f ,. "
Strategies for Strengthening and
Scaling up Community-based
Peter S. Setimela and Petr Kosina
CIMMYT (www.cimmyt.org) is an international, not-for-profit organization that conducts
research and training related to maize and wheat throughout the developing world. Drawing
on strong science and effective partnerships, CIMMYT works to create, share, and use
knowledge and technology to increase food security, improve the productivity and profitability
of farming systems, and sustain natural resources. CIMMYT is one of 15 Future Harvest
Centers of the Consultative Group on International Agricultural Research (CGIAR) (www.
cgiar.org). Financial support for CIMMYT's work comes from the members of the CGIAR,
national governments, foundations, development banks, and other public and private agencies.
International Maize and Wheat Improvement Center (CIMMYT) 2006. All rights reserved.
The designations employed in the presentation of materials in this publication do not imply the
expression of any opinion whatsoever on the part of CIMMYT or its contributory organizations
concerning the legal status of any country, territory, city, or area, or of its authorities, or
concerning the delimitation of its frontiers or boundaries. CIMMYT encourages fair use of this
material. Proper citation is requested.
Correct citation: Setimela, P.S., and P. Kosina. (eds). 2006. Strategies for Strengthening andScaling up
Community-based Seed Production. Mexico, D.F.: CIMMYT.
AGROVOC descriptors: Maize; Agricultural economics; Seed production; Plant breeding;
Pollination; Hybrids; Planting date; Models; Small farms; Training
programmes; International cooperation; South Africa
Additional Keywords: CIMMYT; FAO
AGRIS category codes: E16 Production Economics
F03 Seed Production
Dewey decimal classification: 338.1768
Design and Layout: Marcelo Ortiz S.
Printed in Mexico.
Maize Hybrid and Open Pollinated Varieties: Seed Production Strategies
1 Maize hybrids
1 Hybrid seed production
4 Characteristics of hybrid plants
4 Advantage and disadvantages of growing hybrids
5 Open-pollinated varieties (OPVs)
5 Stages of seed production
6 Standards for certified seed
7 OPV seed production
8 How to harvest, dry, shell, and store seed
Analysis of Community-Based Maize Seed Production and Supply Schemes in
10 Purpose of community-based seed production and supply schemes
11 Elements of sustainable CBSPS
12 Overcoming the problem of isolation distance within communities
12 Accessing foundation seed
13 Selecting a variety for community-based seed production
13 Roles of various partners in the setting up CBSPS
13 Recommended reading
A Case Study of a Community-Based Seed Production Scheme in two Districts
of the Limpopo Province, South Africa
14 Farmer variety selection
16 Provision of basic seed
18 Seed Growers Associations
18 Seed Growers Service Cooperative
19 Collection Processing Point (CPP)
Scaling Up Strategies in Reference to Agroforestry
21 Prong 1. Direct training of farmer trainers and local change agent teams
22 Prong 2. Training of partner staff
22 Prong 3. Farmer-to-farmer exchange
23 Prong 4. Support to existing initiatives
24 Additional Prongs
25 Recommended reading
CIMMYT would like to thank the following organizations for their support and contributions:
* Department of Agricultural Research and Extension (AREX) within the Ministry of Lands, Agriculture,
and Rural Resettlement in Zimbabwe
* Non-governmental organizations (NGOs) involved in Agriculture in Zimbabwe
* Department of International Development (DFID)-Zimbabwe and FAO for financial support.
Maize is the primary source of calories in the diet of many resource-poor farm families in the
Southern African Development Community (SADC). International agricultural research centers,
in collaboration with national agricultural research systems, continue to develop and disseminate
technologies for adoption by smallholder farmers. These technologies are aimed at addressing
the multiple constraints that smallholder farmers face. One of these new technologies is stress
tolerant, more nutritious open-pollinated maize varieties (OPVs) and hybrids.
