acreage at less than one percent. Apart from Argentina, significant commercial plantings in the developing world are limited to China, Mexico and South Africa. In addition, in only a few crops are GM varieties yet of commercial significance, principally soybean, maize, cotton and oil seed rape, potato, squash and papaya.

The lack of support for the development of GMOs in the fight for food security has not gone unnoticed. In June 2000, seven National Academies of Sciences released a report calling for a concerted effort by all sectors, public and private, to develop GM crops - especially food staples - that benefit consumers and poor farmers, particularly in developing nations. It called for the sharing of GM technology developed by private corporations for use in hunger alleviation and to enhance food security in developing countries. The report also proposed special exemptions for the world's poor farmers to protect them from inappropriate restrictions in propagating their crops (NAS 2000).


Broadly speaking, biotechnology involves the use of living organisms for human benefit. It consists of two components: 1) Tissue and cell culture and 2) DNA technologies including genetic engineering. Both components are essential for the production of GM plants and animals.

Plant tissue and cell culture are relatively low cost technologies which are simple to learn, easy to apply and widely practiced in many developing countries. Plant tissue culture aids crop improvement through a range of actions, including: (a) mass propagation of elite stock; (b) the provision of virus-free stock through in vitro culture of meristem; (c) the selection and generation of somaclonal variants with desirable traits; (d) the overcoming of reproductive barriers and the transfer of desirable traits from wild relatives to crops by widecrosses; (e) the facilitation of gene transfers using plant protoplast fusions; (f) anther culture to obtain homozygous lines in a breeding programme; and (g) in vitro conservation of plant germplasm.

The second component of biotechnology, i.e. DNA technologies, including genetic engineering, utilises newly emerging knowledge of the genes and the genetic code to improve crops, trees, livestock and fish.

Important applications of DNA technologies include the use of DNA probes, specific to pathogens and pests, for their identification, monitoring and control. DNA-based markers are particularly useful for gene map construction for gene isolation. This non-controversial technology is being used to enhance efficiency of conventional plant breeding programmes and to characterize genetic resources for their conservation and use.

One of the important features of DNA transformation is the ability to move genes even across kingdoms helping to enlarge the gene pools for all organisms, including crops. Though controversial for the time being, genetic engineering allows useful genes from any living organism to be transferred to crops or animals for improving their productivity. Genetically altered bacteria or trees can be used in soil remediation. Furthermore, biosynthetic pathways can also be manipulated to produce added nutritional compounds in crops, high value pharmaceuticals and other polymers, using plant as bioreactors. The few examples of technologies present today only vaguely present the vast implications for potential importance of biotechnology on agriculture in the next two decades.

Although GM crops currently available target only simple traits and single genes, technological advances now permit the transfer of as many as 12 genes into a single plant genome (Zhang et al., 1998), although not all genes have been expressed. This may permit the manipulation of more complex, but also more valuable, traits such as yield and tolerance to drought, salinity, heat, chill and freezing, as well as tolerance to problem soils such as salinity and aluminum toxicity. In addition, great progress is being made in using GM crops to produce vaccines at low cost and which are suitable for storage conditions in developing countries (Artzen, 1995, 1996; Landridge 2000).

Research into the physiological and biochemical basis for abiotic tolerance has been greatly aided by advances in molecular biology. Japanese (Kobayashi et al., 1999) and American researchers (Jaglo-Ottosen et al., 1998) have independently isolated a transcription factor that when being over-expressed in GM plants resulted in significant tolerance to drought, salt and freezing stresses. Through an entirely different mechanism, tolerance to salinity was achieved in GM Arabidopsis and tomato engineered to over-express a vacuolar Na+/H+ antiport gene (Apse et al., 1999; Zhang and Blumwald 2001). Improving stress tolerance to enhance yield stability while remodelling photosynthesis by genetic engineering may increase crop yield potential. The transfer of photosynthesis genes from maize to rice experimentally resulted in an increase of 35 percent of rice yield as compared with lines of similar genetic background (Maurice et al., 1999).

In forestry, progress has been made in tree engineering to produce wood with reduced lignin contents for the pulp and paper industry. In addition, research is being made to manipulate genes involved in floral development to produce non-flowering trees, thus improving wood productivity (Rick Meilan, pers. comm.).

The knowledge of genes, of clustering of genes with similar expression patterns and their order in Arabidopsis, a dicot, and rice, a monocot, may be used to isolate and characterize the corresponding genes and to understand gene order and expression patterns in other crop plants (Somerville and Somerville, 1999). This knowledge coupled with opportunities to move genes across species barriers, broadens crop gene pools, which has not been possible, or with tremendous difficulties, using conventional approaches. Agricultural biotechnology, particularly that of crops and trees, is benefiting greatly from Arabidopsis and the rice genome sequencing projects.


There have been attempts to classify countries into categories, based on biotechnological capacity, useful for appropriate developmental assistance (Byerlee and Fischer 2000). The gene revolution, as it has been known, started from developed countries and spread to developing countries such as Argentina, Brazil, China, India, Mexico and South Africa. These countries have not only been able to take advantage of technology "spillovers" but also have become technology developers. They have well-established traditional plant breeding programmes and expertise in plant tissue and cell culture. It is, therefore, feasible to build expertise in recombinant technologies on already strong foundations. This is because tissue and cell culture expertise is necessary for DNA transformation. The transformed cells, with desired genes, need to be regenerated into whole plants and evaluated for stable gene expression at acceptable levels in subsequent generations. Although progress made in pollen transformation may make the tissue culture phase unnecessary (Burke et al., 1999; Saunders et al., 1997), a functional plant breeding programme is a prerequisite for any crop biotechnology programme. The GM plants still need to be tested for adaptability and performance under local conditions prior to large scale commercial release.


There have been some national and international initiatives in crop biotechnology. Recently Kenya started field testing of GM sweet potato engineered with the coat protein of the feathery mottle virus (Wambugu 2000). In South Africa, sophisticated research in molecular biology has been in progress for more than 20 years. Recently, work has been carried out to isolate and characterize genes from the resurrection plant Xerophyta viscosa, that are functionally important in osmotolerance with the objective to engineer crop plants which would exhibit greater tolerance to envronmental stresses such as drought. South Africa already has biosafety legislation that allows commercial cultivation of several GM crops such as Bt-cotton and Bt-maize5. In both Kenya and Zimbabwe, recombinant vaccines are being used against animal diseases with great success. Interestingly, while European's reaction to genetic engineering has not been enthusiastic, government officials in Africa, particularly in Nigeria and Kenya, have expressed their wish to have access to biotechnology (Amadu 2000).


In Asia, regulations are in place for field-testing and approval of GM crops in China, India, Indonesia, Japan, Philippines, Taiwan, and Thailand. China has the largest area under GM crops (Teng 2000), having released GM virus resistant tobacco and tomato varieties in 1990. Since then, China has carried out 31 GM field trials, mostly for virus resistance, with canola, cotton, potato, rice and tobacco. Contained experiments with GM crops are being conducted in India, Philippines, Thailand, Vietnam, Singapore, Malaysia, Indonesia and China, while field experiments with GM crops are underway in Indonesia, Philippines, Thailand, India and China.

Thailand has been successful in using molecular approaches in diagnostics and control of virus disease of shrimp. The Philippines government has recognized biotechnology as a major strategy to increase agricultural productivity but the field testing and commercial use of GM crops has been constrained due to public concern (De la Cruz 2000). In India, government support for biotechnology has been extensive and the country has an impressive research capacity. Nevertheless, there has not been any commercial release of GM crops in India. Sri Lanka has recently banned the import of all GM crops and food.

Middle East

Within the Middle East, Egypt, with its Agricultural Genetic Engineering Research Institute (AGERI), is probably the most advanced country in applying biotechnology. It has conducted field testing in West Asia and North Africa for a large number of GM crops. Thanks to the collaboration with the private sector under the USAID-ABSP administered by Michigan State University, AGERI isolated and patented its own indigenous Bt-genes - a successful case of private and public sector collaboration between developing and developed countries (Lewis 2000).

Latin America and Caribbean

This region has benefitted greatly from the FAO-sponsored REDBIO network which incorporates 619 laboratories in 32 countries within a single information network. The region also boasts a number of countries with advanced biotechnology capabilities, such as Mexico, Brasil, Argentina, Costa Rica and Cuba. Research conducted in Mexico on the molecular biology of aluminum tolerance in plants has great potential for the development of GM crops productive on such problem soils. Since 1998, Mexico has evaluated GM corn, cotton, potato and tomato in the field. It is an interesting case study for biosafety as the country is the centre of origin of maize. Mexico has also had commercial planting of Bt cotton. Although Brazil has been the front runner in research and development as well as in field testing of GM crops, a recent court case that bans their cultivation or testing may slow Brazil's earlier progress (Amstalden Sampaio 2000). Brazil is the first country in the developing world that completed the genome sequencing of a bacterial strain of Xylella fastidiosa, that attacks orange trees. This sequencing project was financed in part by the Brazilian orange industry (Simpson et al., 2000; Yoon 2000). Costa Rica is using biotechnology to characterize and conserve biodiversity. The country is also seeking to reduce toxic chemical use in controlling banana diseases (Sittenfield et al., 2000). Cuba offers an excellent model for the developing countries in the applications of biotechnology for agricultural development and medical biotechnology. A dozen GM crops have reached laboratory and/or field testing. Cuba has refrained from using GM tobacco, however, for fear of endangering its cigar and tobacco export industries (Lehman 2000).


High returns to GM adoption

Although there have been arguments that the first generation of GM crops, concentrated on input and simple traits designed for developed countries' industrial agriculture, may not benefit small farmers in developing countries, there is increasing evidence that this may not be the case. The cultivation of GM crops in some developing countries with high capacity in biotechnology, demonstrates that it is already making an impact in them through reduced pesticide costs and risks of poisoning, environmental benefits and productivity gains.

In China, Pray et al. (2000) has presented evidence of higher economic returns to small farmers who planted Bt-cotton, and who required less hospitalisation due to pesticide poisoning, than those cultivating non-Bt cotton. The use of Bt cotton has reduced pesticide use by 80 percent in Hebei Province in China. Since pesticide use in cotton accounts for 25 percent of global pesticide consumption in crops, this has great environmental and health benefit. South African's experience has shown that small farmers can also benefit from Bt-cotton. The number of small farmers' participation in the cultivation of Bt- cotton in this country has increased from 4 to 400 in just 4 years, indicating that they are realising the benefit in growing GM crops (Webster 2000). In Kenya, it has been projected that two sweet potato biotechnologies, GM virus and weevil resistance, will generate annual gross benefit of US$5.4 million and US$9.9 million respectively. Due to the semi-subsistence nature of sweetpotato, the producing households will be the main beneficiaries. The high efficiency of the research projects is confirmed by significantly positive returns on their investments (Qaim 1999).

