Full text of DOI:10.1111/1467-7717.00211

Disasters, 2002, 26(4): 343–355
Do Farmers Need Relief Seed? A Methodology
for Assessing Seed Systems
Catherine Longley
ODI/Intl. Crops
Carlos Dominguez
ICRISAT
Research Institute for the
Semi-arid Tropics (ICRISAT)
Momade A. Saide
ICRISAT
Wilson José Leonardo
ICRISAT
This article outlines a methodology to help agencies better determine whether or not
relief seed is needed by farmers affected by disaster. A brief review of current seed
needs assessment procedures in southern Somalia and Mozambique illustrates
problems of knowing which crops and households are affected, the importance of seed
access (not just availability) and the need to plan interventions earlier than at present.
The development of a Seed Systems Profile (SSP) is proposed to understand better both
the socio-economic and agro-ecological aspects of farmers’ seed systems. A five-step
framework for assessing seed systems in disaster situations is also presented. These
tools are currently being tested and further refined in Mozambique. A better
understanding of farmers’ seed systems will allow for the development of relief and
rehabilitation interventions that effectively enhance the resilience and reduce the
vulnerability of these systems.
Keywords: Somalia, Mozambique, Seed Systems Profile, assessment.
Introduction
Existing guidelines on emergency seed provisioning (e.g., ODI, 1996; Johnson, 1997;
Chemonics Intl., 1996) recognise the need to determine first whether or not relief seed
inputs are needed by farmers affected by conflict or natural disaster. However, they
contain very little in the way of detailed, practical advice on exactly how to do this.
Usually because of inadequate prior assessment, many seed interventions are based on
an assumed rather than an actual need for seed. There are certainly situations in which
farmers do need seed, and appropriate and timely seed inputs can bring very significant
improvements to agricultural production and food security. Yet there are also an
increasing number of studies showing that some emergency seed interventions have
very little impact in relation to their high cost (Jones et al., 2001; Longley et al., 2001).
Although the ability of agencies to implement relief seed projects has improved
markedly over recent years, the rationale on which such projects is based is now
coming under question (Jones et al., this volume). Contrary to assumptions, farmers’
©
Overseas Development Institute, 2002.
Published by Blackwell Publishing, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA.

344
C. Longley, C. Dominguez, M.A. Saide and W.J. Leonardo
seed systems are remarkably resilient, even in the face of severe natural disasters or
war (Friis-Hansen and Rohrbach, 1993; Longley and Richards, 1998; Sperling, 1997).
Unless such systems are better understood, there is a risk that repeated relief seed
interventions may inadvertently work to weaken rather than strengthen farmers’ seed
systems.
This article addresses the present lack of adequate techniques for assessing the
need for seed system support in protracted emergencies and disaster-prone areas. It
describes a method, Seed System Profiling (SSP), which can be used to compile
information about the ways in which farmers manage the seed of various different
1
crops. When used in conjunction with an assessment framework such as that in Box 2,
SSP information allows for a better understanding of the impact of a disaster on seed
systems. The tools presented in this paper are currently being tested and further refined
in Mozambique.
Present approaches to seed needs assessment
In general, there is a wide range of assessment approaches and procedures, each suited
to different types and stages of an emergency situation and the requirements demanded
by the assessment. This article is primarily concerned with detailed ‘one-off’
assessments (suitable for protracted emergencies) and longer term monitoring (suitable
for repeated acute or chronic stress situations) (see Jaspars and Young, 1995).
Southern Somalia
In the 10 years following the fall of the Siad Barre regime in 1991, between 2,000 and
4
,000 metric tonnes of cereal and pulse seed was distributed on an annual basis in
southern Somalia in response to political insecurity, drought and floods (Longley et al.,
2
001). The quantity of seed needed was calculated in relation to food needs, which in
turn is determined by relative vulnerability. Vulnerability is assessed by the Food
2
Security Assessment Unit (FSAU) using the Household Food Economy Approach (see
Seaman et al., 2000).
FSAU has defined a total of 20 Food Economy Areas across Somalia which
describe the different types of terrain (riverine, pasture, highland) and the various forms
of livelihood (pastoral, agro-pastoral, agricultural), together with the dominant crop or
livestock types in each area. Baseline profiles of each of these Food Economy Areas
have been undertaken, detailing the ways in which households of different wealth
groups normally obtain their food and other income. The FSAU uses the baseline
information contained in the food economy profile for analysing monthly and seasonal
data on areas planted, production and yields, livestock conditions, income-generating
activities and coping strategies (Nisar Majid, pers. comm.). Levels of vulnerability are
assessed according to the nature of the external shock (drought, flooding, political
insecurity, price fluctuations) and the relative ability of local communities to cope with
these shocks.