Although many of these new varieties have been released by the national systems, smallholder
farmers continue to encounter difficulties in accessing maize seed. To overcome this problem
some non-governmental organizations (NGOs) have initiated community-based seed production
schemes to assist farmers with selecting, multiplying and saving seed of the improved OPVs to
meet their own needs, and sometimes those of neighboring communities.
Community-based seed production schemes are most commonly initiated where the formal seed
system is unreliable and the seed supply infrastructure is poor. However, NGOs and extension
workers training farmers on seed production require sound knowledge of the differences between
maize OPVs and hybrids, variety characteristics, strategies for community based seed production,
and scaling up methodologies. It is to fill this need that this publication was compiled.
CIMMYT hopes that this publication will become a useful resource for development workers
involved in community-based seed production schemes and related facets of smallholder
agriculture in the region.
Director, CIMMYT Global Maize Program
Maize Hybrids and Open-Pollinated
Varieties: Seed Production Strategies
P. S. Setimela1, X. Mhike2, J. F. MacRobert1, and D. Muungani2
1CIMMYT -Zimbabwe, P.O. Box MP 163, Harare, Zimbabwe. E-mail: email@example.com
2Department of Agricultural Research and Extension (AREX), Ministry of Lands, Agriculture, and Rural
Resettlement, P.O. Box CY550, Causeway, Harare, Zimbabwe
Maize varieties may be either hybrid or open-pollinated. Hybrid varieties are made by crossing
selected parents (sometimes known as inbred lines) in the field, while open-pollinated varieties
(OPVs) are broad populations of many parents. Open-pollinated varieties show greater variability
than hybrids, but have the advantage that unlike hybrids, their seed may be saved for re-planting
without much yield loss. These differences are explained in greater detail below.
Hybrid plants are produced by cross-pollinating two unrelated male and female parents. Crosses
made between two unrelated inbred parents are known as single-cross hybrids (Figure 1). Those
from three parents are three-way hybrids (Figure 2), and those from four parents as double-cross
hybrids (Figure 3). The female of a three-way hybrid is a single cross hybrid, while the male is an
inbred line. The parents of a double cross hybrid are both single-cross hybrids. Another type of hybrid
is the top-cross, in which one of the parents is an OPV, and the other is a single-cross or an inbred line.
Plant breeders carefully select the parents of hybrids over many years of testing and evaluation.
Hybrid parents are chosen based on characteristics such as early maturity, disease resistance,
drought tolerance and yield potential. The resultant hybrids are evaluated for yield potential and
field characteristics appropriate to farmers. Only the best hybrids are advanced to commercial seed
production and sale. In order to ensure consistent uniformity and performance of the hybrid, the
parents are maintained in a very controlled manner, and the production of hybrid seed is done
following strict procedures.
Hybrid seed production
Hybrid seed is produced in a controlled fashion far from other maize fields, in order to avoid
contamination. Male and female parents are inter-planted in alternating rows (Figure 4), usually
composed of 3 to 6 female rows and 1 or 2 male rows. The female plants are de-tasselled (i.e. their
male inflorescence is removed) before a single tassel starts to shed pollen, and pollination of the
female is achieved with pollen from the male plants.
Female Parent x Male parent
(obtained from female)
Figure 1. The formation of a single-cross hybrid
involves crossing a male parent with a de-
tasseled female parent.
Female Parent A x Male parent B
Female Parent AB x Male parent C
(obtained from female)
Figure 2.The formation three-way cross
hybrids involves crossing a single-cross hybrid
(de-tasseled female parent) with a different
inbred male parent.
Female Parent A x Male parent B
Female Parent AB
Female Parent C x Male parent D
x Male parent CD
Hybrid Seed ABCD
(obtained from female)
Figure 3. The formation of a double-cross hybrid involves crossing two single-cross hybrids.
