FAO EMERGENCY COORDINATION UNIT IN BURUNDI: AN OVERVIEW - Jean Alexandre Scaglia

Jean Alexandre Scaglia
FAO Burundi

This paper summarizes FAO's seed aid assistance in Burundi. The context for delivery of humanitarian relief is a challenging one. Most parts of the country are plagued by intermittent fighting, drought is a frequent (and repeated) occurrence and malnutrition rates are exceptionally high in everywhere but the western strip.

To respond to this complex crisis situation, FAO first established an Emergency Coordination Unit (FAO/ECU) in 1996. Its overall objectives have been to alleviate the consequences of the ongoing conflict on the rural communities and improve access to food by increasing and supporting local staple food production; to coordinate agricultural relief operations so as to ensure a rapid response; with partners, to assess needs and target beneficiaries; and to keep the rural "backbones" functional in order to facilitate rapid deployment as soon as the conflict ends. More specifically, in relation to seed, FAO's work has had three specific thrusts: assessment, seed aid provision and rehabilitation of seed multiplication infrastructure. These are briefly discussed below.

Assessment

Assessment of the need for seed and tools has been guided by a range of data sources: crop harvest assessments; UNHCR and UN Office for the Coordination of Humanitarian Affairs statistics, WFP food assessment reports; early warning system networks and quick assessment through the REACT mechanism (a mechanism of UN agencies). Targeting has been an integral component of assessment, with criteria jointly defined by WFP and FAO and also approved by the government. Two broad groups have been distinguished through targeting, those needing agricultural assistance and a "seed protection ration" and those needing both agricultural assistance and food aid. To give an idea of scale, during the second season of 2003, 278 000 households were assisted, with 27 000 tonnes of food (equivalent to 100 days' ration) and 2 800 tonnes of bean seed.

Seed aid provision

Seed aid, distributed with a "seed protection ration" (20 days of food - in collaboration with WFP), has been given continuously since 1997, although the specific zones of distribution have varied markedly from one season to another. Bean seed, the crop most widely distributed, is generally sourced from within Burundi, and includes the local varietal mixes as well as the preferred yellow and large red-seeded grain types. Samples of seed distributed as emergency relief are tested by national institutions to meet recommended standards, and the end-use of seed on-farm is monitored after distribution to local communities. Generally, the results look promising: surveys suggest that 70 percent and upwards of the seed distributed is actually sown, with only small proportions consumed. Aside from beans, vegetable seed, hoes, banana plants and sweet potato cuttings are sometimes included in the aid packages distributed.

Rehabilitation of seed multiplication infrastructure

FAO also carries out activities anticipating some of the longer-term seed security challenges of the country. The main objective is to rehabilitate rapidly the national structures in order to make available seeds and vegetative materials of good quality at the community level and simultaneously to decrease the need for direct distributions. Presently, the focus is on effective multiplication of potato, rice, bean, soybean and maize, regionalized according to more favourable crop-specific production zones, and drawing from a pool of improved varieties selected prior to the current crisis by the national research programme, ISABU (Institut des Sciences Agronomiques du Burundi). The multiplication scheme is streamlined, using "basic seed" to produce "quality declared seed" and then to distribute it via traditional seed exchange networks and direct market sale, as well as free distribution among the more vulnerable (widows, returnees, etc.) In terms of quality-declared seed, since 2001 (second season), about 2 210 tonnes of rice, 496 tonnes of bean, 2 100 tonnes of potato, 481 tonnes of soybean and 103 tonnes of maize have been produced.

Peace is the great unknown factor in terms of planning future FAO activities within Burundi. When enhanced security and peace become a reality, FAO will turn its attention towards the definitive resettlement of refugees and IDPs; a demobilization of arms programme; the promotion of measures to bolster sustainable food security countrywide; and the rendering of widespread support to the Agricultural Redeployment Programme, as defined by the Ministry of Agriculture and Livestock.