In Argentina, the high adoption rate of GM crops show that they have already had an impact there. The illegal smuggling of GM seeds from Argentina to Brazil for cultivation indicate that Brasilian farmers, largely commercial growers, appreciate the benefits of GM crops over conventional ones. In Mexico, cultivation of Bt cotton for two years has resulted in an estimated of US$5.5 million in economic surplus, if which about 84 percent accrued to farmers in and 16 percent to seed suppliers (Traxler et al., 2001). In Cuba, the strategy to apply knowledge-intensive biotechnology is promising to give high pay off in terms of royalty from proprietary technologies. Cuba is developing tool kits for plant disease diagnosis. Cuban's genetic-engineered vaccines against cattle ticks and against an enterotoxic E. coli, have already been sold in international markets and reduced its pesticde imports (Borroto 2000). Its production of a patented bionematicide will allow the reduction of toxic nematicides used in banana plantation (Lehman 2000). The use of recombinant vaccine against cattle tick has reduced Cuba's pesticide imports from US$ 2.5 to only 0.5 million annually (Borroto 2000).

Biotechnology tools are also being used to investigate the mechanism of apomixis in plants for its potential applications in agriculture6. This important trait has enormous potential impact if the technology can be made available to resource-poor farmers who could replant hybrid seeds which retain permanent hybrid vigour in apomictic hybrid varieties.

Experimental work resulting in yield increases of 10 to 35 percent for GM rice using photosynthetic enzymes derived from maize (Maurice et al., 1999) and a quadrupling of yields for GM rice in which a gene has been transferred from barley providing a tolerance to low soil iron in alkaline soils (Takahashi et al., 2001), suggest that considerable further gains are possible.

Improved nutritional and medicinal quality

With 800 million malnourished people in developing countries, there is considerable scope for nutritional genomics which use metabolic engineering to manipulate plant micronutrients for human health (DellaPenna, 1999; Li and della Penna 2001). These have been termed nutraceuticals. And although their production may initially be focused on wealthy consumers in the developed world, genes can be engineered into crops cultivated and consumed by poor farmers to improve their dietary requirements. An existing GM derived variety is that of Golden Rice which contains enhanced vitamin A content in the rice endosperm (Ye et al., 2000). Food quality can also be improved by eliminating the production of certain substances. GM cassava offers reduced levels of cyanogen glycosides (Sayre 2000), thus avoiding toxicity problems associated with current varieties. Currently, there is evidence that GM Bt-crops play an important role in providing safer food than that of traditionally-bred crops through the reduction in mycotoxins produced by fungi infection through insect attack. For crops that are used as bioreactors to produce drugs, the genetic use restriction technologies (GURTs), including the so-called terminator technologies, may be very useful in preventing the contamination of the environment with the drugs and/or vaccines through gene leakages by restricting unwanted transgene flow. The availability of inexpensive, edible plant vaccines against diseases endemic in developing countries such as hepatitis B, cholera and malaria, gives poor people a chance to lead a healthy and productive life. Hopefully, edible plant vaccine against AIDS may one day be developed.

Utilisation and rehablitation of marginal and degraded lands

Considerable areas of land exist in many developing country regions that are unsuitable for agriculture due to to soil and related constraints, other areas are utilised, but yields are sub-optimal. Work by Mexican researchers in elucidating the molecular mechanism of aluminum tolerance and developing GM plants resistant to this toxic ion could have a great impact on the exploitation of acid soils in developing countries (De la Fuente et al., 1997), particularly the opening up to more intensive cultivation of vast areas in the Brazilian Cerrados and West African moist savannah. Since acid soils occur to some extent over some 43 percent of all tropical areas, aluminum tolerant crops would help to extend crop production in many zones. A further 30 percent of cultivated land is alkaline, making iron unavailable for optimum crop production. Japanese workers recently demonstrated that a GM rice, engineered with barley genes, showed an enhanced tolerance to low iron availability and yielded four times more than non-transformed plants in alkaline soil (Takahashi et al., 2001). Encouraging results are also being made in the area of salinity tolerance. In the presence of 200 mM NaCl, GM tomato and canola plants reached maturity with very good fruit set and oil quality, respectively (Apse et al., 1999; Zhang and Blumwald, 2001 and Eduardo Blumwald, pers. comm. 2001). In addition, climatic variability such as sudden drought or frost may have severe consequences for resource-poor farmers, living in a marginal environment. Biotechnology applications of research on environmental stress tolerance may assure poor farmers of a stable harvest

Biotechnology applications could also have positive impacts on the environment degraded through conventional practices, e.g. restoration of degraded soil using phytoremediation with engineered crops and/or microorganisms. As a source of renewable energy, GM crops can be engineered to produce fuel directly or indirectly through the processing of their biomass. Production of biomass for fuels such as alcohol would not necessarily contribute to additional carbon dioxide to the atmosphere and could be especially beneficial if such fuels, instead of petroleum-based fuels, met the growing needs of the Third World (Guy et al., 2000).

Reducing encroachment to marginal environment

As with any other productivity enhancing technology, the increase in crop productivity due to biotechnology can reduce the pressure to open up new lands, thus reducing the need to encroach on fragile environment in the tropical and subtropical regions. It is often asserted that thanks to increases in yield, land has been saved with diminished pressure on the environment such as less deforestation than otherwise would have taken place (FAO 2000). It has also been reasoned that a crop productivity increase of just 1 percent per annum, equivalent to a cumulative 69 percent increase from 1997 to 2050, would reduce the amount of new cropland needed to meet future demand to 325 million ha. compared to a habitat loss of another 1600 million hectares if crop productivity is kept at the level of 1997 (Goklany 2000). This is not an unattainable goal as biotechnology research is moving very fast.

Input replacement

One of the main criticisms of the Green Revolution has been that it bypassed poor farmers living in marginal environments and those who cannot afford the cost of inputs such as pesticides, fertilisers and infrastructure cost for irrigation. The `gene' revolution is, to some extent, redressing this imbalance by developing crops that produce their own pesticides. Research in crop nutrient uptakes and biological nitrogen fixation also holds great promise for resource-poor farmers who cannot afford the cost of fertiliser inputs. This would also help to protect the environment through the saving of fossil fuel needed to produce nitrogen fertilizer. Non-leguminous crops, such as maize or rice, may be able to be modified to fix their own nitrogen. Alternatively an expansion of the host range of nitrogen fixing bacteria would permit more crops to maintain symbiotic relationships with these bacteria. For phosphorus, Mexican researchers have demonstrated that genetically engineered tobacco plants showed significant increase in their ability to take up phosphorus as compared with the control plant (Shmaefsky 2000). In addition, a research group at Purdue University has cloned a phosphate transporter gene from Arabidopsis. These genes were also found in other crops such as tomato, potato and alfalfa. GM plants are also being developed for more efficient uptake of phosphate (Prakash 2000).

Animal husbandry

In parallel with crop biotechnology, poor farmers can also benefit from advances in animal biotechnology due to a livestock revolution occurring now in most developing countries. Research has shown that the rural poor and landless get a higher share of their income from livestock than better-off rural people. Hence, an increase in consumption of animal products can actually help increase the food purchasing power of the poor and this livestock revolution could become a key means of alleviating poverty in the next 20 years if proper policies and investments are in place (IFPRI 1999). Animal biotechnology can supply abundant and healthier animal protein at lower cost. In the pipeline, there are pigs engineered to have less fat, chicken designed to resist illness-causing bacteria and beef that can grow twice as fast on less feed. Poor farmers who own few livestock will have their investments more protected due to an improvement in animal health through better and cheaper vaccines produced by recombinant DNA. The detection of such diseases by using molecular-based diagnostics will also benefit village herds in controlling the spread of such diseases and upgrade livestock health, thereby helping the rural communities at large and providing household food security at the individual family level.

Immediate impacts of tissue culture and micropropagation

Although DNA technologies are beginning to benefit small farmers of developing countries, the immediate impact for many countries, particularly those with low technical capital, will be in the production and distribution of disease-free and high quality planting material of the native clones of vegetatively propagated plants. These include banana, plantain, cassava, yams, potato, sweet potato, pineapple, sugarcane, many fruit trees such as apple, pear, plum, date palm, mango, and litchi, and many ornamental shrubs and flowers. The benefits of micropropagation are immediate, and the availability of cheap labour in the developing countries provides a competitive edge in the use of this technology. Micropropagation of banana and sugarcane has created rural jobs in Cuba and promoted exports of propagules of ornamental plants from India to Europe. Within the last five years, nearly one hundred plant- micropropagation companies have been established in India by the private sector. In Cuba, if micropropagation capacity can be scaled up to satisfy domestic demand, the country can save US$15 million annually for expenditure on import potato seed stock. The Cuban cottage industry, based on tissue culture, is providing part-time employment opportunities for rural housewives. In China, micropropagation of virus-free sweet potato seed in Shandong which resulted in an average yield increase of at least 30 percent, gave an internal rate of return at 202 percent and a net value of US$550 million (Fuglie et al., 2001). In Kenya, disease-free banana plantlets have greatly increased yields from 8-10 to 30-40 t/h (Anonymous 2000).


There is always inherent risk in any technology, old and new. There are concerns of increased pest resistance as a result of Bt crops that may result in the loss of Bt as an important pesticide. Such risks can be addressed through scientific-based risk analysis and risk management, including post-commercial monitoring, coupled with proper management of cropping systems. Recent experience with large scale Bt GM crops in US support this view. Tabashnik et al., (2000) reported that, contrary to expectations, increased insect resistance has not been observed in the Bt cotton growing region in Arizona. In addition, it is also expected that further genes for insect resistance in addition to Bt will be identified, reducing potential risks in the future.

Increasing the focus on crops and livestock constraints of importance to small producers

In order to fully realize the benefits of agricultural biotechnology to contribute to poverty alleviation and equitable growth in developing nations, it is crucial that a concerted effort be made to ensure that the benefits of biotechnology are available to a broad spectrum of small farmers in a range of developing countries. Not withstanding the achievements and progress described above, the vast majority of funded research has been dedicated to crops such as soya, maize and canola, which are primarily of interest to large commercial producers in the industrialized world. This will only change if considerably expanded international and national public resources are devoted to `poverty' crops such as sorghum, banana, beans and lentils, as well as to increased disease resistance in small ruminants and poultry, and to traits of particular importance to low-income producers (e.g. biological nitrogen fixation). The sharing of the rice genome sequence data with researchers in developing countries by private sector (IFPRI 2000) and the recent announcement of the sequencing of the banana genome ( are steps in the right direction toward using biotechnology for the benefit of developing nations.

Without doubt, the concerns of the public in industrialised countries as to safety and environmental impact of biotechnologies comprises a key constraint to increased international funding. These concerns must be dealt with, both through improved communication of the real evidence concerning risks and benefits of biotechnology use and adoption, and through improved safety measures. As the introduction of GMs may pose differerent risks in different countries, it is important that international organizations, private and public sectors, donors and other stakeholders in developing countries adequately address public concerns and ensure effective monitoring and regulatory environments.