Field inspection is done regularly to ensure that emerging tassels on the females are all removed,
and that there is no contamination by pollen from the female rows or surrounding fields (see
isolation distance below). To avoid seed mixing during harvesting the male parent rows are cut
down once they have provided pollen to the female parents.
A key factor in both yield and quality of hybrid seed production is the coincidence of flowering of the
male and female parent plants, known as nicking. If male pollen flow begins after the emergence of
the female silks or ends before the female silks are fully emerged, not only will less seed be set, but
the risk of contamination by foreign pollen increases.
In most countries the isolation distance for hybrid seed production is about 400 meters. This is
the distance that has to be maintained all around the hybrid seed production unit. Four hundred
meters is greater than the normal dispersal range of maize pollen, so the isolation ensures that
pollen from neighboring fields does not reach the hybrid seed production site.
Both parents of single-cross hybrids (inbred lines) have low vigor and low seed yield, and
therefore need excellent crop management for good seed production. Three-way and double-
cross hybrids have as one of their parents, a single-cross hybrid, which is vigorous and high
yielding, but they, too, require good crop management.
Figure 4. Hybrid seed production field showing alternating two male rows with tassels
and six female rows without tassels.
Characteristics of hybrid plants
Hybrid plants are uniform in colour, maturity, plant
height and other plant characteristics (Figure 5).
The uniformity of hybrid plants enables farmers to
carry out certain operations, such as harvest, at the
same time, which is useful for those farmers using
combine harvesters. Hybrids are usually higher- -
yielding than OPVs, but they generally require
much higher standards of field management than
OPVs to achieve their yield potential. The higher
yield of hybrids is the result of a biological effect
known as heterosis, which describes the superiority
of the first generation over the parents.
However, hybrid plants will not reproduce uniform
plants similar to themselves if the hybrid grain is
recycled as seed in the following season. Furthermore,
the yield following recycling will decrease by 30-50%,
depending on the type of hybrid originally grown,
due to an effect known as segregation.
Figure 5. Hybrid plants, especially single-
Single-cross hybrids suffer much greater yield r hybrid plant, e ecia singe-
cross hybrids, are uniform in plant height
reductions from recycling, than do three-way or and plant characteristics.
double-cross hybrids. For best results, therefore, fresh
hybrid seed should be purchased every year.
Advantages and disadvantages of growing hybrids
Hybrids are higher-yielding than open-pollinated varieties (OPVs). Research has shown that the
yield advantage of hybrids over OPVs is about 15%, depending on the varieties being compared.
Some improved OPVs may yield as well as some poorly adapted hybrids. Therefore, farmers in
high potential environments who can afford inputs such as fertilizer, stand to benefit the most
from growing adapted and suitable hybrids (Table 1.1, Figure 6).
Table 1.1. Farmers' checklist on when to grow a hybrid or OPV.
When to grow a hybrid When to grow an OPV
* The farmer expects to harvest more than The farmer does not expect to harvest more than
2 tons/ha (15 bags per acre) of maize grain. 2 t/ha of maize grain.
* The costs of hybrid seed will be recovered from its yield The costs of hybrid seed may not be paid for by
advantage. Hybrid seed costs about 10 times the price its yield advantage over the OPV.
of grain, and therefore the yield advantage of the hybrid
should be at least 250 kg/ha.
* The farmer is located in a high potential environment and The farmer is located in a low potential environment
can afford inputs such as fertilizer and pesticides. and cannot afford extra inputs.
* Hybrids adapted/suitable for local conditions are available. No locally adapted/suitable hybrids are available.
* The farmer can obtain seed for timely planting. The farmer cannot readily obtain seed.
Open-pollinated varieties (OPVs)
Open pollinated varieties are known sources of open pollinated plants. Some have been improved
and have variety names. The grain of an OPV may be saved for replanting if it is isolated from
other maize varieties, or harvested from the middle of the field. The recycled seed will grow and
yield as well as the original plants. Compared to hybrids, OPVs are less uniform (Figure 7) and
usually lower-yielding than hybrids in optimal environments. OPVs, however, have the advantage of
being more stable than hybrids in low-yielding or stress environments.