While the information collected by the FSAU relates largely to food security,
it has also been used by the FAO to determine seed security: areas which were found to
be food insecure were also deemed to be seed insecure. Calculations regarding the
actual quantities of seed required were made according to figures released by the FSAU

A Methodology for Assessing Seed Systems 345
on the harvest output together with numbers of vulnerable households and the average
seeding rates of different crops. It was generally assumed that if a harvest was good,
the need for seed distribution was low; if a harvest was poor, the need for seed
increased.
Mozambique
Emergency seed inputs have been provided to farmers in Mozambique almost every
season for over a decade (Rohrbach et al., 2001), initially in response to the civil war
(
1976–1992) and, more recently, in response to both drought and floods.
Responsibility for regular data collection at national level is held by the Early Warning
Unit of the Ministry of Agriculture and Rural Development. There have been changes
3
to the monitoring system over the years, and both the methods and organisational
structures relating to the national monitoring system are currently under revision.
Despite changes taking place at national level, there exist clearly defined
assessment procedures at district level. The following description is based on
information collected by the authors through interviews with district agricultural staff
and farmers in 11 districts (see Saide et al., 2002). Once an emergency situation has
been reported within a district, it is taken for granted that seed will be needed and the
district agricultural officer is expected to recommend the crops and quantities of seed
required. Quantities of seed required are determined by the number of farming
households affected by the disaster. Information is compiled by local community
leaders, based on existing household lists, and usually verified by district field staff.
Existing data tables detailing the proportional areas planted with different crops in the
district indicate the crops for which seed is required. Information from the data tables,
together with average farm size and standard crop seeding rates are used to calculate
the quantity of seed needed. Examples of these procedures are provided below.
Critique
From the brief descriptions above, a number of similarities and differences in the
existing needs assessment techniques can be distinguished. In both countries, existing
monitoring systems manage a wealth of data of direct relevance to the assessment of
seed security, but at present this information is not immediately accessible to those
responsible for assessing whether or not relief seed is required.
In Somalia, the need for food aid is the key variable in determining the need
for seed among agricultural populations; although FSAU regularly collects a
considerable amount of information that usefully relates to seed, there is no explicit
assessment of seed need (Buzz Sharp, pers. comm.). In Mozambique, there is a clear
procedure for determining seed needs at district level, in which the key variable is the
number of farming households affected by the crisis. Both in Somalia and
Mozambique, the assessment exercises start from the assumption that seed will be
needed, and the data collected are merely used to calculate the quantities of seed
required and how these should be targeted. The possibilities that farmers either already
have seed of their own or can acquire seed locally and therefore do not need seed aid,
or that other types of external support might be more appropriate, are simply not
considered under present assessment processes.

346
C. Longley, C. Dominguez, M.A. Saide and W.J. Leonardo
Problems of knowing which crops and households are affected
The use of crop data tables in Mozambique — at least in theory — assumes that all
crops are equally affected by disaster (DDADR-Moamba, 2000; DDADR-Chibuto,
2
001). In Matutuine District, for example, once the number of affected households is
known, the cultivated area affected by the disaster is calculated according to an average
household farm size of one hectare. Using the crop data tables, the areas of the
different crops that have supposedly been affected are then determined. Cropping
systems differ, however, for the upland and lowland areas. Thus, the crops affected
will largely depend on the type of disaster: for example, floods affect the lowland
crops, whereas drought primarily affects the upland crops.
4
In Machaze District, Mozambique a meeting with community leaders
responsible for determining the number of households affected by the crisis revealed
that while it was relatively easy for them to know which households were affected by
flooding (namely, those situated in low-lying areas without access to upland farms),
this was considerably more difficult in a drought situation. The leaders were not able
to articulate the precise criteria used in these decisions, but it is thought that the poorer
households were deemed to be the most needy. When referring to seed security, the
issue of access is vital and must be made more explicit in seed needs assessment
exercises (Remington et al., this issue).
Issues of access, availability and timing
Although the issue of access is widely recognised as an important component of food
security, it has yet to be fully recognised as a component of seed security (see
Remington et al., this issue). In Somalia, although access to food is used by the FSAU
is determining food requirements, the ways in which farmers access seed did not
feature at all in the assessment of seed needs. Instead, the use of harvest data implies
that the immediate availability of seed was paramount in determining seed security.
However, this fails to take into account the ways in which farmers normally save and
acquire seed. The amount of seed saved by a household does not depend on the size of
the harvest but rather on the amount to be planted the following season. If, in a crisis
situation, the household is unable to save seed, there are mechanisms through which
seed can be acquired from local traders or other farmers.