Developing regulatory frameworks

Assisting in economic planning

Promoting technical assistance among national, regional and international institutions


The impact of biotechnology in the next 30 years will depend largely on the strategies that countries adopt to improve their technical capacity and thus capture the benefits of biotechnology. Although biotechnology cannot by itself stimulate economic growth and alleviate poverty, this new innovation certainly provides an additional tool in the fight against hunger. Theodore Schultz showed more than 30 years ago, that poor farmers are effective business people who use resources and technology at their disposal to obtain maximum return to their investments. The problem is that they reach an equilibrium at a very low level. To bring this equilibrium to higher levels, new innovations are needed.

In the green revolution, many small producers were left behind due to lack of access to the inputs required, as well as inappropriate policies. The `gene revolution' may finally provide the opportunity for them to share in the benefits of technology, provided appropriate enabling policies and investments are in place. It is important to review proposed strategies (Byerlee and Fisher 2000; Spillane 1999), learn from case studies (Paarlberg 2000) and to devise a global strategy to harness the benefits of agricultural biotechnology for the poor. In this strategy, international institutions (donors, the public and private sectors, advanced research institutions and NGOs), within their varying mandates, pool their scarce resources and co-ordinate their activities through regional technical networks, such as the REDBIO, for technology development and transfers. These activities, aiming to maximize the benefits while minimizing the risks through cost effective biosafety measures, rest on several pillars, i.e. ethics, public dialogue and biosecurities. In this respect, FAO would be in position to collaborate with partners to assist Member Countries. Since 1999, the FAO Conference and FAO's Committee on Agriculture (COAG) has recommended that the agency strengthen its capacity to assist Member Countries in harnessing the power of biotechnology (FAO 1999; and /index .asp?lang=en). However, the success of the strategy will depend largely on national governments who, at the end, are responsible for the development of appropriate policies and allocate sufficient resources in agricultural research in their respective countries.




The Guinea savanna zone in West Africa, which lies within the Cereal-Root Crop Mixed Farming System, was selected as one of the case studies because of the potential for development of this major under-utilised natural resource, within which current productivity and household incomes are low. Because of the complexity of the land use systems and the range of interventions required, this case is presented in a general format.

Climate, water and land resources

The Northern and Southern Guinea savannas extend in a broad band through most West African countries. They have a warm tropical temperature regime. The length of the growing period (May/June to September/October) ranges from about 150 days - often with short drought periods in the early part of the growing season, near the border with the Sahel - to about 210 days in the southern part. During most of the growing season, probabilities of drought are small, although a minor mid-season dry period may occur near the southern border with the derived savanna and forest zones, where mid-season droughts have an impact on agriculture. The length of the growing period and the total rainfall vary not only from year to year, but also on a scale of several decades (relatively moist in the thirties to sixties, dry in the seventies and early eighties, relatively moist since then). Average annual rainfall in the area varies from 800 mm in the north to 1200 mm in the south. Potential evapotranspiration is between 1500 and 2000 mm, exceeding annual rainfall. The rainfall surplus during the wet season is generally below 300 mm. Groundwater occurs at varying depth in the valley bottoms and the river plains.

Several rivers pass through the Guinea savanna zone. The Gambia River is strongly seasonal; during the dry season salt water intrudes as far as 250 km upstream. Without major dam construction, it is estimated that no more than 2400 ha could be irrigated in the dry season. The sources of the Senegal River are in the Guinea savanna zone. The annual discharge leaving the zone is estimated at about eight km3 but during the dry season the river often runs dry. The irrigation potential in the area within the Guinea savanna zone is estimated at about 15000 ha. The annual inflow of the Niger entering the Guinea savanna zone from the south is estimated at 40 km3. In Mali, several tributaries with a joint discharge of about 16 km3 join the Niger before it leaves the moist savanna zone northward towards the inner delta, where much water is lost due to evaporation. The discharge is about 30 km3 where the Niger re-enters the zone of the moist savanna and about 40 km3 where it leaves the Guinea savanna to the south. It is difficult to estimate the irrigation potential for the moist savanna zone, but the total irrigation potential for the Niger is about 2.8 million ha.

The sources of the Volta River are located in the zone of the moist savanna. Where the Volta leaves the zone southward, the discharge is estimated at about eight cubic km. The irrigation potential is roughly estimated at 0.7 million ha. The extent currently equipped for irrigation in this zone is in the order of 84000 ha, excluding the areas of cultivated wetland and inland valley bottoms.

The major soils in this zone (Lixisols) contain low activity clays but have a relatively high base saturation. Associated with these are similar but more acid soils with low base saturation (Acrisols) in the more humid parts and acid sandy soils (Arenosols) in the drier, northern part of the zone. Very poor, largely unproductive soils with a lateritic pan at shallow depth occur scattered over the area and are locally dominant, as in parts of Northern Ghana. The sandy soils are relatively easy to work but nutrient deficient and retain little soil moisture. The acid soils in the more humid part generally have a low natural fertility, and many need lime and phosphate applications to increase yields. In the valleys and major river plains, soils are generally fertile but in many parts they are also intensively used. They may become depleted quite quickly therefore, if not regularly replenished with nutrients.

The Guinea savannas form part of an extensive agro-ecological zone also occurring in Eastern and Southern Africa and in the Cerrados region in Northeast Brazil, as well as in parts of several Southeast and South Asian countries. Agricultural technology from other parts of this zone could be usefully explored for possible adaptation and use in the Guinea savannas, given similar general land conditions8.

General land use and farming systems

The Guinea savannas, less densely populated than areas nearer to the coast, still have land that is very lightly used, particularly at a distance from the roads. The more easily accessible land is largely used for annual crops, generally with low external inputs, and producing low yields. Crops include maize and sorghum; millets in the northern part; cotton, cassava, soybean and cowpea; yam near the southern border, and wetland rice in parts of the river plains and valley areas. Homesteads often have some vegetables and fruits. Cattle, mainly N'Dama, are held on many farms for draught or for milk but are less common near the southern border because of the tsetse threat. Small livestock is often held in the homesteads. Manure is used to maintain the productivity of the homestead garden, and some manure may be applied on the nearest fields as well. Part of the farms have draught power, mainly oxen, some of the largest farms have a tractor, but many farmers cultivate their land with the hoe, and therefore the extent of their cultivated land is severely limited. Generally the household consumes a large part of the farm produce; some is sold at harvest time.

Five main farming systems or variants (FS) have been identified9 in the Guinea savanna zone, with different population densities, proportions of land under cultivation and cattle densities (Table 1.). The first system (FS1) covers most of the Nigerian part of the zone, a large area in Burkina Faso, parts of Togo and Benin, and a part of Senegal. It is a mixed farming system where population densities, proportion of cultivated land and cattle density are medium high. There still is room for further expansion as well as intensification. This system tends to occupy a greater proportion of land near the gradual boundary with the two intensive crop-livestock mixed systems practised in densely populated areas and around urban centres (e.g. the Jos Plateau and around Kano in Nigeria, Ouagadougou in Burkina Faso and Dakar in Senegal). In one of these (FS2) the high population density still allows for a high proportion of cultivation and a high cattle density; in the other (FS3) the percentage of land under cultivation and the number of cattle kept show the effect of the urban centres and the very high urban and peri-urban population density. In these two systems, agricultural production increases will have to come from further intensification rather than from expansion.

FS4 covers large areas of Benin, Ghana, Ivory Coast, Mali, Guinea, Guinea-Bissau and Senegal, and the remainder of the countries mentioned above. As may be expected from the lower population density the proportion of cultivated land and the cattle density are also low. The system comprises mixed farming around villages, and agro-pastoralism and pastoralism in surrounding areas. In addition to local, trypanotolerant N'Dama herds (e.g. in Guinea), there are important numbers of Zebu cattle in areas with relatively low tsetse pressure. These migrated south with their pastoralist owners during the droughts of the eighties (e.g. to Ivory Coast), and many have settled in this zone. Settlement of increasing numbers of people in those places in the Sahel zone to the north where water is available has been restricting the possibilities for rainy-season grazing and further induced the pastoralists to settle in this area. Crop agriculture and livestock herding often co-exist with little integration, giving rise to frequent conflicts. Lack of formal land tenure or uncertain tenure arrangements are contributing factors. In the more humid edge of the zone in Guinea, there is an area with a still more extensive, slightly more livestock-oriented mixed system (FS5).

The areas of these two extensive farming systems have an important potential, particularly for expansion of integrated crop-livestock systems, in so far as factors such as the presence of locally widespread poor soil conditions or the prevalence of onchocercosis or animal trypanosomosis do not inhibit the further development of these mixed agriculture systems.

Table 1. Farming systems in the Guinea savanna zone, West Africa


Land under

Cattle density

Ha cultivated
/100 people

Cattle number
/100 people


FS 1







FS 2







FS 3







FS 4







FS 5







* Values not adjusted for the presence of urban land and population.

Infrastructure, services and institutions

In Guinea savanna zone countries, public investments in rural transport have generally concentrated on improving main highways. An estimated 50 to 70 percent of the existing feeder road network in the Guinea savanna zone are in poor or very poor state. Also, the connecting access tracks are generally ill maintained and often lack minor but basic structures such as culverts, becoming inaccessible to vehicles especially during the rainy season. These conditions make transport of produce to local or more distant markets expensive, slow and sometimes impossible.

A number of deficiencies affect the agricultural marketing system, particularly the smaller rural markets and the secondary markets that form the link between these and the main consumption areas. Besides the transport problems, these include a lack of common facilities resulting in wastage and spoilage, distortions in produce flow, and inadequate price formation and market transparency. As a result, the marketing system is slow to respond to producer and consumer demands, has high transaction costs and entails a high degree of losses, especially of perishable commodities, including cassava.

Only small percentage of the rural population in the Guinea savanna has safe water supply in easy reach. Most of the traditional supplies, usually from streams and ponds, require considerable time and effort to utilise, may dry up during drought periods, and have very poor water quality in the dry season. Water-borne diseases are prevalent. In most of the countries of the Guinea savanna zone, less than 15 percent of the villages are connected to the national electricity grids. The lack of electricity limits the opportunities for education, rural water supply, irrigation and local processing.

In most countries of the zone the T&V system of extension was introduced, including the unification of sub-services such as livestock, forestry and fisheries. The system has brought some advantages but has generally proven to be financially unsustainable and often fails to function properly within a decentralised government structure. In the recently resettled oncho-freed areas the flow of extension information is often poor or even non-existent.

Readily accessible savings and loan services are generally lacking in the zone. There are occasional credit facilities linked with cotton cultivation. Informal credit is generally at high rates of interest.

In the Guinea savanna zone, farmer-based organisations are generally weak or are lacking, because several areas have been opened for settlement only recently, after eradication of onchocercosis, and villages are further apart and communications are poorer than near the coast. However, most villages have some social cohesion based on traditional structures and relationships. On this basis, rural farmers' organizations could be promoted.