Stages of seed production
For both hybrid and OPV seed production there is a three-stage process of seed multiplication, as
follows (see also Figure 8).
The first stage, the production of breeder's (or pre-basic) seed, is carried out by the breeder.
This stage produces the least amount of seed, with the highest level of varietal purity.
Breeder's seed is produced in small plots where a breeder can monitor the plants to ensure
that there is no pollen contamination, and that the plants are 'true to type'. The breeder or the
developer of the variety is responsible for maintaining breeder's seed, i.e. ensuring that its
genetic purity is maintained.
Figure 6. How to grow hybrids.
Figure 7. Maize OPVs usually exhibit plant variability.
Foundation or basic seed is the first increase of the breeder's seed. The seed is produced in
isolation and with great care to ensure the variety remains true to type. The foundation seed is
produced by breeders or by seed companies. The seed companies will sell foundation seed or
give it to selected farmers to produce certified seed under contract.
Certified seed is the last stage in seed multiplication. Seed companies contract farmers to
produce certified seed. The seed is produced from foundation seed and grown in isolation and
under controlled standards to ensure purity and identity. This seed is processed and packaged
for sale to farmers to produce grain.
The quantity of seed produced increases from one stage to the next. Only small quantities of
breeder's seed are produced, while large volumes of certified seed have to be produced to meet
the demand for the grain production (Figure 8).
Standards for certified seed
Before seed is certified, the authorizing agency must check that the seed meets all the set
quality and purity standards. The agency also provides labels and makes spot checks at points
of sale. Physical seed quality tests include laboratory testing for moisture content, seed purity,
germination, viability, and absence of pests and diseases. This ensures that the seed is viable and
healthy. Seed certification for maize is mandatory in most countries. Some countries may allow
OPV seed to be sold as 'truthfully labelled' or standard seed. Truthfully labelled seed meets
minimum standards for purity, and its producer guarantees the quality as described on the label.
* Half the seed is saved as reserve stock
Figure 8. Stages of seed production.
In order to protect farmers from poor quality seed, the government seed certification agency
verifies the source of the seed, through field inspections (of foundation and certified seed
production fields) at specified stages of the growing season and after harvest. Field inspectors
check that the seed is planted according to the quality standards required, isolation distances are
met, the fields are free of weeds, and all the 'off type' plants are removed from the field. If the
field inspector finds that some of the requirements are not met, he/she may recommend that the
crop is destroyed, or that the harvest is used only as grain, or that certain field operations are
carried out immediately (rules vary from country to country).
OPV seed production
Farmers who are far from seed retails shops can benefit from growing maize OPVs, as the grain
they harvest may be replanted as seed without significant yield loss. OPV seed may be recycled
for a maximum of three seasons without significant yield loss. The longer a farmer recycles OPV
seed, the greater the risk of contamination by pollen from neighboring fields, and deterioration
of the variety.
Seed production of OPVs is simpler than that of hybrids, because only one variety is planted and
there is no need for detasseling. Nevertheless, maintenance of the OPV and production of the
foundation seed of OPVs requires some special procedures, similar to hybrid seed production.
The OPV seed production fields may be isolated in space or time, or both. The isolation distance
for OPV seed production is normally 250-350 m. Isolation by time requires a planting interval
(difference in sowing time) of 4 to 6 weeks to ensure that there is no pollen contamination of seed
plots by other maize fields. The number of days to tasseling and silking of the seed field and
neighboring fields helps to determine the proper time isolation (Figure 9).
If farmers cannot isolate by time or distance they may harvest from the middle of their fields to
minimize contamination, and the field should be at least about one hectare.
Maize field (OPV)
Planted first Planted 30 days or
,,...--, \more later "
a) by time b) by space Bush fallow/other Crop
Figure 9. Isolation by time and space of an OPV maize seed field.