PLACING THE FARMING FAMILY AT THE CENTRE OF THE SEED DEMAND PROCESS: CRS BURUNDI'S EXPERIENCE WITH SEED VOUCHERS AND FAIRS IN KIRUNDO PROVINCE - Stephen Walsh, Bonaventure Ngendahayo and Christophe Droeven

Stephen Walsh, Bonaventure Ngendahayo and Christophe Droeven
Catholic Relief Services, Burundi

Direct seed distribution relies on the formal seed sector and is characterized by a logistical command and control approach where community involvement is minimal, where the residual economic impact of seed purchases on the community is negligible and where, at the point of receipt, farmers have virtually no choice as to what seed and other farming inputs they receive. This paper will explore the experience of Catholic Relief Services in Burundi with the seed voucher and fair approach in Kirundo Province, northeastern Burundi and provide insights for seed aid practitioners in Burundi and elsewhere.

In October 2001, CRS Burundi piloted the seed voucher and fair approach to respond to seed needs for 517 farming households in Kirundo Province, northeastern Burundi. In the subsequent 16 months and over the course of three agricultural seasons, CRS employed this same approach in Kirundo to respond to the seed needs of nearly 33 000 farming families. Approximately US$180 000 was injected into the Kirundo economy as a result of these fairs, with the average gross income per seed trader of US$160.

Background to the area of intervention

Kirundo Province is in the extreme northwest of the country, bordering Rwanda and covering an area of 1 700 km². It has a population density of 230 people/km² and an average farming area of 0.8 ha/family. Small farm cultivation provides for over 90 percent of the population's livelihoods. This region is traditionally a bean and sorghum producer, but bananas, coffee, cassava and sweet potatoes are also cultivated there.

Agricultural production and food security at the household level have been devastated by the combined effects of drought and political crisis. Kirundo has experienced a severe rain shortfall with declines of 70 percent from the norm for 2000 and 2001.

Principles of the CRS seed system diagnostic framework

The CRS diagnostic framework under which seed systems are assessed is based on distinguishing between access and availability. Derived in part from the entitlement approach of Amartya Sen, this diagnostic framework highlights that total output and availability is only one of several factors that determine entitlements, that is, the bundle of commodities over which a farming family establishes command in order to meet its seed needs.^[12] Entitlement here is defined as endowments from land holding, labour, cash, trade, kinship and social ties.

Even though this fact is elementary enough, it is remarkable that food analysis and seed analysis are often conducted merely in terms of production and total availability rather than taking note of the processes through which people establish their entitlements to food and seed (Drèze and Sen, 1989, pp. 22-25).

The seed security framework (Table 1) allows a deeper and more robust diagnosis of seed security by providing seed aid practitioners with a framework that, at times, can shift the focus from availability (output) to access (entitlement).

TABLE 1
Seed security framework

Source: Remington et al., 2002.

TABLE 2
Seed system problems and appropriate responses

Source: Sperling and Cooper, 2003.

The shift from viewing seed need as output or availability driven to being access or entitlement driven has profound implications for seed aid practitioners. Over the past three years, CRS has pioneered the application of this seed security framework through the seed voucher and fair approach (Remington et al., 2002). This approach has been used to address the emergency seed needs of approximately 150 000 farming families in 14 countries in Africa.

The "availability" or "access" characterization of the seed system diagnostic framework points to the need for sharply differing responses by seed aid practitioners. Sperling and Cooper's seed system matrix further categorizes these parameters under "acute"and "chronic"while suggesting four distinct appropriate responses (Table 2).

Characteristics of the farmer seed system in Burundi

The Burundian farmer seed system is characterized by broad crop and varietal diversity and the continuous search for new seed and new varieties.^[13] Seed production is usually part of crop production and there is often significant seed acquisition off-farm. Poor farmers and wealthier farmers access seed differently - poorer farmers are much more likely to meet their seed needs through the market whereas richer farmers tend to access seed from their own stocks.^[14] The farmer seed system is also marked by a limited ability to access new materials.