Historically, development of this area has suffered from two major constraints: onchocercosis and trypanosomosis. The Onchocercosis Control Programme (OCP), a major ongoing effort since 1974 to eradicate river blindness in West Africa, has prevented several hundred thousand cases of blindness and freed up to 25 million hectares of previously infested cultivated land. Much of the Guinea savanna is now onchocercosis-free, and the programme is working on eradication from the remaining endemic areas, for example in parts of Nigeria. The programme thus has been removing a major constraint to socio-economic development and opened important new economic opportunities for the population in the region. Important parts of the additional agricultural land becoming available as a result of Onchocercosis control (so-called `new land') is likely to be relatively fertile and have access to surface water or groundwater, because Onchocercosis was most prevalent near rivers where the vector (blackfly) laid its eggs.

While onchocercosis has been effectively controlled in more than 60 percent of the area, tsetse-transmitted African animal trypanosomosis (AAT) still is a major constraint to agricultural development. Distribution of tsetse, and therefore of trypanosomosis risk, is not uniform over the entire area. Besides the presence of different species adapted to either savanna or riverine vegetation), tsetse distribution is affected by climate. Therefore in the less humid northern band of the Guinea savanna zone, tsetse will generally be present only along main drainage systems. This enables livestock keepers to minimise exposure of cattle to tsetse by using adapted grazing patterns. However, in the southern, more humid band tsetse colonises not only the drainage systems but also expands seasonally into adjacent savanna and trypanosomosis risk increases towards the south.

The historical response from farmers to this threat has been trypanotolerant livestock, i.e. livestock that tolerates trypanosomes, the parasite causing AAT. While it has been shown that trypanotolerant livestock do make a contribution towards reducing the trypanosomosis problem, trypanotolerance alone does not enable farmers to keep animals as and where they wish. The presence of the tsetse fly remains a major obstacle to cattle raising and inhibits particularly mixed farming development and intensification. AAT often prevents farmers from making the step from subsistence agriculture to more advanced livestock raising systems. A recent FAO study10 highlighted the areas where tsetse removal would be most beneficial to livestock. Major areas in the Guinea savanna zone include the West African cotton triangle (Burkina Faso - Mali - Ivory Coast) and the Nigerian middle belt. An ongoing study is assessing the economic benefits of large-scale control operations in more detail.

Tsetse also transmits human sleeping sickness. This is not a major concern in this area, however, since sleeping sickness is confined to a series of foci in the more humid parts south of the Guinea Savanna zone.


The opportunities and development interventions with the greatest effect on poverty alleviation and on productivity and sustainability of farming and household systems in the Guinea Savanna zone, are focussed on the intensification of the mixed crop-livestock farming systems, including: localised irrigation; improvement of services to the pastoral/agro-pastoral system; improving crop-livestock integration (draft animals, feed from cover crops and crop by-products, manure from intensive livestock production); diversification through crop introduction and rotations; promoting local post-harvest value-adding enterprises; and improving infrastructure, services and institutions designed to serve the people and the enterprises in the zone.

The specific opportunities and promising interventions discussed individually below should not be seen in isolation; they are designed to assist households and communities to improve their livelihoods, building on the resources and farming systems of the people in the area, and there are important synergies among them. For example, conservation farming reduces the amount of draft power needed for land preparation, so draft animals could handle more cropland. Conservation farming would reduce free dry-season grazing on cropland but will produce cover crops and increased amounts of crop by-products that can be used to improve livestock nutrition in the dry season. While some of the interventions will already have considerable benefits in the current situation, most will have much greater effect if they are accompanied by improvements in infrastructure, services and institutions such as those mentioned in the last section.


Farm households can achieve significant improvements in economic and nutritional status through modifications in their approach to soil, crop and pest management. These changes in the farming system lower input costs, create a more even distribution of the workload throughout the year, improve the effect of applied fertilisers and organic materials, reduce weed pressure and the need for insect and disease control activities, and lead to higher and more stable yields. There are three main elements or aspects of these changes.

Conservation agriculture is based on minimum- or no-till methods, direct seeding and continuous soil cover11, improving resilience against drought and gradually reducing weed pressure. Integrated plant nutrition management, based on crop rotation including legumes, continuous soil cover and maximum cycling of nutrients - including from manure and post-harvest residues or waste - improves plant nutrient balance and health and raises the efficiency of applied mineral fertilisers. Integrated pest and plant management is based on crop management, to increase its resilience to weed and pest pressures, and systematic periodic observation by the farmers of the balance between pest populations and those of their natural predators. IPPM thus reduces and may eliminate the need for pesticide application, lowering production costs, and improves plant health and yields. An example for all three elements is the cultivation of a mucuna cover crop for weed control, enhancement of soil fertility and livestock feed. This practice is already widespread in some parts of the Southern Guinea savanna zone, but still unknown in large areas.

Such improvements in the existing farming system, as well as other changes discussed below, cannot be simply packaged and delivered. They need to be explored, tested and adapted through practical experience, participative learning and discussion, to be thoroughly understood and successfully applied. Methods along the lines pursued by Farmers Field Schools, Farmer Research Committees, Participatory Technology Development, Promoting Farmer Innovators, and Campesino a Campesino have proved effective in raising and maintaining farmers' active interest, co-operation, and initiative. Farmers' organisations, NGOs, and small-scale private extension service providers can all be held accountable by local communities for enabling those communities to make better decisions, based on sound scientific understanding rather than on standard recipes.

A suitable entry point for improvements may be the introduction through participative learning and discovery of no-till and continuous soil cover methods with suitable crop rotations, since these can drastically increase yields and drought tolerance within 2 to 3 years, even with low external inputs. Further improvements can then be explored after this initial success.

Investment for Conservation Farming Equipment

Availability of farm power, particularly to carry out tillage and planting operations, is becoming a bottleneck to increasing or even maintaining agricultural production. Lack of human labour due to migration and diseases, lack of animal draft power due to the occurrence of trypanosomosis or lack of fodder and lack of motorised power due to capital and service shortcomings, are all causing delays in planting and hence low yields, or are restricting the area that can be planted. Conservation agriculture provides a chance to carry out timely planting operations with much less farm power, achieving higher yields and releasing human labour for other, more rewarding activities.

Investment assistance may be necessary for the introduction of this technology, although in the form of credit, rather than as free handouts. The main investment on equipment will be for the planter. In addition, there might be a need for a one-time treatment by ripper or subsoiler on highly degraded soils, and for a sprayer mainly for weed control, particularly in the first 2 to 3 years. Cover crops and crop residues after harvest can be managed with equipment such as the knife roller. This can be easily manufactured locally at low cost12.

Investment in livestock development

As discussed above, removal of the tsetse constraint is expected to have a major impact on general agricultural development in the area. Therefore the results of the ongoing joint cost-benefit studies on tsetse control and eradication by FAO and DFID may pave the way for investment in the area. Privatisation of veterinary services and extension is the way forward, but should be based on a review of this sector throughout the region, highlighting the failures and successes, to identify present needs and investment opportunities.

The pastoralists in this zone are now increasingly settled, occupying areas distant from villages that are not yet used for arable cultivation. Some countries still need to put in place policies and legislation to establish legal title or tenure for well-defined holdings, occupied both by pastoralists and farmers. Participatory approaches will be essential for resolving land tenure problems. A good starting point will be to provide legal support for actual, existing settlement and occupation patterns rather than to direct pastoralists to newly created pastoral zones.

The pastoral and increasingly agro-pastoral systems will become more profitable and sustainable through strategic investments combined with the promotion of pastoralists' organizations and training. Specifically, the establishment of well distributed, readily accessible cattle markets and strategically located slaughterhouses is essential to ensure that livestock owners themselves profit from the added value of their cattle. Accessible veterinary care, vaccination and quarantine facilities will improve animal health and productivity, but also enable certification of animals or meat for export, opening new markets and increasing widespread acceptance of the products. These services, as well as markets and slaughterhouses, could be under private enterprise if government controls on health and sanitary quality are put in place.

The promotion of draft animal power is a key factor towards improved integrated livestock and crop agriculture. This issue should not be addressed in isolation but as part of a whole package enabling sustainable draft animal power use within a region. Animals should be drawn from local herds rather than from government ranches. There may be scope for the introduction of improved breeds in those herds, but the emphasis should be on proper selection.

Improved livestock husbandry techniques should be introduced in the mixed farming systems, including night stabling for manure collection, or permanent stabling, with proper feeding and maximum use of crop residues and by-products as well as fodder crops. Introduction of better nutrition through the use of fodder crops and crop by-products would be most profitable in the framework of milk production or stall feeding during the dry season. Short-cycle livestock production at the village level, such as poultry and egg production for local markets and small ruminant fattening for annual occasions can be enhanced using minimal inputs. More complex is the introduction of improved husbandry practices in semi-intensive poultry, small ruminant and pig production. These need functioning extension and veterinary services, access to animal feed and feed additives, as well as reproduction services. All these services can be provided in response to well-organised demand from farmers' associations and with complementary investment in improved access to local markets.

Milk co-operatives in peri-urban areas have not always obtained positive results. Two reasons are often cited for lack of success: competition with cheap, easily stored, imported milk powder, and the fact that stakeholders often are not involved in the processing, retail and delivery systems. While the price of milk powder will remain a major constraint, fully involving cattle owners' associations from the start of such projects might be part of the answer. If successful, a next step can be the genetic improvement of local breeds. Intensive peri-urban poultry, pork and milk production systems are already the focus of private sector investment in several places. Nearby farming areas can profit from the organisation of commercial links with these systems, through sale of feed crops and crop by-products, as well as through the use of manure from these intensive systems.

Affordable, small-scale irrigation facilities

In recent years, simple, affordable equipment for precision drip irrigation has been developed and widely applied in several countries. Systems have proved successful at scales ranging from about 20 square metres, used in homesteads or backyards, to a few hectares, serving several horticulturists.Such systems, already tested in some African countries, will allow dry-season cultivation of vegetables and fruits - for local consumption, as well as for the market - near surface water or where groundwater is available at shallow depth. A few buckets of water per day are sufficient for 20 square metres of vegetables; a hand pump or the Swiss model treadle pump can supply the requirements for several hundred square metres. Motor pump sets will be needed in larger systems. The treadle pump, already used in Senegal and Tanzania, for example, can be manufactured locally using low-cost standard materials.

Introduction of oil palm cultivation and processing

The recent development of high-yielding, precocious oil palm clones adapted to certain environments outside their traditional range provides an opportunity for their introduction in parts of the Guinea savanna zone, specifically in valleys and river plains with moderately accessible groundwater.

The economic importance of the oil palm derives from its extremely high productivity as oil producers. In the plantation sector, large-scale mills produce crude palm oil for edible oil refining industries, which further process it into cooking oils, margarine and baking fats. Crude oil for household use and local sale is generally produced by traditional artisanal processes. These are relatively simple, but tedious and inefficient with very poor recovery rates. Manual and fully automated processes that are much less tedious and have higher extraction efficiency are widely used for palm oil extraction at the small-scale level. Small-scale extraction equipment is locally produced in a number of countries in the region. There is however a need for improving the efficiency of this equipment.

Local manufacture and marketing of cassava and soybean products

Cassava is generally processed in the household or the village prior to consumption or marketing, not only for the purpose of detoxification, but also to increase its shelf life and facilitate its transport. Primary food processing options for cassava include the production of chips by drying. There is currently a large trade in cassava chips for the feed industry, but it is extremely competitive. Since distribution (transport) costs are an important factor in European cassava chip imports, ports in West Africa may have an advantage with respect to traditional suppliers from Asia. Cassava chips could also become important among the local feed sources for farmers' and pastoralists' livestock.