How to harvest, dry, shell, and store seed
At least 500 cobs should be harvested to ensure that the OPV maintains its own distinct
variability. It is recommended that farmers harvest their ears at a grain moisture content of
about 16-20 % (i.e. once the husks are dry) and sun-dry to about 14 % moisture content before
shelling. The cobs are husked, selected for uniformity and quality (i.e. pest- and disease-free, not
discoloured) and spread out in the sun for drying. In most areas, farmers may sun-dry their seed
to the desired moisture content. When sun-drying, ensure that the seeds are turned frequently to
prevent "sun-burn" damage to the embryo. Most farmers sun-dry their maize seed on a concrete
slab, but if this is not available, the seed may be spread out on black plastic sheeting.
Shelling may begin once the grain has reached a moisture content of 14%. Farmers can apply
a simple test to check whether the seed has reached 14% moisture content: Shell a cob and mix
about 100 kernels with one gram of salt; if the salt feels moist after 5 minutes then the cobs need
to be dried further, but if salt remains dry then the shelling can begin.
Shelling by hand minimizes seed damage, but if a mechanical sheller is used, lower the settings
(speed and severity of threshing action). The shelled seed may then be cleaned (winnowed), and
any chipped or diseased seeds removed by hand.
The seed is then treated with insecticides and fungicides against storage pests and fungi. When
treating with pesticides, observe safety recommendations, and ensure that the seed is uniformly
treated. Treated seed must never be used as food for animals or humans. The seed is then stored
in bags or in bulk a cool, dry place, away from direct sunlight. Ideally, seed should be stored at
12% moisture; low seed moisture increases the viability and storability of the seed.
RECYCLING OPV SEED
Certified OPV seed Farmers' field production Grain for food,
feed and sale
Shell, dry, and then store the seed Select good cobs for seed from
under appropriate conditions. the middle of your field.
YOU MAY RECYCLE AN OPV FOR 1-3 SEASONS BEFORE BUYING NEW CERTIFIED SEED.
Figure 10. How to grow OPVs and retain your own seed.
Analysis of Community-Based Maize
Seed Production and Supply Schemes in
1CIMMYT -Zimbabwe, P.O. Box MP 163, Harare, Zimbabwe. Email: firstname.lastname@example.org
Farmer recognize seed as one of the most important inputs in agricultural production. Seed
carrying traits such as earliness, disease resistance and increased yield potential can improve
productivity as well as the value of the commodity on the market.
Regardless of the crop, most of the seed in sub-Saharan African comes from the informal seed
sector, namely farmers' own saved seeds, farmer-to-farmer seed exchange, and local seed
markets. This sector supplies mainly local landraces, local and improved varieties, and recycled
hybrids. The seed markets in the informal sector are unstructured and usually lack marketing
strategies. The seed quality is assured primarily through existing trust between seed buyers and
sellers. Sometimes the seed provided through these channels may be of poor quality (low genetic
purity, contaminated with pests and diseases, or poor germination).
To improve seed availability and quality, government institutions and non-governmental
organizations (NGOs) and have initiated community-based seed production and supply
schemes (CBSPS) with the aim of improving the availability of quality seeds of improved
varieties. The CBSPS are informal in the sense that seed quality standards required in seed
production are not as stringently enforced as in the formal seed sector. Many CBSPS have
been effective in disseminating improved varieties; others have been unsustainable, and this is
attributed to various factors, including the following:
* Farmers and NGOs staff have inadequate knowledge and skills in seed production, quality-
enhancing measures and seed systems.
* Advice on seed issues and extension, mainly provided by government seed service units,
does not reach small-scale farmers due to poor rural infrastructure, and sometimes, too few
* Some plant breeders and researchers do not make an effort to promote their varieties, so they
remain unknown to farming communities.