It is common for bean farmers to access their bean seed from a number of different channels within a single season. These channels include own stocks, market, social networks and limited amounts of new material for experimentation from extension agents or research. Strategies for seed acquisition vary by wealth with a pronounced increase in the use of the market as a seed source among poorer households (David and Sperling, 1999).

Chronic stress in the farmer seed system in Burundi

Chronic stress in the Burundian seed system can be characterized as having two main thrusts, a physical and psychosocial insecurity stress driven by the ongoing war and a demographic stress driven by reduced land holdings (Figure 1). Both of these stresses result in a decrease in seed holdings at the household level and ultimately to a reduced capacity of the household to exercise entitlements from land.

FIGURE 1
Sources of chronic stress in the Burundian seed system

Physical and psychosocial insecurity stress is caused by theft and looting, which are by-products of the ongoing war in Burundi. This stress leads to less risk-taking by seed holders. Seed holders, both small traders that form the backbone of the farmer seed system and individual farming households, may reduce holdings as a means of mitigating risk. The result is a disruption of market channels in the farmer seed system and reduced seed holdings at the household level.

The direct seed distribution experience

In Burundi, as elsewhere in Africa, diagnosis of emergency seed needs has been based on household food security assessments, without distinguishing between access and availability issues. Assessments have been produced and based on seasonal calculations, without much regard for seed system problems that may be more chronic. Such assessments are liable to misdiagnose chronic problems as "acute". This misdiagnosis may manifest itself in repetitive or regular characterizations of the seed problem in a given region, territory, or country as being "acute" without adequate analysis and appreciation of more systematic and chronic factors.

Direct seed distribution does not place the farming family at the centre of the process; rather, the farming family is typically a passive recipient of relief. In the best circumstances, the recipient farming family or its local representative is queried as to its seed needs and the resulting relief package is determined in consultation between local authorities and the implementing agency. At the point of receipt of conventional seed relief, there is rarely any distinction among what farming families receive despite their differences. Uniformity and standardization are the watchwords for conventional seed relief.

The seed voucher and fair experience

The seed voucher and fair approach places the farming family at the centre of the process by which its seed demand is met. In the worst circumstances, the recipient farming family negotiates for seed with a large number of seed traders and selects crops and varieties that best meet its needs. The resulting relief package is determined through consultation between voucher holder and seed traders.

The package is dependent on the quantity, variety and quality of seed brought to the seed fair by farmers and traders in the area of intervention. At any one seed fair, a multitude of seed packages are chosen by farming families. Choice and individual empowerment are the watchwords for seed vouchers and fairs.

The role of the implementing agency is to act in consultation with local authorities and in coordination with other seed aid implementers to target the area of intervention and to determine the beneficiary list in consultation with the local authorities and beneficiary representatives. The dates and locations of seed fairs are determined through the same consultative process.

Farmers and traders are sensitized to the expected monetary value of the seed fairs in their area, and the dates and locations.^[15] The intent of such sensitization is to insure that an adequate quantity, variety and quality of seed is present on the day of the fairs. This sensitization is through multiple channels: churches, community associations, governing authorities and local and international NGOs operating in the area of intervention. Seed traders are registered prior to fair commencement and seed is inspected. Participants are sensitized in advance with a focus on educating voucher holders that the vouchers are akin to cash and only redeemable through the seed fair.

Kirundo seed fairs: what seed was exchanged?

At the Kirundo seed fairs, the cost of seed per beneficiary farming family was approximately US$6. The average seed package received by voucher-holding farm families over the three agricultural seasons was 20 kg beans, 1 kg sorghum, 0.5 kg maize and 0.33 kg groundnut. Prices for seed exchanged during these seed fairs were also quite competitive with local market prices.

Seed fair prices were on average 12 - 20 percent above local market prices, the increase being due to the temporal nature of the vouchers. Seed traders at the fairs know that vouchers are valid only on the day of the fair and thus voucher holders are required to spend their vouchers.