Secondary processing options include the production of flour (from chips), and a variety of products such as fufu, a moistened, highly perishable product prepared from cassava flour, gari, a fermented coarse meal, attieke, a fermented pre-gelatinised meal similar to gari, lafun, a fermented cassava flour, and chickwangue, a pre-gelatinised cassava paste usually in the form of balls. All of these products are generally produced at the household to village level, with the use of relatively simple equipment for peeling, grating, de-watering, sieving and heating (frying or steaming). With an investment of a few hundred dollars, a village-level cassava processing industry can be established. A few larger-scale automated lines for the processing of gari, attieke and cassava flour also exist.

Although village and household-level cassava processing is labour-intensive, it offers considerable potential for income generation in rural areas. It would however greatly benefit from improved technologies and training inputs.

Soybeans offer exceptional nutritive value for humans as well as for domestic animals. However, cooking, roasting or extrusion is required in order to render the protein content of the bean digestible for humans. Soybeans can be easily processed at village level. Soymilk and its derivatives such as tofu, yoghurt and ice cream can be relatively easily processed. The production of soymilk is simple, does not require specialised equipment, and can be undertaken in the household. With an investment of a few hundred dollars, a soy milk cottage industry can be established. Fully automated lines for soymilk production are also available from many manufacturers around the world. Small additional investments are required for the production of tofu or yoghurt from soymilk. The solid fraction remaining after filtration of the milk forms the basis for several popular food dishes in Nigeria.

The extruder-expeller process can produce pressed soy oil at small to medium scales. Using this process, about half of the oil contained in the bean is recovered. The resulting high-protein, high quality meal can be milled to flour and used to fortify several other food products for human consumption, increasing their protein content. The meal can also be used as a high-protein component in animal feeds, particularly for intensive peri-urban milk or livestock production systems. Full fat soy meal can be more cheaply produced by direct extrusion, yielding a high-protein high-calorie feed ingredient.

Local maintenance, repair and manufacturing of equipment

From the very beginning of the introduction of conservation agriculture and zero tillage/direct planting, the machinery input supply and service sector must be involved. In different parts of the world equipment for zero tillage and direct planting has been developed and is readily available for a wide range of crops, farming systems and farm sizes. As a result, basic technology is exists for the introduction of zero tillage in West Africa. However, the equipment must be adapted to the local conditions, available materials and handling habits. This requires a close collaboration from the start between rural mechanics or blacksmiths, workshops and small manufacturers as well as with farmers using the new technologies.

Experience has shown that such a technology development process for equipment is best done by the commercial private sector rather than by research stations. From an early point in time the mechanical workshops/manufacturers have to assume ownership for the equipment development. The potential future manufacturers or repair-service providers will need to collaborate closely with the farmers using the technologies, to make sure that the local technologies developed serve the farmers in the practice of conservation agriculture and provide a good long-term business for the commercial sector.

Support for the introduction of conservation agriculture should therefore include mechanisms to facilitate the establishment of a small-scale industry that produces, distributes, services and further develops tools and implements for conservation agriculture. An example of the effective creation of such a small-scale industry can be found in the southern states of Brazil.

Investment in supporting infrastructure, services and institutions

The effective spread of people-driven exploration, innovation and improvements in farming systems and related economic activities will require an effective support structure. Issues to be addressed include markets to absorb the increased supply of produce and derived products, market information, timely availability of the right inputs, seasonal and longer-term credit, technical and scientific support on demand to farmers and their associations, a well maintained transport infrastructure. External assistance will be needed in many cases to finance the needed strengthening of institutions and services and the improvement or rehabilitation of infrastructure, but their operation and maintenance will need to be paid for through service fees or other types of payments by beneficiaries. Management could be by local communities and institutions, private enterprise or provincial or national self-financing entities, accountable to the contributing beneficiaries. Investment in developing financial services should be focussing on training, promoting savings and loan groups in rural areas, facilitating the organisation of local banking institutions and training their staff, rather than on providing external funding for credit.


Water Harvesting and Soil Rehabilitation in India and Africa:
Potential and Practice13


Most rural people in developing countries are dependent upon a resource-based subsistence economy using products obtained from plants and animals (Agarwal & Narain 1989). However, a large portion of the world's rural poor today live in highly degraded lands and environments. `Ecological poverty' can be described as the lack of an ecologically healthy natural resource base that is needed for a human society's survival and development. Healthy lands and ecosystems can provide the wealth that is needed for economically viable, healthy and dignified lives.

The challenge today lies in empowering and mobilising people to enable them to escape from their `ecological poverty', in order to create natural wealth, and to develop a robust local economy. Experience in India and Africa has repeatedly shown that major economic change in rural communities has occurred wherever they have organised themselves to regenerate and manage their natural resource base (Tiffen, et al., 1994). However, externally imposed technocratic resource management systems have invariably failed or been cost-ineffective, making them irrelevant in the financially constrained world of the poor.


Over the last one hundred years, the world has seen major shifts in water management. During much of the last century individuals and communities steadily relinquished their role in water management to the state. The simple technology of storing and using rainwater declined in importance and the exploitation of rivers and particularly underground aquifers became pre-eminent. However, as water from these sources comprises only a small portion of total precipitation, this focus inevitably led to a growing pressure on these sources. In many cases, governments of water-scarce countries have encouraged massive interventions into the hydrological cycle but have done little to sustain the integrity of the hydrological system itself.

In reality, rainwater harvesting and the utilisation of groundwater sources are complements, not substitutes. In drought years, when rain is scarce and rivers dry up, groundwater becomes an important source of water both for drinking and irrigation (Agarwal 2000). However, were effective rainwater harvesting sytems in operation, the need for groundwater extraction in normal years would be reduced to minimal levels in many areas, while even in semiarid areas groundwater sources would be no more than a supplementary role. Water harvesting and groundwater together can drought-proof the country and create local food security.

In the case of India, for example, average annual rainfall is 1170 mm. But more than 50 per cent of this rain falls in about 15 rainy days and in most cases in less than 100 of a total of 8760 hours in a year. It is therefore very important to capture this extremely transitory resource before it is lost for human use (Agarwal & Narain, 1997). Effective rainwater harvesting offers the opportunity to achieve this. Adequate and safe potable water for human consumption could be guaranteed in all areas of the country, and sufficient water would generally be available to grow at least the less water-intensive crops every year.

Rainwater harvesting can also serve another purpose, by reducing the dependence of farmers on weak state institutions. In the African context, attempts to create large-scale capital-intensive agricultural systems have failed in most cases as a result of poor management, policies encouraging inappropriate and inefficient use of water resources, and financial problems that have constrained maintenance and staffing.

This paper presents two case studies. The first is from India and describes the transformation from a state of ecological poverty to a state of sustainable economic wealth. The second case, from Niger in West Africa, describes the use of indigenous technologies to improve water capture and utilization. These case studies are important because they show not only how these simple technologies can radically improve water availability for poor farmers, but also because they illustrate the associated ecological, social and economic impacts. The experiences of Sukhomajri village in India now span over 20 years. The experience of the case in Niger is more recent, but it is still as dramatic and significant.

Over the 1970s and 1980s, India has seen a number of micro-experiences of successful community resource management. What is remarkable is the short time it takes to transform a poverty-stricken, destitute and ecologically devastated village to a rich and green village. This is true for both examples presented in this paper.



The village of Sukhomajri near the city of Chandigarh has been widely hailed in India for its pioneering efforts in micro-watershed development. In 1976, Sukhomajri - a small hamlet with a population of 455 situated in the sub-Himalayan Sivalik foothills - had a sparsely vegetated environment, with poor agriculture, and high levels of soil erosion and runoff. Though the annual average rainfall was about 1137 mm, groundwater was not available at a reasonable depth. Soil erosion and gully formation was steadily leading to a decrease in farm area. As agriculture was riddled with uncertainty, villagers traditionally kept herds of livestock to minimise risk. They cultivated about 50 ha of unirrigated land and kept about 411 heads of animals. Open grazing by livestock suppressed regeneration and kept the surrounding hills bare.

In 1979, facing a severe drought, the villagers built a small tank to capture the monsoon runoff and also agreed to protect their watershed to ensure that their tank did not get silted up. Since then the villagers have built a few more tanks and have protected the degraded forest that lies in their catchment area (Mishra et al, 1980). The tanks have helped to nearly triple crop production and the protection of the forest area has greatly increased grass and tree fodder availability. This, in turn, has increased milk production. With growing prosperity, Sukhomajri's economy has undergone a change.

Economic Impact

The following economic and ecological changes have taken place in the village over the years:

In Sukhomajri, a major incentive for the villagers to protect their watershed came from changes in forest department policies. For example, the forest department would traditionally auction the grass in the degraded watershed to an outside businessman who would charge the villagers high rates to harvest the grass. The villagers argued that as they were protecting the watershed, they should get the benefits from the increased biomass production and not the contractor. The state forest department agreed to give the grass rights to the village society as long as the villagers paid the forest department a royalty equivalent to the average income earned by the department before the villagers started protecting the watershed. The villagers pay their village society a nominal amount to cut grass in the watershed. A part of this is used to pay the forest department and a part is used to generate community resources for the village.

A village-level institution for natural resource management

A crucial role in this entire exercise was played by a village-level institution specifically created for the purpose of watershed protection. The Hill Resources Management Society, as this institution is called, consists of one member from each household in the village. It provides a forum for all households to discuss their problems, manage the local environment and maintain discipline amongst their members. The society makes sure that no household grazes its animals in the watershed and in return it has created a framework for a fair distribution amongst all the households of the resource so generated, namely, water, wood and grass.

Future operational strategies in the Indian case

Despite the success of Sukhomajri and other communities with similar experiences in Ralegan Siddhi and Tarun Bharat Sangh, little adoption of these practices has occurred in other areas. Critics have often used these facts to condemn these examples as non-replicable creations of remarkable individuals who have persevered to bring change. But that is not the correct picture. These examples have remained scattered because a governance system that would foster local control over natural resources does not exist in the country. The current examples exist despite the system rather than because of them. It therefore takes enormous perseverance for individuals to bring change at the micro-level. However, if the system of governance enables local communities to improve and care for their resource base, it would be easier to bring about change. The Rajiv Gandhi Watershed Development Mission of the Madhya Pradesh government has shown that the state can replicate these community-based efforts if there is adequate political will and pressure on the technical and administrative bureaucracy to deliver.

In order to develop a sustainable village-level natural resource management programme, it is essential to develop a conceptual framework that addresses both the private and common property resources of the village, its diverse biomass needs, and the interests and requirements of different socioeconomic groups within the village community. Such a programme sets into motion a series of ecological successions, beginning with increased quantity and productivity of croplands as a result of increased water conservation. This in turn leads to an increased availability of water for irrigation, expanded grass production, and slowly increased production of fodder and timber resources from the tree and forestlands. Each of these stages of ecological succession generates its own economic impacts on the village society, which slowly unfold over the years.