* The seed demand within the community is difficult to estimate.The formal seed sector is
unreliable for crops that it considers unprofitable such as mungbeans; cowpea; minor millets; and
even maize OPVs, on which private seed companies do not hold exclusive marketing rights.
* Issues pertaining to sustainability are not well thought through in the establishment of
* Poor linkages of CBSPS to agricultural research, seed market information, seed companies, and
extension have resulted in limited supply of foundation seed.
* Many of the new improved varieties do not meet farmers' needs, and so farmers do not
* Seed policy pertaining to community-based seed production is not well understood, or it
does not exist.
* Due to the small land holdings, isolation distance requirements for quality seed production are
problematic, if not impossible to achieve in some community areas.
Strengthening the capacities of partner organizations involved and interested in CBSPS is key to
empowering farmers in availing and accessing quality seed of improved varieties.
Most of the seed policies in sub-Saharan Africa are aimed at the formal seed sector, and the
informal sector, including CBSPS, has been long neglected. It is only recently that the importance
of the informal seed sector, which provides more than 85 % of the seed requirements of
resource-poor farmers in sub-Saharan Africa, is being appreciated (Monyo et al., 2004, Figure
11). It is therefore necessary that CBSPS are strengthened, as a means of empowering farming
communities, increasing their productivity and ultimately their incomes and food security.
Purpose of community-based seed production
and supply schemes
Setting up a CBSPS starts with understanding farmers' existing seed systems, including their
seed channels/suppliers, the type of crops they plant, and their major farming constraints. Before
setting up a CBSPS, its purpose should be well defined. This might be to increase farmers' access
to improved and preferred varieties in order to generate income, or to achieve better seed security.
Usually the income generation aspect is only thought of after a CBSPS is in place, but without the
business skills to manage the seed production/sale as a business, the scheme often falters.
Seed production starts with a few kilograms of 1"
a specific preferred variety, which a seed grower
multiplies and makes available to other farmers in
the community. The demand for seed of a preferred
variety within the community should be assessed,
so that quantities to be produced are estimated, to
prevent under- or over-supply of seed.
85% ] Informal
Fig. 11. Seed supply to resource-poor
smallholders in sub-Saharan Africa.
For CBSP, certain conditions must be met, e.g.
* Planting material must meet specific quality standards, to ensure variety attributes and the
quality of seed.
* The seed grower must know the variety and be able to recognize it; this minimizes
contamination from other varieties.
* Adequate crop protection and other quality/yield enhancing techniques should be used.
* The produced seed should pass germination/ quality tests.
As such, seed production is a unique (specific) business, and not any farmer can be a seed
grower. Seed growers are usually lead farmers in the community, who understand the technical
aspects of seed multiplication. They should also be socially accessible, so that other farmers can
readily access the produced seed.
Elements of sustainable CBSPS
Unfortunately, most of the CBSPS started in southern Africa have not been sustainable. Once
external support is terminated the schemes have collapsed and the farmers have reverted to
experiencing shortages of improved seed.
If a CBSPS succeeds and becomes efficient and sustainable, the farmers could become future
contract seed producers for seed companies, and continue meeting their communities' seed
requirements. Information on new improved OPVs, their characteristics, and prices should
be provided to farmers by the producers of the OPVs, and extension workers. This, and other
elements of sustainable CBSPS, are summarized in Box 1, and Figure 12.
Box 1. Elements of sustainable CBSPS
Continuous exposure of farmers to new promising and preferred varieties in a way
that stimulates the market. Smallholder farmers should be informed of the advantages of
buying fresh seed.
Strong linkages between national agricultural research and extension systems
(NARES), international agricultural research (including the CGIAR) centers, seed
companies, small-scale seed producers, and extension.
Ready supply of improved germplasm and relevant information.
Seed production models that suit the target communities.