TABLE 3
Kirundo seed fairs: bean varieties available and quantities sold (kg)

Figure 2 synthesizes the type and quantity of seed on offer and that actually sold over the course of 31 discrete seed fairs held over three agricultural seasons in Kirundo.

Seed availability was not a problem as there was nearly one-third more seed supplied than demanded.

The preponderance of bean in the Kirundo seed system is evidenced by the rich diversity on offer at the Kirundo seed fairs. There were a total of 12 different bean varieties on offer during the fairs (Table 3).

FIGURE 2
Kirundo seed fairs: seed traded

TABLE 4
Kirundo seed fairs: value of seed sold over three agricultural seasons

*900 F Bu = 1 US$

Varieties here are defined as self-identified by local farmers. Hence, "tout venant" or "coming from everywhere" (i.e. a mixture of bean varieties) is included as a variety. The composition of "tout venant" differs by trader and by location. Voucher holders make clear qualitative distinctions in assessing "tout venant".

Kirundo seed fairs: assessing the economic impact

Beyond responding to relief seed demand in a timely and efficient manner, seed fairs stimulate local markets through injecting capital into the economy by way of independent entrepreneurs (both large and small seed traders). Table 4, based on payout lists from 31 discrete seed fairs in Kirundo, illustrates how seed fairs can equitably inject capital into cash-strapped local economies and provide important income opportunities for women.

The relatively equitable injection of capital into the economy as a result of the seed fairs is evidenced by the large number of traders participating and the median gained per trader, which ranged between US$50 and US$130 over the course of three seed fair seasons.

The number of female traders, as a percentage of all traders, grew considerably over the course of three agricultural seasons. This large number of participating female traders seems to indicate that seed trading presents an important employment opportunity for rural women in Burundi.

Exit interviews conducted among seed traders indicated that seed fair proceeds would be used to pay for health care and education expenses, to help to build homes and for investment in agriculture.

Over the 31 different fairs held during these three agricultural seasons, there was an average of 37 seed traders per fair, with a maximum of 72 and the minimum of 12. Of these traders, 48 percent brought between 200 and 1 000 kg of seed to the fairs while 37 percent of all traders brought less than 200 kg.

TABLE 5
Kirundo seed fairs: voucher holders over three agricultural seasons

The traders at these fairs were an assorted group-young and old, male and female, young men in their late twenties, grandmothers paired with children in their late teens, traders on bicycles, others with rented vehicles and a very few traders with their own vehicles. The one common denominator was that they were all independent entrepreneurs.

Table 5 shows the percentage of male and female voucher holders over the course of three agricultural seasons. Male voucher recipients, as a percentage of all recipients, increased by nearly 50 percent between the first and third seasons of fairs. This may indicate classic rent-seeking behaviour on the part of male household members. As the seed voucher and fair system becomes appreciated by households receiving vouchers, male household members - accustomed to working with cash crops - may increasingly seek to represent the family at a seed fair.

Kirundo seed fairs: lessons learned

The evidence from the Kirundo seed fairs supports the assertion that even in emergencies, farmer seed systems are resilient and can often be expected to be the prime source of farmer seed in good times and bad.

Moreover, the Kirundo seed fairs demonstrate that when the seed security problem is diagnosed as being "access" as opposed to "availability" driven, a voucher-based response relying on the farmer seed system can effectively address the access gap. This access gap is addressed through the seed fair process of subsidizing demand at the household level, stimulating supply to respond to the demand and providing a regulatory framework through which supply and demand meet.

While a formal comparison of direct seed distribution versus the seed fair approach is beyond the scope of this paper, the use of seed fairs in Kirundo was a more appropriate response than direct seed distribution. Vulnerable farming families were provided with adequate seed quality and supply in a timely and efficient manner; the most vulnerable were empowered with choice and placed in the centre of the process by which their seed demand was met; and there were residual economic benefits accruing to the local economy as a result of the seed traded.