This case study teaches us that a bundle of policy measures are also needed for good natural resource management. These measures include changes in the current institutional, legal and financial framework in order to engender community-level, participatory democracy. It is only once this policy package is implemented that these `isolated' micro-experiences will bloom into a `million' villages.

Structures with a social process

Ecological change in Sukhomajri started with water harvesting. Building water harvesting structures is a comparatively easy task. But starting off a process of self-management in village communities is much more difficult. This is possible only if each water harvesting structure is the result of a co-operative social process. Strong social processes precede each structure to build `social capital'. This is an area where the track-record of government agencies is literally non-existent and inflexible government rules militate against the very principle of social mobilisation. Social mobilisation means, firstly, creating awareness and confidence in the people that water harvesting works. Once this is achieved, it means creating village institutions that will decide where, when and how the water harvesting structures will be built, who will build them, and how much the villagers will contribute to construction costs. Once the structure is built, a key consideration is how the benefits - that is, the water - will be shared amongst the villagers, especially in the early years when water is scarce, and how will its use be regulated. Every part of the community will have to be involved by making each section - from the landed to the landless and women's groups - appreciate the benefits it will derive from the exercise. And by making efforts to ensure that benefits do indeed flow to each section of the community.

It is for this reason that water harvesting works best when combined with watershed development. It is in the nature of structures to be of value primarily to those who have land leaving the landless without any benefits and therefore alienated from the exercise. The development of watersheds to conserve both water and soil increases both soil and water conservation and leaf and grass production on what are usually common lands, thus greatly benefiting landless households. In addition, the process extends the life and effectiveness of the structures that benefit the landed by reducing siltation.



This second example presents the experience of the IFAD-funded Soils and Water Conservation sub-programme in Niger and its success in promoting the tassa technique, a low cost soil and water conservation and water harvesting technique. Tassas are a farming technique that helps to soften up deeper lying soils and enrich their organic-matter content; the technique consists of digging small holes and then planting seeds on the ridges formed by the dirt removed from the holes. Application of simple, low cost techniques such as the tassa makes it possible to recover degraded and abandoned land and transform it into productive cropland through individual and group action. There are many similar examples from Africa (Reij et al., 1996).

The IFAD-funded Soil and Water Conservation (SWC) Project started in 1988 and targeted the Illéla district (400 mm average rainfall), south of Tahoua. Population pressure and catastrophic droughts (1972-1973 and 1982-1985) had led to degradation of croplands (shortening and, in some cases, disappearance of fallow periods), pasture land and wood resources, as well as progressive fragility of production systems. The large majority of the population in Illela district is sedentary farmers (mainly Hausa). The IFAD project concentrated its activities mainly on 77 villages with about 100000 inhabitants. Rainfed agriculture is the dominant mode of production with millet, sorghum and cowpeas as the major crops. However, most people cannot survive on rainfed crop production alone. Livestock and commercial activities are important sources of income. Seasonal migration of young men is also common.

The programme was designed as a pilot action. But an overall area of 6350 ha was developed, more than twice the initial expectations. Off-farm measures consisting of collective silvo-pastoral initiatives were carried out on 585 ha, while collective cropland actions were carried out on 5800 ha. Although large-scale erosion-control activities (i.e., construction of stone bunds) fell short of expectations, the tassas - which were not among the initially planned SWC techniques - were a resounding success, and their use continued to spread on individual plots even after the project closed.

The techniques focused mainly on bringing land back into production, reducing inter-annual variability of production and enhancing the resilience of agricultural systems to climatic risk. The tassa technique, in particular, is spreading at a surprising rate, encompassing an additional 2 to 3 hectares per annum on some holdings. Tassas are best suited to landholdings where ample family labour is available, or where farm hands can be hired. The technique has spawned a network of young day labourers who have mastered this technique and - rather than migrating - go from village to village to satisfy farmers' growing demands. There have even been cases of land being bought back by farmers who recognised early on the profit that can be earned from this land. The SWC sub-programme in Illéla can be considered one of IFAD's most successful actions to develop rainfed agriculture in semiarid zones and improve food security. On average, food availability in participating households rose between 20 percent and 40 percent, depending on local rainfall conditions.

A new approach: Low cost and replicable techniques

A major objective of the project, in contrast to previous efforts, has been to introduce simple, low-cost soil conservation and water harvesting techniques, which could easily be mastered by farmers. The main targets of the project were to construct contour stone bunds on 2300 ha in four years and to develop 320 ha with half moons (demi-lunes). The project changed course in the second year. Ten farmers were sent to the Yatenga region in Burkina Faso where they observed various types of SWC techniques employed, including improved traditional planting pits (zaî). These reminded them of a traditional technique (called tassa in Hausa) used in their own region, which had been more or less abandoned. The traditional tassa consisted of just tiny pits made with a hoe to break the surface soil crust before the onset of the rains. The improvements in the zai technique consisted in increasing their dimensions (from a diameter of 10 cm to 20-30 cm and from a depth of 5 cm to 10-25 cm) to collect more rainfall and putting organic matter in the pits to improve soil fertility. The organic matter attracts termites, which digest it and make the nutrients more easily available to the plant. Termites also dig channels thus increasing the infiltration of water into the soil.

On return, the improved tassas were tried out on four ha of land in the village of Nadara. The impact was spectacular. Some 70 ha of degraded land were rehabilitated in 1990. Only in the tassa pits was a reasonably good yield obtained during the drought of 1990. This convinced farmers of the great advantage of this technique so that in 1991 they applied the technique to 450-500 ha and in 1992 a further 1000 ha. By the end of 1995, about 3800 ha had been treated in Illéla district alone. These figures underestimate the achievements of the farmers because they are based exclusively on what the extension agents have been able to measure. No figures are available for other districts, however, field observations indicate that they are increasingly being adopted elsewhere.

The Extension System

The project organised exchange visits between villages, to share experiences and train villagers in the various aspects of SWC works. These exchange visits became one of the key activities, very appreciated by farmers, and which had an important impact in accelerating the diffusion of the SWC technologies.


The IFAD-funded project has measured the impact of the tassa, half-moons (demi-lunes) and contour stone bunds on a large number of farmers-managed demonstration plots. A comparison of the different SWC/WH techniques on millet yields show that in years of drought, demi-lunes perform on average slightly better than tassa because demi-lunes have a larger catchment area, so more runoff is available to plants. In contrast, in a year of good rainfall, tassas do slightly better than demi-lunes when only manure is used. These results show that water harvesting on its own has an important impact on yields and adding manure increases yields further. The rehabilitation of degraded barren land using tassa is clearly an economic proposition as farmers and traders are now increasingly getting interested in buying degraded land.

Impact on food security: While in a year of good rainfall, most families in the Illéla area produce more or less sufficient food crops, in all other years, and in particular in low rainfall years, these families get into serious trouble. To procure their cereals requirements, they are obliged to sell livestock or earn cash through migration. Families who have invested in SWC still have a cereal shortage in a drought year, but much lower than before. If they store some of their surplus gained during a good rainfall year, they will be able to meet their cereal needs even in years of low rainfall.

Management of rehabilitated soils: Maintenance of conservation works and soil fertility is necessary to ensure the sustainability of yield levels. A survey conducted in December 1998 showed that farmers rarely re-dig tassas every year. They usually continue to use the same pits. The demi-lunes need to be cleaned and the ridge repaired every two to three years. In general, maintenance is irregular, but more so for the demi-lunes since their total maintenance requirements are more demanding.

Because livestock raising is an important component of livelihood systems in the semiarid region, most farmers have access to adequate quantities of manure. The major bottleneck is the transport of manure to the fields. Only a minority of farmers applies manure to their tassa every year. The majority does so every second year. In current economic circumstances, mineral fertilisers are neither available to the majority of farmers in Niger nor would their use on millet be economic.

Role of paid labour: To rehabilitate degraded land with tassa requires a considerable investment in labour. Family labour is usually not sufficient and farmers have to either hire labour or organise traditional work parties. In several cases, young men have organised themselves in small groups of 5 to 10 persons, which can be contracted out by individual farmers to undertake specific SWC activities. Many families increasingly rely on hired labour. Many analysts assume that resource poor farmers in particular benefit from this emerging labour market. This new source of cash income means that they don't have to sell all or most of their small livestock in case of a bad harvest and they can avoid migration.

An emerging land market: In the Illela project, the use of simple and effective SWC techniques that could be applied by individual farmers meant that they could treat lands over which they had land use rights. This situation is very different from other SWC operations in which large blocks of land were treated, which almost always created land tenure conflicts.

Badly degraded land on the plateau in Illela has become productive again and a land market has now emerged. Farmers buy and sell degraded land for prices that increased considerably between 1992 and 1994. The emergence of a market for degraded land shows that farmers believe that tassa are an efficient and cost-effective means for bringing degraded land back into production.

Recommendations for Future Operational Strategies and Actions

The identification and adaptation of local technologies is what characterised the successes observed by the tassas in Niger. Farmers decide whether or not to adopt and replicate a particular SWC technique based on how easy it is to implement, how it fits into the crop year and, especially, whether it has an immediate impact on production. Other than the tassas, measures such as stone bunds and demi-lunes have also been widely replicated in West Africa. These measures were largely appreciated by those interested in bringing degraded land back into production. They triggered a significant shift in the rural exodus as they offered appealing options for immediate returns and paid work.

An appropriate conservation programme strategy normally requires a slow and modest start to implement. Governments and funding agencies must be prepared to re-examine programme content and budgets after some time. It is essential to combine the short-term perspective with the longer-term view of how productivity increases and conservation will continue to be supported. In this context, it seems more relevant to develop `programmes' rather than `projects' to integrate SWC activities properly with long-term efforts to develop agricultural production.

The Niger project shows that the first step must be to reinforce mechanisms for the identification and analysis of local technologies and know-how. The project teams must be encouraged to undertake this kind of `inventory'. It also means that extension agents need to be trained in diagnostic methods for their individual areas of technical expertise. The adaptation of appropriate technologies entails close co-operation between farmers and researchers and means steering them toward the adjustment and local adaptation of principles tested elsewhere.

In most cases, the research and innovation potential (and not only "traditional practices") of the farmers themselves has not been adequately recognised and exploited. In view of the very limited facilities available for institutional research, and hence its inability to respond to diversity (and often to even perceive that diversity), only by mobilising the research/experimentation capacities of smallholders can innovation needs be met.

Once a promising technology has been identified and adapted by farmers, it is very important to support the local systems of farmer-to-farmer diffusion. This method of diffusion is low cost and usually produces very good results as well as creating an informal network between farmers, which can lead to other initiatives.


Local soil and water management provides the key to the transformation of the ecological and economic base of communities dependent on natural resources. Both these examples illustrate the need for fundamental changes in current water management policies and strategies.

Both examples illustrate the power and value, to individuals, communities and wider regions, of combining individual and collective knowledge and energy in soil and water harvesting and conservation.