The way in which seed needs have been assessed in Somalia in recent years
requires that crop harvest data are available before the quantity of seed aid can be
determined. Given that there are two main planting seasons, by the time the harvest
data are available, it is practically too late to procure seed for the forthcoming season.
This last-minute planning can be avoided by developing a greater understanding of the
ways in which farmers maintain and acquire seed at local level, thus allowing an
agency to predict (prior to harvest) whether or not external seed inputs are likely to be
required. The ways in which farmers manage seed and planting material is referred to
5
as the local or farmer seed system, and is here summarised in terms of an SSP.
Seed Systems Profile (SSP)
Field staff members often have some knowledge of local seed systems, but this must be
documented and used more explicitly in planning interventions. A failure to

A Methodology for Assessing Seed Systems 347
understand how local seed systems operate may lead to seeds-and-tools interventions
that inadvertently undermine these local systems over time. In the longer term, the
ability of farmers to respond to and cope with disasters depends partly on the resilience
of the local seed systems. The capacity of local seed systems must therefore be given
greater significance by relief and rehabilitation managers so that external assistance
effectively strengthens, not weakens local seed systems (Longley, 2000).
In order to develop a greater understanding of the ways in which farmers
maintain and access seed, it is suggested that an SSP be developed. This is best done
prior to an emergency, and can very feasibly form part of a disaster preparedness
strategy. The main questions to be answered in developing an SSP are shown in Box 1.
The questions in Box 1 deal with not only agro-ecological and technical
considerations but also socio-economic aspects of agricultural production and seed
management. As such, a seed system has been described as a socio-technical ensemble
(
Richards and Ruivenkamp, 1997). The socio-economic and agro-ecological aspects of
seed systems are elaborated further in Box 1.
Socio-economic aspects
How farmers use different crops and different varieties can provide important
information relating not only to the availability of seed but can also help to understand
the dynamics in cropping systems due to insecurity and changing market conditions
associated with disasters. For those crops that are sold (even in very small quantities),
planting material may be available to farmers in local markets, either as grain that can
Box 1 Questions to ask when developing a Seed Systems Profile
•
What crops and crop varieties are planted by farmers, and how are these used (for food,
for sale, as forage, etc.)?
•
What are the main features of the cropping system, i.e. in what ecologies are the crops
planted, what is the cropping calendar for the different crops and crop types and who
(
i.e. men or women) is responsible for various agricultural tasks?
•
•
For each crop, do farmers normally save the seed from the previous harvest? How is
i
seed saved and what are the main constraints?
If seed is not saved, how do farmers normally acquire the seed for the different crops
ii
iii
iv
(
where, through what means, from whom )?
Notes:
i. The term ‘seed saving’ covers more than just seed storage and includes seed production, seed
selection and processing. Seed production is generally regarded by farmers as a component of
agricultural production, i.e. seed multiplication takes place within farmers’ fields, not as a
separate activity. If farm size is exceptionally small, it may be difficult for farmers to produce
enough for both food and seed; theft may be a problem where there is population displacement
and inadequate means to secure fields and seed stores.
ii ‘Where’ might refer to whether seed is accessed within the settlement, from neighbouring
settlements or from local markets.
iii. The means through which farmers acquire seed might refer to whether seed is bought for cash,
exchanged (e.g. for labour, for grain or other seed), borrowed (to be repaid either with or
without interest at harvest time) or provided as a gift (generally through existing social networks
based on relations of reciprocity).
iv. Farmers might acquire seed from relatives or other farmers, from petty traders, or (rarely) from
formal sector sources (e.g. extension agents, seed suppliers or projects). It is important to know
whether those providing seed tend to be male or female.

348
C. Longley, C. Dominguez, M.A. Saide and W.J. Leonardo
be used as seed, or as seed/planting material. Farmers tend to take much greater care of
the seed of crops that are not normally sold, and ‘farmer seed specialists’ often exist
within communities. Such specialists are often able to maintain the seed of different
varieties of such crops, even under very extreme crisis conditions, thus providing a
source from which other farmers can acquire seed that they themselves may have lost.
Labour is often the main constraint to agricultural production, both in ‘normal’
situations and in times of stress, particularly in situations when some household
members are displaced, conscripted or killed. In situations of chronic conflict, labour
inputs to agricultural production are generally reduced and farmers often find it
difficult to control pests and diseases, thus reducing harvest outputs. By constructing
cropping calendars in relation to the main labour tasks and who is responsible for
specific agricultural activities (planting, weeding, seed selection and processing, seed
storage, seed acquisition), it may be possible to predict how seed systems will be
affected in relation to the timing of the crisis and the impacts it may have on local
populations, particularly in terms of changing household compositions. Knowing
whether men or women are responsible for specific tasks may also help to target
appropriate interventions according to gender roles. In many agricultural communities,
it is often women who are responsible for seed selection and storage.