CRS Burundi's experience in Kirundo also points to how the seed voucher and fair approach can strengthen the farmer seed system. The evidence suggests that this occurs in three ways: by letting farmers strategize about which crops and varieties they should use in stress times; by letting farmers continue to access seed through traders they know and whose quality standards they know; and by supporting local seed traders who will continue to serve farmers, with or without seed fairs.

The major insight that should be drawn from the CRS experience with seed fairs in Kirundo is the need for greater focus and attention to be paid to farmer seed systems. The fact that the farmer seed system, as opposed to the formal system, remains the dominant channel through which seed needs are met - particularly by more vulnerable households and even during times of profound seed system stress - points to the need for an increase of resources to be channelled to the farmer seed system.

References

David, S. & Sperling, L. 1999. Improving technology delivery mechanisms: lessons from bean seed systems research in Eastern and Central Africa. Agriculture and Human Values, 6: 381-388.

Drèze, J. & Sen, A. 1989. Hunger and Public Action. Oxford, UK, Clarendon Press.

FAO. 1997. Developing seed security strategies and programmes for food security in developing countries. In Proceedings: International Workshop on Seed Security for Food Security. Rome.

FAO. 1998. Proceedings of the International Workshop on Developing Institutional Agreements and Capacity to Assist Farmers in Disaster Situations to Restore Agricultural Systems and Seed Security Activities, 3-5 November. Rome.

Remington, T., Maroko, J., Walsh, S., Omanga, P. & Charles, E. 2002. Getting off the seeds and tools treadmill with CRS seed vouchers and fairs. Disasters, 26(4): 316-328.

Sen, A. 1981. Poverty and Famines. Oxford, UK, Clarendon Press.

Sperling, L. & Cooper, D. 2003. Understanding seed systems and strengthening seed security. Background paper prepared for the workshop Effective and Sustainable Seed Relief Activities. Rome, 26-28 May 2003.

Sperling, L. 1994. Analysis of bean seed channels in the Great Lakes Region: South Kivu, Zaire, southern Rwanda and select bean growing zones of Burundi. Summary of three case studies. Africa Occasional Publication Series No. 13. Cali, Colombia, International Center for Tropical Agriculture.

SEED RELIEF IN MOZAMBIQUE: A REVIEW OF RECENT INTERVENTIONS - Carlos Dominguez, Sérgio Gouveia, Joaquim Cuna, Stefano Gasparini and Birte Hald

Carlos Dominguez
International Crops Research Institute for the Semi-Arid Tropics

Sérgio Gouveia
Directorate of Economics of the Ministry of Agriculture and Rural Development

Joaquim Cuna
Directorate of Economics of the Ministry of Agriculture and Rural Development

Stefano Gasparini
FAO Input Distribution Unit

Birte Hald
FAO Input Distribution Unit

Introduction

Mozambique lies in the semi-tropical region of southern Africa, bordered by the Indian Ocean on the east. Its agricultural areas are characterized by erratic rainfall. About 30 percent of farming areas are located in a semi-arid region in the southern part of the country. Agriculture benefits from an intricate river system, but is also affected by river flooding during periods of heavy rainfall. Irrigation infrastructure is not developed except in small patches, where mostly commercial crops are grown. Agricultural production is vulnerable to frequent floods and droughts; emergency or relief seed distribution programmes have been organized almost every year during the past ten years (Rohrbach et al., 2001).

Agriculture in Mozambique is dominated by the "family sector", that is, subsistence farming on small landholdings in which a diverse range of crops are grown. Households produce largely for their own consumption, but surpluses are sold in incipient markets in the villages or along the main roads. This smallholder agriculture accounts for around 90 percent of the country's planted area. Over 70 percent of the population depends on agriculture for subsistence. Women are the main workers and producers and, by tradition, are responsible for the family's food security. Subsistence farmers use mainly farm-saved seed (Dominguez, 2001), supplemented by small amounts of seed/grain from the market, with a few farmers using small amounts of commercial seed. Commercial agriculture is incipient but growing as new marketing opportunities develop. This sector uses certified seed produced by a few seed companies or imported from neighbouring countries.