One of the key issues facing farming systems with many small-scale producers largely dependent upon low yielding staple crops is that of income generation. Few farming households are completely susbsistence based, and cash earnings are vital not only for adequate nutrition and access to services, but also as a source of demand for local goods and services that provide livelihoods for many more households. Even where yields or cultivated areas are increasing, declining terms of trade and domestic prices for staple crops in many farming systems are leading to reductions in smallholder incomes, pushing producers ever deeper into poverty and strangling rural economic growth. Yet creating the initial conditions for rural income growth has generally proven difficult. Development projects tend to focus on strengthening technologies, human capital and infrastructure rather than on increasing farm incomes directly, and in any case their impact tends to diminish once the implementation period is concluded.

The following case study documents an example of substantial and sustained growth in smallholder family incomes that has occurred within a part of the Mesoamerican hillside maize-beans system as a result of diversification of small-scale producers into export horticulture. The case is particularly interesting because the beneficiary population is largely indigenous, many not even speaking Spanish, and because their domination of the U.S. market for snow peas has been achieved entirely on the basis of micro-level production by over 20000 family production units.

This impressive position has been reached without any external support from Government or development agencies. Rather it arose from the activities of a private sector that was, in turn, responding to emerging international market opportunities. However, the study argues that the effectiveness of the private sector actions, and hence the success of the small-scale producers, was greatly enhanced by the concurrent emergence of a private sector association dedicated to non-traditional products, and a series of national governments generally supportive of the needs of an emerging export sector. Thus major gains for poor, indigenous farming families resulted, in part at least, from a favourable environment for business development.


The Mesoamerican hillside maize and beans farming system occupies 650000 km2, and encompasses the long mountainous chain connecting North America to South America through Mexico and Central America16. It is bounded for most of its length on both the Pacific and Caribbean sides by the Coastal Plantation and Mixed Farming System. As is common in the region, the Mesoamerican system is highly dualistic, with large numbers of poor and extremely poor small-scale producers occupying marginal and sub-marginal hillside and slope lands, while the better soils of the valleys and intermittent lowlands within the system are incorporated into large estates or commercial family farms with coffee, sugar cane, rubber, cattle or other agricultural activities. The proportion of population in rural areas is high by Latin American standards, reaching over 60 percent in Guatemala17, and agriculture plays an important role in GDP, contributing as much as 28 percent in Nicaragua.

Poverty levels are high within the system - exceeding 80 percent of population in extreme poverty in the highlands of Guatemala, where this case is drawn from - and are often directly correlated with the percentage of indigenous people in the system (65 percent of the population in Guatemala, possibly exceeding 90 percent within the area of the farming system). Rising population levels have led to an increasing fragmentation of holdings and a concomitant reduction in farm sizes18, which in turn has increased pressure on land and water resources and led both to the expansion of the agricultural frontier (often onto steep or otherwise constrained soils incapable of sustained arable production) and to further poverty. In the 1980s, FAO studies estimated erosion and serious soil degradation to be affecting as much as 35 percent of lands in Guatemala, suggesting that the problem on slope lands within the Mesoamerican system was likely to be even more serious. The crucial importance of land is clearly illustrated by the numerous rurally-based armed conflicts which have erupted over the last 30 years within system boundaries.

Smallholder production of maize and beans is a core activity for most households and is largely consumed on-farm. Yields are low19, reflecting the limited use of external inputs and marginal soils. Coffee is the favoured cash crop where altitude and soils permit, but communities near urban centers may have traditions of fruit and vegetable production. Cattle can play an important role, but are of less importance at higher altitudes, except in the Guatemalan and Mexican altiplanos where sheep are common. More isolated communities often resort to seasonal or long-term migration as a means of supplementing cash incomes. Despite these alternatives, maize and beans continue to occupy a key role culturally, nutritionally, and financially for the majority of small-scale producers, and the sale of surpluses of these products provides the basic income for the majority of families. Yet real domestic prices for these commodities have been stagnant or declining in recent decades as a result of expanded external access to domestic markets, and lower levels of domestic market protection.

The presence of public services and infrastructure is weak throughout much of the system, and diminishes rapidly as one moves away from administrative centers. Widespread civil strife has added to the gulf between large urban centers and rural communities. Even in Costa Rica, often seen as a model for economic development in the region, recent data shows the human development index is nine times higher in key urban areas than in rural indigenous communities20. Over the next 30 years the future looks bleak for those small-scale producers dependant on maize and beans for their livelihood, and even bleaker for their children, who will inherit ever smaller and more degraded holdings from their parents.


Over a period of approximately 20 years, from 1974-1994, a number of key changes occurred in Guatemala that combined to profoundly affect the lives of over 150000 poor inhabitants of the Guatemalan rural highlands. In broad terms these changes can be grouped into three categories:

Although described separately below, in order to clarify the actions taken and illustrate clearly the role of each group, the development and impact of the three sets of factors were clearly inter-related, and should be understood as such.

The Emergence of Snow Peas and Broccoli in Guatemala

In the early 1970s, growing demand for snow peas (Pisum sativum) in the U.S. faced a major constraint; fresh supplies from California were only available from June to October, yet the frozen imports from Taiwan were viewed as poor replacements. An alternate source of fresh snow peas might be very profitable. Initial interest had focused on Chile, but in 1974 an American entrepreneur started to experiment with snow pea production in Guatemala. Snow peas were agronomically well suited to the altiplano of Central and Western Guatemala where the temperate climatic conditions permitted harvesting from October to May. Results from pilot plots were encouraging and a number of agribusinesses commenced production.

Demand for snow peas continued to grow rapidly through the 1970s and 1980s, but expansion in Guatemala proved difficult; obtaining land within the densely populated highlands, where few owners have legal title, was expensive and time consuming. Supply could not keep up with demand and the agribusiness corporations had to turn increasingly to independent producers to supply their buyers. By the beginning of the 1980s, a number of smaller producers started to bypass the agribusinesses and deal directly with small exporters and processors.

Despite their lack of formal education or capital, local indigenous producers enjoyed a number of advantages. Many of the early producers, at least, were familiar with horticultural production, having grown onions, tomatoes and similar crops for local markets. Their land was available at no cash cost, and their desire to maximise the use of family labour was ideally suited to a crop which required an input of 516 person/days/ha over a four month period.

With labour comprising 35 percent of total costs, even at the low wage rates paid in rural Guatemala, smallholders using unpaid family labour could earn extremely high returns from snow peas compared with traditional alternatives. A quarter hectare of snow peas (a typical holding) could generate US$500 before returns to land and labour were taken into account. By contrast, the same area of maize would earn only US$50. Agribusiness operations were unable to compete21. Furthermore, with supplemental irrigation, it was possible to produce two consecutive vegetable crops - snowpeas and broccoli - in one year and maize in the next. As a result, broccoli production, already known in Guatemala but previously not considered attractive by small growers, increased greatly in importance. The very small areas of land required to generate significant increases in household income also permitted producers to continue with maize and beans on the remaining cultivated areas. By the mid-1990s, an estimated 21500 families were involved in snow pea and/or broccoli cultivation, producing 23000 MT of snow peas and 43000 MT of broccoli annually on approximately 4350 has.

The growth of independent producers was only possible because of a parallel growth in intermediaries and exporters. Managing the daily output of so many producers scattered over perhaps 3000 km2 of poorly serviced rural highlands, and ensuring that it is selected, packed and shipped within no more than 24 hours of harvest, requires a sophisticated distribution system. No estimates exist of the number of intermediaries involved, but at the time of the studies in the mid-1990s, at least 50 companies were making regular export shipments of fresh snow pea during the harvest season; some using air freight, others reefers (the latter primarily to Europe)22.

By the 1990s many producers - estimated at more than 60 percent - had regular contractual arrangements with local intermediaries, who in turn generally represented specific exporters. For growers, the benefits of contractual production lay in access to working capital, generally in the form of seed and agrochemicals. For exporters, on the other hand, the opportunity to plan and co-ordinate harvesting dates and volumes so as to maximise volumes at key points in the price cycle was the key incentive. Some growers stayed independent, selling their output, often in lots of less than 100kg, at specialist auctions in communities around the altiplano. Starting at 5 p.m. and ending at around midnight, intermediaries with pickup trucks would receive contracted deliveries and bid for non-contracted supplies to fill their nightly quotas, either for direct delivery to an exporter, or for onward sale before dawn at one of five major wholesaling centers that evolved to serve the market. As much as 1.5 million lbs. (650 MT) per week might flow through these channels during the peak of the harvesting season. Through the 1980s and 1990s other smallholder non-traditional products emerged, increasing diversification. These included mini-vegetables, baby corn, sugar-snaps, blackberries and especially raspberries.

The Role of GEXPRONT

In 1978 the Government of Guatemala created a public-sector agency, GUATEXPRO, to promote national exports, but the institution lasted only two years. Nevertheless, it convinced a number of exporters of non-traditional products that, lacking the power of the traditional sugar, coffee or livestock sectors, they needed a forum that would provide influence and co-ordination to the mostly small and medium-scale enterprises active in the sector. Thus in 1982 the Guatemalan Non-Traditional Exporters' Association (GEXPRONT) was founded. GEXPRONT received a major boost in 1986 when USAID commenced financial support. Although GEXPRONT comprises five commissions, this case study deals only with that active in the area of agricultural products. By 1995, the agricultural commission of GEXPRONT (now referred to as AGEXPRONT) had over 250 registered members. In co-operation with its sister commissions, and often in collaboration with other private sector entities such as the Chamber of Industry and Commerce, AGEXPRONT lobbied for, and contributed to, a number of key changes within the export sector. It placed little emphasis on direct market development, although some activities were designed to support individual enterprises in their commercial relations with overseas buyers. Instead, AGEXPRONT focused on resolving systemic problems impeding its members in conducting and expanding their operations. These were primarily issues of importance to small businesses, and often involved the management of functions hitherto considered the sole mandate of the public sector. Apart from contributing to changes in Government policy (see below), the key changes instituted by AGEXPRONT in their first decade included:

By the beginning of the 1990s, yields had begun to stagnate for both broccoli and snow peas and supply was once again a problem despite the much wider range of non-traditional crops now being grown. Poor rural infrastructure rendered opening up new production areas difficult and the best lands for broccoli and snow peas had already been brought into production. Increasing concerns over chemical contamination were emerging in foreign markets. As a result, the support programmes offered by AGEXPRONT also started to change focus:

In the second half of the 1990s, GEXPRONT faced a further challenge; USAID financial support, which had been diminishing for some years, ceased. Member dues and income from sponsored activities were on their own insufficient to maintain the demand-driven research programmes and other high cost activities. Instead, AGEXPRONT started to forge ever-closer links with the Ministry of Agriculture, convincing the government that a private sector agency could make more effective use of public funds than could the Ministry itself. As a result, many of these programmes have since been maintained in operation, with public support.