It is necessary to answer the third and fourth questions of Box 1 as they relate
to different wealth groups. Three levels of household seed security have been defined
according to relative household wealth (Cromwell, 1996): seed-secure households are
usually better off and are able to save their own seed for planting but may use off-farm
sources (from relatives, neighbours and markets) in order to experiment with new
varieties; crisis-prone households are seed secure in most seasons but less well off, so a
situation such as harvest failure or death or illness in the family can force them to augment
their own saved seed with that acquired from markets and other, more seed-secure farmers;
chronically seed-insecure households are poorly resourced (small farm size, and a shortage
of labour resources) and so in most seasons are unable to harvest sufficient crops to meet
both seed and domestic consumption needs. Chronically seed-insecure households
regularly rely on acquiring seed through loans that are repaid after harvest, as gifts from
better-off farmers, or by exchanging their labour for seed. Farmers who are least seed
secure in ‘normal’ times also tend to suffer the most in terms of reduced access to seed in
times of crisis (Longley, 1998).
Agro-ecological aspects
In describing the crops and crop varieties and on what ecologies they are grown (e.g.,
lowland or upland fields or plots situated near the homestead), particular attention
should be given to the relative diversity of the cropping system and the range of
choices available to farmers. The resilience of both agricultural systems and seed
systems is at least partly determined by the ability of farmers to change their cropping
patterns, for example, by substituting crops or varieties according to local constraints
and opportunities, or by relying on alternative agro-ecologies in the situation that a
specific ecology or field type is rendered unproductive or unusable.
Predictions concerning the impact of a crisis on particular crops and seed
systems can be made by assessing the type, timing, duration and extent of a disaster in
relation to local crop calendars and cropping patterns. As mentioned above, a natural
disaster such as drought or flood will have greatest effect upon those crops growing in
the particular agro-ecologies affected. In the case of flooding, the planting material of

A Methodology for Assessing Seed Systems 349
Table 1 Seeding and multiplication rates of particular crops
Crop
Seeding rate (kg/ha)
Multiplication rate
Beans
100
8
6–10
15
Groundnut
Cowpea
Barley
120
90
100
15
Wheat
100
25
Rice
20 (upland), 80 (swamp)
50
Maize
20
10
5
100
100
200
Sorghum
Pearl millet
Source: ODI Seeds and Biodiversity Programme, 1996: 41.
some crops (e.g., cassava) can withstand prolonged waterlogging, depending on their
level of maturity when disaster strikes; other crops can be ‘rescued’ and re-planted on
higher ground, as reported by farmers in Matutuine District, Mozambique. In war
situations, the broad scale of crop and seed losses can be assessed in terms of the actual
time that farmers are prevented from farming and at what periods of the agricultural
calendar (Longley, 1998; Pottier with Wilding, 1994). If farmers are able to harvest
their crops, then at least some seed will be available, even when the output is very low.
A total lack of seed within a community is thought to be extremely rare, occurring
primarily following seasons in which farmers have not planted anything at all and
markets are not functioning. This happened in the vicinity of Baidoa, southern Somalia
in 1992–3 when there was widespread displacement; farmers were living in the bush
and it was not possible to farm for a period of two years (Longley et al., 2001).
The capacity of seed systems to recover following disaster can be remarkably
rapid, particularly if there is sufficient and timely rainfall to allow a good harvest.
Those crops with high multiplication rates and low seeding rates (that is, grain crops
such as millet and sorghum) have the potential to recover within a single season,
because the output is relatively high and the amount of seed required for the following
season is relatively small (see Table 1). Legumes, on the other hand, generally tend to
have lower multiplication rates and higher seeding rates, and it may take several
seasons for the seed system to recover fully. In situations of chronic conflict, farmers
are often able to continue to farm although they may experience difficulties in
maintaining or accessing the seed of legumes. Not only do legumes have lower
multiplication rates and higher seeding rates, the stored seed is generally more
susceptible to pests such as weevils. Where external support is to be repeated over
more than one season following an emergency, priority should be given to those crops
that farmers find most difficult to manage, for example beans, groundnut, cowpea.