Prevailing means of seed exchange

According to Howard et al. (2001), "Most of the Mozambican production systems and almost all smallholding production are characterized by very low levels of improved input use." Seeds are the most important input - and often the only input used. Smallholders are conscious that seed is an essential input for food security and retain planting seed from their harvests. This seed system is diverse in terms of crops, varieties, adaptation to environmental conditions and the local methods used to store and safeguard seed. Women are primarily responsible for saving and maintaining seed.

Farmers use a wide range of local varieties or landraces that have specific names in each region. The extent of genetic similarity (or differences) between landraces in different regions is not fully understood. According to Ferguson (ICRISAT, 2002), varieties with the same name have widely different morphological and genetic characteristics and little homogeneity.

Local varieties are well adapted to specific environmental conditions. They have specific agronomic and taste characteristics developed through many years of selection by farmers. Because of this continuous selection, varieties with the same name might actually be different. Germination and physical purity in local varieties are acceptable, but yield losses due to late planting are common. Many farmers use multiple plantings to minimize the risk of crop failure. For example, farmers in southern Mozambique, where rainfall is particularly uncertain, plant some maize after each rain. Multiple plantings are also used in other crops, such as groundnut. Even though the rainfall season is short, farmers tend to plant their seed in stages, prepared to sacrifice some yield (due to late planting) in order to spread their risks. Planting the same crop or variety on highland (vulnerable to drought but relatively safe from flooding) and lowland (susceptible to floods) areas is also used as a food and seed security strategy (ICRISAT/INIA, 2002).

Farmers commonly exchange seed with their neighbours or with farmers from nearby villages. Seed may be provided as a gift, as a loan to be repaid at harvest, or exchanged for labour or other products. In many villages, some farmers are recognized as "seed providers". Seed donations are more common among relatives but seed exchange is always practised if seed is available.

The grain market is an important source of seed. Farmers buy grain for use as seed, but are careful in selecting the right variety. Market traders bring in grain from distant areas and farmers are aware that not all varieties are adapted to the local conditions.

Small-scale farmers rarely use commercial seed, except for vegetable crops. There are several reasons for this, mainly the limited number of retail outlets in villages and the high cost of seed compared to grain.

In summary, small-scale farmers use a combination of sources to obtain planting seed. Data from the Ferguson study (ICRISAT, 2002) indicated that, under normal conditions, the main source is their own seed (72 percent) supplemented with grain purchased from the markets and/or gifts or loans from relatives and friends (16 percent). A few farmers (12 percent) rely solely on purchasing grain for seed. Surprisingly, it is the poorest farmers who most rely on this purchased grain for seed - possibly a day-to-day survival strategy for families with very low incomes.

Impact of disasters on the seed system

During emergencies (droughts and floods) farmers travel further to other villages/regions with similar environmental conditions to exchange or buy seed. The local market may remain an important source of seed even during emergencies, but particularly in this situation the poorest farmers cannot afford to buy seed. Furthermore, not all villages have markets; farmers in remote areas far from markets are more vulnerable to seed insecurity.

According to Ferguson (ICRISAT, 2002), the main seed sources following a flood emergency were a combination of seed relief and market purchases (44.6 percent), followed by a combination of seed relief and gifts/loans from other farmers (16.7 percent). Only 13.9 percent of the farmers interviewed used relief seed exclusively. Relief seed is not necessarily assimilated into the farming system. Out of 38 farmers who received cowpea seed under relief programmes, only 8 farmers were still growing the variety after two and a half years.

In contrast, most of the smallholders who received the maize variety Matuba (developed by the National Institute for Agricultural Research) are still using it. Farmers like the variety's characteristics, which are similar to those of local types, and therefore save the seed for future use (Libombo and Uaiene, 1999). Further studies should be undertaken to examine the effect of seed relief on genetic diversity in the farming system.