The Role of Government

While there is no doubt that the key players in the success of snow peas and subsequent export crops have been the smallholders and the exporters with whom they trade, the Government has also played an important role over the last 20 years in setting an appropriate legislative and policy framework that has permitted the non-traditional export sector to thrive. Key measures have included:

In the late 1990s, the Ministry of Agriculture (MAGA) drastically reduced its staffing and internal expenditures, but increasingly funnelled significant amounts of public funds through GEXPRONT for small-producer support activities. It also formed regional development groups, incorporating the private sector, which have tried to co-ordinate private and public sector development in less favoured rural areas. AGEXPRONT has played a major role in the work of these committees.


The Results

There is no doubt of the enormous impact that the development of snow peas and broccoli exports has had on small-scale indigenous farmers of the Guatemalan altiplano. From 1980 to 1993, the Guatemalan share of the OECD market for fresh, frozen and processed vegetables quintupled, from 0.09 percent to 0.45 percent24, even while the commercial-scale production of these products was declining to almost nothing. By 1995, Guatemalan supplied one third of U.S. imports of snow peas; for a value of US$55 million per annum. And because no single family could command the labour, capital or water supply to cultivate a large area of these intensive vegetables, the returns were widely spread within the indigenous community, with average holding sizes devoted to the crop of 0.24 has. No producer was recorded with a planted area of more than 0.5 has.

By 1996, it was estimated that 21,500 indigenous families were involved in direct production of these two crops, generating estimated gross farm incomes in the region of US$30 millions. This equates to almost US$1400 per family. On the basis of an estimated 516 person days of labour per hectare for snow peas, and 191 person days for broccoli, it can be calculated that this family income was achieved for an average of 0.5-0.6 person years of labour input per family, or an earning of approximately US$2500/annum/person year employed25.

A further US$28 million was calculated to accrue annually to the wholesaling, processing, packing and export sector within Guatemala, some of which would benefit rural inhabitants engaged in collection, packing and transport activities. In fact, a study undertaken in 1994 estimated an indirect labour multiplier of 0.26 in relation to non-traditional agricultural activities26, suggesting that as many as 27000 families may have derived employment from these activities, without counting those occupied in producing mini-vegetables, raspberries and other later crops. With a conservative family size estimate of six persons, these two non-traditional crops may have contributed to poverty alleviation for over 160000 rural poor in Guatemala. Further more, these numbers do not take into account providers of good and services in rural areas, who were able to establish business in response to rising rural demand. These certainly exist but no data on their numbers ia available.

Key Contributing Factors

Although the innovation and risk taking of indigenous producers and small-scale trading enterprises was critical to the success of the snow pea/broccoli system in Guatemala, a number of other factors were probably equally important. The low costs of entry into both production and export of snow peas resulted in a system that was fiercely competitive, with no apparent advantages of scale such as were seen in broccoli freezing (perhaps contributing to the much lower returns for this crop). As a result, over 40 percent of the destination market price for snow peas was captured by the growers, a high proportion for a perishable export crop.

As important, however, was the role played by AGEXPRONT in creating (and in relation to Government, promoting) a framework within which such competitive behaviour could flourish, providing new entrants with market support and facilities that would otherwise have been very costly to develop. In general, AGEXPRONT resisted the temptation to pick winners, undertaking only a single exercise of this type in the period to 1996. Perhaps this was best; enthusiasm for asparagus, identified in the 1987 study, led to a number of large investments which were all failures.

Unusually for an exporters' trade association, however, AGEXPRONT took a leading role in promoting enterprise-producer linkages, seeing it as a key to increasing product availability, and hence turnover, for its members. From an early point, AGEXPRONT encouraged exporters in the same product line (melons, mangoes, snowpeas, broccoli, cut flowers, etc.), to form sub-commissions aimed at identifying and alleviating common obstacles to the development of that product. It was this strategy that directly led to the establishment of the USAID financed shared-cost field research programmes, and subsequently the privately funded extension services, as exporters agreed on a common need to deal with low yields, contamination issues or other problems. In 1997 AGEXPRONT created its newest sub-commission; for exporters of environmentally friendly products and services.

The relationship between the private sector and the Government of Guatemala has also been crucial for the rapid development of the non-traditional export sector, and hence ultimately for income generation among small-scale producers. Although Government attitudes have varied over the last 20 years, and not all policies have been to the sector's advantage (following the readjustment of exchange rates in the 1980s, there was a long period of overvaluation of the currency again in the 1990s), the Government has, in general, been supportive of small-scale enterprises. In the longer term, the willingness of MAGA to utilise AGEXPRONT as an executive arm for channelling and managing public sector funds has shown that international funding is not the only way for such private sector bodies to access the financing they need to continue their activities.


Small-scale non-traditional vegetable and fruit producers in Guatemala face intense competition from many other producers. Nonetheless, it has managed to maintain and even expand its position, utilising its favourable climate, low family labour costs, and efficient marketing system. The most serious threats it faces are a consequence of its greatest strength; its highly dispersed and atomistic production and marketing system. This has raised serious concerns over contamination, firstly from prohibited agrochemicals, and more recently as a result of the suspected biological contamination of raspberries with cyclosporin. With so many producers and exporters, determining the source of any contamination or infection is very difficult, and entire shipments may be contaminated by a single producer. A second possible threat to system sustainability are the difficulties still experienced by exporters in matching peak supply with periods of highest demand (and hence price).

Controlling output is certainly easier where production is concentrated among a small number of producers, but is not impossible for dispersed production systems. The future of the current system may depend on convincing small growers of the critical importance of proper chemical and sanitary management of their products, and the need to match planting times to market demand. Curiously enough, it may be these needs that propels participating indigenous farmers into 21st century production techniques, through the use of internet-based harvesting forecast systems, computerized coding of producer batches, and automated sampling and analysis of products.

On the positive side, it seems inevitable that the U.S. share of its own domestic market will decline significantly in the future, even as demand for exotic and high-value horticultural products and fruits continues to increase. California, the chief production area in the U.S., has many competing demands for its land, and the continued production of high labour-input crops, even behind high levels of tariff and pseudo-tariff protection, can have little long-term future. This may open up major new growth prospects for Guatemala, if Mexican competition can be held at bay27.

Finally, it is a cause for some concern that the sector has achived so little progress at creating added value through processing. In the mid-1990s, nearly 90 percent of the value of non-traditional agricultural exports was still accounted for by fresh products, and even the 10 percent that was processed was exported primarily to regional markets in Central America, not to cosmopolitan markets.



A number of important lessons can be drawn from the case presented above:


Because this case study describes a series of events, rather than a specific intervention, the question of replicability may be less relevant than for other case studies. Nevertheless, some comments can be made. None of the elements that contributed to the development of snow pea and broccoli exports in Guatemala are unique. A number of niche crops, requiring intensive labour and with limited barriers to entry, offer potential for diversification among poor smallholder populations in many farming systems. There are many countries where exporters' associations exist, although they are rarely as active as GEXPRONT and are difficult to sustain once external funding dries up. There are also many countries where Governments have made an effort to promote non-traditional exports. Finally, there are numerous examples of cases where the private sector has provided the driving force for the emergence of new non-traditional crops and products, whether they be cut flowers in Colombia, shrimp in Ecuador, or orange juice in Brazil. What is unusual in the Guatemala case, is the degree to which the private sector has created a successful partnership with poor, small-scale producers to dominate the largest market in the world for a specific commodity28. Other crops have also found success, using the same formula.

For Governments and external financing agencies, perhaps the key to replication of this experience must be to focus diversification and income support projects less on the producers themselves, and more on the environment in which they function. This environment clearly includes elements under Government control, such as infrastructure, legislative controls, fiscal policies and exchange rates, but also depends heavily on private sector activity in marketing, financing, input supply, and employment creation, none of which can be sustainably provided by the public sector, nor adequately dominated by producers themselves.

Yet the private sector is likely to face obstacles, sometimes very serious ones, to these activities. If these can be removed, the environment for diversification and income generation will be much more conducive and - as long as the underlying agronomic, socio-cultural and market conditions are appropriate - the probability of success greatly enhanced. However, the private sector must be convinced of the potential of the small-scale producers as partners. Small-scale enterprises are inherently more willing to accept this premise (large enterprises may find, in any case, that transaction costs are too high when dealing with small-scale producers). Furthermore, only the smaller enterprises are able to adequately identify and assess the importance of the constraints faced in developing atomized production and marketing systems. Hence the importance of enabling small and medium-sized enterprises to play a leading role in the process.

1 This paper is condensed from Tran and Nguyen (2001).

2 Addressed within an integrated national development programme that gives priority to the agricultural sector by improving access to inputs and new technologies.

3 However, it is also recognised that rice yields on some farms can sometimes exceed those achieved on research stations.

4 This case study has been condensed from Le (2001).

5 Bt-crops are those that have been engineered with genes, derived from the bacteria Bacillus thurigiensis. These genes encode for endo-toxins which are toxic to specific classes of pests.

6 Apomixis can be defined as the process of reproduction (e.g. seed production) without fertilization, therefore ensuring exact duplication of the genes of the parent stock.

7 This paper is condensed from Brinkman (2001).

8 An example is described in Nachtergaele and Brinkman (1996).

9 FAO 1999.

10 Swallow 2000.

11 Examples of practices in smallholder conservation agriculture are described in FAO (1997).

12 Examples of local development and production of machinery for conservation agriculture in Brazil are described in FAO (2000).

13 This paper is a synthesis of case studies prepared by Agarwal (2001) and Mascaretti (2001).

14 This case study is condensed from Gulliver (2001).

15 This case study draws heavily from Contreras, 1996 and Gulliver et al., 1996.

16 See Gulliver et al (2001) for further description of the Mesoamerican hillside maize and beans farming system.

17 Proyecto Estado de la Region 1999.

18 According to national agricultural census data, the number of holdings growing maize in Guatemala has increased from 321000 in 1964 to 667000 in 1996 - an increase of over 100 percent in 32 years. Average holding sizes of maize producers in 1996 was approximately 3.6 ha.

19 Average yields are less than 1.5 t/ha. for maize and 0.75 t/ha. for beans.

20 Proyecto Estado de la Region 1999.

21 Cost of production data for commercial operators in the early 1990s show that, with labour at Q.4/personday (equivalent to approximately US$1 at that time), they faced an additional US$2064/ha. in labour costs.

22 Frozen snow pea has never accounted for more than 10 percent of overall snow pea production, in contrast to broccoli which is exported almost entirely in frozen form.

23 Initially these cessees were set at Q.0.01 per lb., generating some US$65000 annually.

24 Inversiones y Desarrollo Corp., 1995.

25 It should be remembered, however, that this return does not take into account the value of the land utilised.

26 Samayoa Urrea 1994.

27 Mexico has managed to capture almost all of the production gains associated with declining tomato production in the Southern U.S.

28 Two other examples can be readily identified. The development of banana exports among small-holders in the Eastern Caribbean in the 1960s to 1980s was led by Geest, a private sector enterprise. Nevertheless, in Geest, growers faced a monopsony buyer, and thus were never equal partners, while the British Government played a major support role for many years. The case of horticultural exports from Kenya is also similar, but there the role of the commercial agribusiness sector in direct production has always been much greater.

Previous PageTop Of PageNext Page