The ability of farmers to save seed or planting material also depends on which
part of the crop is consumed. In the case of cassava or sweet potato, for example, it is
the roots and tubers that are eaten; the planting material comes from the stems and is
therefore unlikely to be consumed by the household in times of food shortage. The
same is true for vegetables such as pumpkin, cucumbers and tomatoes, in which the
seeds can be removed before consumption. In some cases, however, crops that might
normally be eaten when they are green (e.g., beans, peas, okra) must be maintained on
the farm until they are completely dry in order to produce viable seed.

350
C. Longley, C. Dominguez, M.A. Saide and W.J. Leonardo
Assessing the need for seed system support in a disaster
situation
An SSP can be used as a baseline to both predict and understand the impacts of disaster
on seed systems. Many of the variables necessary for assessing the need for seed
system support following an emergency situation have already been mentioned above,
namely the features of the crisis (type, timing, duration, scale and intensity), and the
socio-economic impacts on local populations (displacement, changes in household
composition). Other important aspects that also need to be considered are the
functioning of local markets, the mobility of both farmers and traders (in relation to
security and transport systems), and the assets available to farmers, including their
ability to draw on existing social networks.
The ways in which these different types of information can be used are
summarised in the framework presented in Box 2. The framework consists of five
steps that must be undertaken in order to understand: the characteristics of the disaster;
the impact on harvested output; seed availability; seed access; and the problems to be
addressed. For Steps 1–4 the first column of the framework highlights the basic
questions to be asked, and the second column identifies the sources of information that
can be used in answering the questions specified, using existing information,
information from the SSP and information that can be collected from the field. The
framework also provides an indication of the ways in which the information might be
interpreted, based on recent field experience in various different African contexts.
Box 2 Five-step framework for assessing seed systems affected by disaster7
Step 1 Characteristics of the disaster
Questions to be asked
Sources of information
What is the type of disaster, its relative Ongoing monitoring systems will describe
frequency over time, duration, extent, onset, the disaster and provide an indication of its
and timing?
geographical scale. Rapid assessments
What areas have been affected, both undertaken
by
individual
agencies,
geographical and agro-ecological? Are these coordinating bodies, and/or government
areas that have been affected by similar departments will report the direct impacts of
disasters in the past?
the disaster, the degree to which local
What are the direct impacts of the transport and market infrastructure are still
disaster, particularly on transport systems and functioning, and levels of population
market infrastructure?
displacement.
Additional micro-level
Has there been any large-scale population information may need to be collected from
displacement, either from or to areas of crop the field, where possible.
production?
Indications from Step 1: The situation can be defined as either a chronic or acute disaster.
In chronic disaster situations, it may be necessary to identify and address the underlying
causes of potential seed insecurity through longer term interventions (i.e. beyond emergency
seed relief). The SSP can usefully identify constraints within the seed system that might
need to be addressed.
Step 2 Impacts on harvested output
Questions to be asked
Sources of information
At what phase of the agricultural The SSP provides detailed crop calendars and
calendar has the disaster occurred?
information on which crops are planted in which
Which crops — planted at which agro-ecologies. This can be supplemented with
times and in which agro-ecologies — cropping data that might be available through on-
have been affected?
going monitoring systems or from existing baseline

A Methodology for Assessing Seed Systems 351
How severely have these crops surveys. Existing agricultural research reports will
been affected, i.e. what was/will be the also provide background information on cropping
level of harvested output? (Remember systems. Information on the extent to which crops
to take account of local coping have been affected by the disaster can be found from
strategies, e.g. harvested output of the ongoing monitoring systems. Where possible, this
same crops planted at different times information should be verified or supplemented by
and/or in different agro-ecologies.)
field visits.
Indications from Step 2: A situation is which there is no harvested output at all is very rare.
Different crops will be affected differently, and although the harvest for some crops may be
very small, it is likely that at least some farms will produce something. Farmers prefer to plant
seed from their own harvest than that from unknown sources.
Step 3 Seed availability
Questions to be asked
Sources of information
Will/have farmers been able to save seed or Basic agronomic knowledge of different crops,
planting material of the affected crops?
together with details from Steps 1 and 2 will
What are the seed system constraints help to determine whether farmers will have
and relative multiplication rates of the crops saved seed. The SSP provides information on
for which seed saving may be affected?
seed system constraints, ‘normal’ seed saving
For the seed affected, where and how and seed acquisition practices. The scale and
do farmers normally acquire seed of these geographical extent of the crisis, together with
crops, and is seed/planting material SSP information from neighbouring areas will
available from these sources?
help to determine the availability of seed in
nearby locations.
Indications from Step 3: If farmers have been able to harvest their crops, seed will generally
be available, even if the harvested output is very low. Where household stocks of own-saved
seed are limited, it is possible that seed or grain suitable for planting will be available within
the wider community, neighbouring communities and/or local markets.