Floods may have a more severe effect than drought on genetic erosion but, in the long term, drought also affects seed security as farmers exhaust their seed and food reserves. According to Ferguson's study (ICRISAT, 2002), cowpea varietal diversity was higher in areas not affected by floods compared with the affected areas (0.552 versus 0.476), suggesting that some recovery had taken place but that the impact on the diversity was still evident.

Seed and tool distribution has been widely used to assist the poorest families to re-initiate farming after an emergency. NGOs and other agencies have distributed many tonnes of seeds under these programmes. These programmes have suffered from several problems. Because Mozambique's seed industry is weak, emergency requirements cannot be met from existing seed stocks and therefore non-adapted or foreign varieties have been introduced to affected areas. Farmers rarely retain these non-adapted varieties in subsequent seasons, although small quantities may be found in fields, mixed with local varieties. Few, if any, studies have been undertaken to determine the effectiveness of the seed and tools approach in restoring crop diversity of the main crops used by farmers in disaster-affected areas.

Seed fairs

Recently, seed fairs have been proposed, tested and implemented as an alternative to seed and tools distribution and this approach has proven to be successful. They are now designated as input trade fairs (ITFs) and have been coordinated by the FAO Input Distribution Unit (IDU) and local government agencies in southern Mozambique. The International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) has provided technical support by adapting a methodology developed by Catholic Relief Services (CRS) to Mozambican conditions.

Under this approach, seed distribution is more market/commercially oriented. Instead of receiving free seed, beneficiaries receive vouchers of fixed value that can be exchanged for seed at a specially organized seed fair in their area. Seed sellers at the fair include local farmers, seed companies and grain retailers. They collect vouchers as payment for their seed and after the fair they exchange the vouchers for cash from the organizers. This system has a number of advantages. It builds on and stimulates traditional methods of seed exchange. Donor funds are spent largely within the community, rather than on bulk seed purchases from outside, benefiting local development. Moreover, the system can be used for relief while simultaneously facilitating seed market development, thus ensuring sustainability. The Government of Mozambique has adopted this system for seed relief and is encouraging local authorities and NGOs to implement it. A number of ITFs are planned for implementation at the beginning of next season, supervised by FAO-IDU, and with financial assistance from the United Kingdom Department for International Development (DFID).

The main advantages of the ITF approach are the following.

a) It allows beneficiaries to choose what they need among a range of agricultural inputs (e.g. selecting the type, variety and quantity of seeds and tools). The community is able to use external aid more effectively, while addressing real needs and suitability to local practices and conditions.

b) A significant percentage of external aid is retained within the target area, as against disbursement elsewhere to procure inputs or free distribution. The impact of the aid intervention on the community is thus greatly increased.

c) Local products are valued, thus encouraging their production and trade. Putting a clear value on local crops/varieties also has beneficial effects on genetic diversity and food security - a diverse crop mix, especially of locally adapted varieties, can be instrumental in reducing risks of crop failure. Smallholders' terms of trade are also benefited.

d) New crops and varieties can be traded in small quantities. This encourages farmers to buy and test them in their own plots.

e) Seed fairs provide people and institutions with a forum to meet and discuss issues related to agricultural assistance. This hands-on approach is extremely direct and was impossible in the older approach.

f) One aspect that could (and should) be promoted is to use the fair as an opportunity to raise awareness about HIV/AIDS.

Lessons learned from experiences in 2001 and 2002 have helped refine the methodology. For example:

a) The system is simple and can be implemented easily by local government agencies or NGOs at the local level, or at wider level by a responsible organization. Staff training is needed, but the necessary information is available.

b) Voucher values, determined exclusively on the basis of the value of previously adopted input kits, turned out to be too high in some areas. Overly high values led to a lottery syndrome: when a large sum of money has to be spent in a short time, normal selection mechanisms and purchasing judgements may be abandoned. This has also led, occasionally, to sellers artificially increasing prices.

c) Smallholders are willing to buy seeds if they are available at a convenient nearby location and sold in small packages that they can afford.

d) Seed of hybrids and new, unfamiliar varieties should be promoted separately. Farmers must be clearly and correctly informed about their advantages and disadvantages.

e) The selection of beneficiaries was sometimes imperfect. Selection should be done by the local authorities in collaboration with local people and organizations, with clear guidelines. Information on selection criteria and methods is available.

Planning interventions in the local seed system

After the terrible floods in 2000, ICRISAT and the Ministry of Agriculture and Rural Development (including INIA, the National Directorate of Agriculture [DINA] and local authorities) conducted a study sponsored by USAID. The purpose of the study was to identify a methodology for assessing seed requirements in emergency situations that takes into account farmer knowledge, attitudes and local seed acquisition practices, and does not disrupt normal market behaviour. Results from the study were summarized in Guidelines for planning local seed system interventions (ICRISAT/INIA, 2002).

This methodology comprises three modules:

• Module 1: Preparing a seed system profile (SSP). Describes the steps involved in preparing the SSP and where and how each action should be conducted. Also describes the collection and analysis of information from various primary and secondary sources used in the development of the SSP.

• Module 2: The framework of assisting seed security. Presents a simple matrix to describe the characteristics of a disaster, the impact on agricultural production, the availability of and access to seeds and the identification of seed-related problems. Includes tips to help community leaders assess the impact of the disaster.

• Module 3: Identifying the most appropriate interventions. Problems were identified in Module 2. This module helps identify the causes of the problems and short- and long-term interventions needed.

This methodology can be used to gain a better understanding of the seed systems used by farmers at local level. The published Guidelines spell out what, when and how to act when a natural disaster (mainly drought or flood) affects the normal functioning of the seed system. Interventions should be planned and implemented only after the problem has been clearly understood and defined so that they meet the real needs of those who are genuinely affected. This methodology has been successfully tested in three districts and is now being implemented in 20 others. ICRISAT will coordinate this scaling-up, with sponsorship by DINA through the PROAGRI system (Agricultural Assistance Programme). The methodology may also be adapted for implementation together with other existing seed assessment approaches, or to complement existing early warning and monitoring systems for food security. It may also prove useful in institutional analysis to identify rehabilitation needs.

As local authorities and NGOs improve their understanding of the seed system, interventions will become more rationalized and sustainable. The dynamics of local seed systems will be enhanced, thus improving small-scale farmers' access to good quality seed.

References

Dominguez C. 2001. Sistema informal de sementes: causas, consequências e alternativas. Second edition. Maputo, Ed. Associação do Centro Cultural de Matalana.

Howard, J., Low, J., Jeje, J., Boughton, D., Massingue, J. and Maredia, M. 2001. Constraints and strategies for the development of the seed system in Mozambique. Research Paper Series. Research Report No. 43E. Maputo, Directorate of Economics of the Ministry of Agriculture and Rural Development. Available at http://www.aec.msu.edu/agecon/fs2/mozambique/wps43e.pdf (accessed November 2004).

ICRISAT. 2002. Assessment of the impact of the 2000 floods on crop diversity in Mozambique. Final Report prepared by Morag Ferguson. Nairobi, Plant Biodiversity and Genomics Center.

ICRISAT/INIA. 2002. Guia para planificar intervenções no sistema local de sementes. Maputo, ICRISAT/INIA.

Libombo, M. & Uaiene, R. 1999. A cadeia de sementes em Moçambique: limitantes encontradas desde a libertação de semente melhorada, sua produção e distribuição. Paper presented at the Workshop on the National Seed System, Maputo, 16-17 June 1999.

Rohrbach, D., Low, J., Pitoro, R., Cuco, A., Massingue, J., Boughton, D., Rafael, G., Paulo, A. & Jocene, D. 2001. Investment priorities for the development of Mozambique's seed system. Research Paper Series. Research Report No. 44E. Maputo, Directorate of Economics of the Ministry of Agriculture and Rural Development. Available at http://www.aec.msu.edu/agecon/fs2/mozambique/wps44e.pdf (accessed November 2004).