Step 4 Seed access
Questions to be asked
Sources of information
Are the sources of available seed Other emergency assessments should indicate the
physically accessible? For example, are extent to which transport systems and markets are
markets functioning within reasonable functioning. Ongoing monitoring systems will
proximity, and is there transport that will indicate the levels of household asset loss and
allow farmers to visit markets or other vulnerability, and (where market information is
communities with seed/planting material? available) the relative price of grain and
If seed/planting material is available seed/planting material in local markets. Social
and physically accessible, do farmers in networks play an important role in allowing
the affected areas have the means farmers to access seed from other farmers —
(
financial or otherwise) to access this details of this are provided by the SSP.
seed?
Indications from Step 4: If seed is available, it is likely that at least some better-off farmers
will be able to access seed, but that poorer farmers may have problems accessing seed.
Breakdown of transport infrastructure and/or markets may make it difficult to transport seed
from surplus to deficit areas.
Step 5 Identifying problems to be addressed
The answers to the above four steps can be summarised in terms of problems to be addressed:
For each crop affected, is there a problem of seed and, if so, does this problem stems from lack
of availability or lack of access?
Prioritise the problems according to the crops for which seed systems have been most
severely affected.
Indicate which types of farmers in which areas are likely to have problems of seed
availability and/or access.
Prioritise the problems according to whether they need to be addressed in the short term
(
i.e. the forthcoming planting season) or the long term.

352
C. Longley, C. Dominguez, M.A. Saide and W.J. Leonardo
The framework is based on the assumption that considerable information will
already be available, either from existing early warning information and monitoring
systems or from other more general assessments that will likely have been undertaken
by operating agencies. Where possible, it is also necessary to supplement or verify
existing information with field visits to observe farms and markets and talk to both
farmers and traders. An additional source of relevant background information rarely
used by relief and rehabilitation managers is the findings of agricultural researchers.
For most countries, there is often a large body of literature (both published and
unpublished) dating back to colonial times describing local agricultural systems,
agricultural constraints and efforts made to overcome these constraints. Even in
countries affected by prolonged conflict where government offices and agricultural
research stations housing such reports have long been destroyed, the former
agricultural researchers themselves can often be traced either within or outside the
6
country, and it is often possible to locate extensive collections of documents. In the
case of Afghanistan, US researchers have been able to compile a well-informed
strategy for emergency seed supply based solely on existing documents, survey
information, and key informants (Dennis et al., 2002). In protracted emergencies, there
is a need for researchers to compile and present relevant agricultural research
information in a format that is useful to relief and rehabilitation managers.
Once the problems to be addressed have been identified, appropriate
interventions can be designed that effectively build on the strengths and address the
weaknesses of local seed systems. With a clearer understanding of the problem, the
objective of the intervention will then be more clearly understood, thus allowing for
better planning and targeting.
Suggestions for practical applications
The SSP, together with the steps outlined in Box 2, can be used to identify the strengths
and weaknesses of seed systems affected by disaster and highlight appropriate
interventions. Ideally, the SSP should have been developed prior to a disaster, but can
also be developed during the course of a protracted emergency or following an acute
disaster, provided that it is possible and appropriate to gather information from farmers.
The framework presented in Box 2 should be implemented following a disaster, or as a
first step in planning for seed aid and/or other agricultural interventions.
Given that many of the types of information necessary for an SSP and
assessment of seed system support are already being collected by existing monitoring
and early warning systems, it would not take much additional effort to implement the
methods described in this article. SSPs can be undertaken at a local level by individual
agencies in the course of baseline studies, or can be incorporated as a module into
national-level monitoring systems. Once an SSP has been developed for a given
administrative area or even the Food Economy Areas defined by the Household Food
Economy Approach, an assessment can be undertaken by local government authorities
or implementing agencies in the aftermath of a disaster to determine the ways in which
local seed systems will have been affected.

A Methodology for Assessing Seed Systems 353
Closing remarks
There is an urgent need for donors and operational agencies to move beyond
emergency seed supply to provide alternative forms of assistance that more accurately
reflect the needs of farmers affected by disaster. The development and use of suitable
assessment tools is just one step in this direction. The methodological approach
described in this paper is presently being tested on a pilot scale and further refined by
8
ICRISAT and ODI in Mozambique (ICRISAT-Mozambique, 2002), where
conventional seeds-and-tools interventions have been implemented for many years and
some agencies are now beginning to recognise the need for change. Even if the
assessment approach described in this paper proves to be applicable at a purely
technical level, its effective application in disaster situations and associated changes to
conventional practice depend largely on the political will of the donors, coordinating
bodies, implementing agencies and government departments concerned with
agricultural relief, rehabilitation and development.
Acknowledgements
The inputs of Tom Remington, Richard Jones, and Paula Bramel are fully
acknowledged in developing many of the ideas presented in the article. Thanks also to
Richard Jones and Mohamed Hussein Ahmed (‘Sufi’), both team members of the
Somalia seed sector study, commissioned and funded by a grant under the European
Community Development Cooperation. Comments on the paper were gratefully
received from Louise Sperling (CIAT), Rob Tripp (ODI), Elizabeth Cromwell (ODI),
Jo Macrae (ODI), Helen Young (Tufts University and ODI), Alison MacColl (FAO-
Somalia), Buzz Sharp (FSAU-Somalia), Nisar Majid (formerly FSAU-Somalia), Jean-
François Detry (formerly CIS-Mozambique) and John Seaman (SCF-UK). Funding for
the development and testing of the methodology in Mozambique was provided by the
USAID mission in Mozambique. The authors are particularly grateful to the staff of
DDADR and NGOs in Massinga, Panda and Chibuto districts who have been involved
in testing the methodology in Mozambique.
Notes
1.
2.
3.
The framework presented in Box 2 is not considered to be definitive. The seed security
assessment framework presented by Remington et al. (this volume) offers a slightly different
perspective in that it usefully emphasises aspects of seed and varietal quality. We gratefully
acknowledge their collaboration.
The Food Security Assessment Unit was initially established under the auspices of the World
Food Programme and is currently managed by the FAO, funded by the European
Commission and USAID-Somalia. FSAU works closely with FEWS-Somalia and WFP-
Somalia, among other partner agencies.
The Early Warning Unit was originally established with FAO assistance in 1989. In 1992,
the Food Security Consolidated Information System (CIS) was established by Médecins Sans
Frontières to provide timely analysis on household vulnerability and community needs
(
AEDES, 1996). CIS continued to operate with technical support from the Agence
Européene pour le Développement et la Santé (AEDES) until 2001. FEWS NET is also
currently operational in Mozambique.

354
C. Longley, C. Dominguez, M.A. Saide and W.J. Leonardo
4
5
.
.
The authors are grateful to Emidio Felipe Paulino of Concern Worldwide for organising the
meeting with leaders in Chipopopo and allowing the ICRISAT team to participate.
The farmer seed system is also sometimes called the informal seed system (Dominguez,
2
001) to distinguish it from the formal seed system. The latter includes the development,
multiplication and dissemination of new seed varieties by plant breeders, projects and seed
companies in both the private and public sectors. In times of crisis, the formal seed system
generally ceases to operate effectively.
6
.
For southern Sudan, there was such a wealth of information that ICRISAT contracted a
Research Assistant to visit UK libraries housing archival collections and the offices of
Norwegian Church Aid in Norway to review the reports available (Slaymaker, 2001a,
2
001b). Several former Sudanese researchers were found working for local agricultural
authorities and NGOs in southern Sudan, and some former British colonial agricultural
researchers were traced through the Tropical Agricultural Association.
7
8
. This framework draws on the work of Tom Remington (see Remington et al., this volume)
whose inputs and collaboration are gratefully acknowledged.
.
Guidelines for developing the SSP, assessing seed systems in disaster situations and develop-
ing appropriate interventions are available in Portuguese and English from Carlos
Dominguez (C.Dominguez@cgiar.org).
References
AEDES (1998) Food Security Consolidated Information System in Mozambique. Ministry of
Agriculture and Rural Development, Maputo.
Chemonics Intl. and USAID (1996) Seeds for Disaster Mitigation and Recovery in the Greater
Horn of Africa. Report submitted by DEFSIL Project of USAID/G/ENV/ENR to
USAID/BHR/OFDA/PMPP, Washington.
Cromwell, E. (1996) Governments, Seeds and Farmers in a Changing Africa. CAB Intl. and
ODI, Wallingford.
Dennis, J., A. Diab and P. Trutmann (2002) The Planning of Emergency Seed Supply for
Afghanistan in 2002 and Beyond. A draft paper for the Tashkent conference. Available at
www.afghanseeds.org.
Direcção Distrital de Agricultura e Desenvolvimento Rural (DDADR)-Chibuto (2001) Plano
face a tragédia. DDADR, Chibuto.
Direcção Distrital de Agricultura e Desenvolvimento Rural (DDADR)-Moamba (2000) Informe
sobre o início da campanha agrícola 2000/2001. DDADR, Outubro, Moamba.
Dominguez, C. (2001) Sistema informal de sementes, causas, consequências e alternativas.
Centro Cultural de Matalana, Maputo.
Friis-Hansen, E. and D. Rohrbach (1993) SADC/ICRISAT 1992 Drought Relief Emergency
Production of Sorghum and Pearl Millet Seed: Impact Assessment. Working Paper 93/01,
ICRISAT, Hyderabad.
ICRISAT (2002) Guia para planificar intervenções no sistema local de sementes. ICRISAT-
Mozambique/INIA, Maputo.
Jaspars, S. and H. Young (1995) General Food Distribution in Emergencies: From Nutritional
Needs to Political Priorities. Good Practice Review, No. 3, RRN, Overseas Development
Institute, London.
Johnson, D. (1998) Distributing Seeds and Tools in Emergencies. Oxfam, Oxford.
Jones, R., C. Longley, T. Remington and P. Bramel (2001) Seed Security, Seed Relief and
Alternative Interventions in Southern Sudan: Preliminary Findings from the Project Seeds for
Survival: Increasing the Effectiveness of Emergency Seed Security in the Greater Horn of
Africa. ICRISAT/ ODI/CRS, Nairobi.
—
— , P. Bramel, C. Longley and T. Remington (2002) The Need to Look Beyond the
Production and Provision of Relief Seed: Experiences from Southern Sudan. Disasters
2
6(4): 302–15.

A Methodology for Assessing Seed Systems 355
Longley, C. (1997) Effects of War and Displacement on Local Seed Systems in Sierra Leone. In
L. Sperling (ed.) Network Paper No. 75, AgREN, Overseas Development Institute, London.
—
— (1998) Seed Management Strategies of Farmers Affected by War: A Case Study from
Northern Sierra Leone. Report to ESCOR, DFID, London.
—
— (2001) Beyond Seeds and Tools: Opportunities and Challenges for Alternative
Interventions in Protracted Emergencies. Humanitarian Exchange No. 18, HPN, Overseas
Development Institute, London.
—
— and P. Richards (1999) Farmer Seed Systems and Disaster. In Restoring Farmers’ Seed
Systems in Disaster Situations. Plant Production and Protection Paper 150, FAO, Rome.
— , R. Jones, M. H. Ahmed and P. Audi (2001) Supporting Local Seed Systems in Southern
Somalia: A Developmental Approach to Agricultural Rehabilitation in Emergency
Situations. Network Paper No. 115, AgREN, Overseas Development Institute, London.
—
ODI Seeds and Biodiversity Programme (1996) Seed Provision During and After Emergencies.
Good Practice Review 4, RRN, Overseas Development Institute, London.
Pottier, J. with J. Wilding (1994) Food Security and Agricultural Rehabilitation in Post-war
Rwanda, August–September. Report to SCF-UK, London.
Remington, T., J. Maroko, S. Walsh, P. Omanga and E. Charles (2002) Getting Off the Seeds-
and-tools Treadmill with CRS Seed Vouchers and Fairs. Disasters 26(4): 316–28.
Richards, P. and G. Ruivenkamp, with contributions from R. van der Drift, M. Gonowolo, M.
Jusu, C. Longley and S. McGuire (1997) Seeds and Survival: Crop Genetic Resources in
War and Reconstruction in Africa. International Plant Genetic Resources Institute, Rome.
Rohrbach, D.D., J. Low, R. Pitori, A. Cucu, J. Massingue, D. Boughton, G. Rafael, A. Paulo and
D. Jocene (2001) Investment Priorities for the Development of Mozambique’s Seed System.
Directorate of Economics Research Report No. 44E. Ministry of Agriculture and Rural
Development, Maputo.
Saide, M., C. Dominguez and W. Leonardo (2002) Procedimentos utilizados na avaliação das
necessidades em sementes em situaçőes de emergência. INIA/ICRISAT, Maputo.
Seaman, J, P. Clarke, T. Boudreau and J. Holt (2000) The Household Economy Approach: A
Resource Manual for Practitioners. SCF UK, London.
Slaymaker, T. (2001a) Literature Review of Agricultural Research Conducted in Southern Sudan
Post-1945. Consultancy report for ICRISAT/CRS/ODI seed study, Nairobi.
—
— (2001b) Review of Norwegian Literature on Agricultural research in Southern Sudan.
Consultancy report for ICRISAT/CRS/ODI seed study, Nairobi.
Sperling, L. (ed.) (1997) War and Crop Diversity. Network Paper No. 75, AgREN, Overseas
Development Institute, London.
Address for correspondence: Catherine Longley, Overseas Development
Institute, 111 Westminster Bridge Road, London
SE1 7JD.
E-mail :
<
<k.longley@odi.org.uk>>