Farmer empowerment with business skills and training on technical aspects of seed
Analysis of the entire production chain-from seed/grain production through to
consumers/buyers-to discover where interventions would be most beneficial.
Agricultural policies that favor and strengthen the informal seed sector.
Marketing and distribution strategies for small seed producers.
Promotion of alternative income sources e.g., a women's group in Uganda
produces OPV seed and also re-packages fertilizer for sale to smallholder farmers
who cannot afford to buy large quantities.
Fcormnl eed sysrlem) production scheme
Informal seed system In nme ngenerainon
Figure 12. Important linkages in community-based seed productions schemes.
Overcoming the problem of isolation distance within communities
In many smallholder communities in the region, it is difficult to find farmers with large
enough fields to meet the isolation distances needed to produce maize seed (see Chapter 1).
Several approaches could help overcome the problem of isolation distance in a community, as
summarized in Box 2.
Box 2. Overcoming the problem of isolation distance in a community
Farmers involved in seed production should be allowed to plant before other farmers (time
isolation). However, this is often difficult because of irregular rainfall in many areas.
Farmers in the immediate surrounding of the seed growers can be offered for free (or at a
lower price) seed of the same variety, to minimize contamination.
Seed production can be done during the off-season (time isolation) if irrigation is available.
Usually this is the best time to produce seed as there are only a few crops in the field, and
disease pressure is lower.
Accessing foundation seed
Seed growers within communities should be linked to NARS or other institutions such as seed
companies, which can provide them with foundation seed. Both the seed growers and their
support service providers (such as NGOs and extension services) should be aware of seed policies
governing CBSPS in their countries. In most countries, seed policy allows CBSPS to market and
sell seed only within the communities. If a CBSPS plans to markets and sell its seed farther afield,
it has to follow additional regulations and process, such as registration as a seed producer.
Private seed companies
* Market seed from CBSPS
* Foundation seed production
* Development of new varieties
* Credit system
* Community mobilization
* Community organization
* Quality control
* Advocacy for policy change
* Foundation seed production
* Development of new varieties
Selecting a variety for community-based seed production
Because of the above policy restrictions on the sale of seed, CBSPS can only multiply and sell
seed of crop varieties that are adapted to local farming conditions, and are known and preferred
by other farmers within their community or region. The process for variety selection for CBSP is
usually as follows:
* NARS issues promising varieties for testing by farmers in a community.
* Farmers' skills and knowledge in the areas of crop agronomy, and variety selection is
* Group of farmers, individually or by consensus rank the varieties of their choice.
Roles of various partners in the setting up CBSPS
The success of a CBSPS depends on the support and goodwill of multiple parties from the public
and private sectors.
* Government should develop agricultural/ economic policies that promote CBSPS.
* NARS should support communities with foundation seed of improved varieties.
* Extension and NGOs should train farmers and their communities in quality seed production,
entrepreneurial skills, and sustainable resource management; most farmers lack initial credit to
start CBSPS, and NGOs are well placed to source the initial credit to support farmers.
* The private sector should link more closely with the communities. As the communities acquire
more experience in seed production they can become contract seed growers for private seed
companies. Such linkages would also provides a sustainable market for the community to sell
* The international and national research communities should continue to develop and
promote new improved varieties that farmers want to grow, and information targeted at
extension and farmers.
Monyo, E.S., Mgonja, M.A., Rohrbach, D.D. (2004) New Partnership to strengthen seed systems in Southern
Africa: Innovative community/ commercial seed supply models. In Setimela P.S., Monyo E.S. and Banziger
M. (eds.). 2004. Successful community-based seed production strategies/Produgao de sementes de cultures
alimentares na regiao da SADC, CIMMYT. Mexico D.F.
Case Study of a Community-based Seed
Production Scheme in two Districts of the
Limpopo Province, South Africa
J.J. Mkhari1, M.R. Matlebjane1, K.P. Dlomu1, N.D. Mudau1 and K. Mashingaidze2
Depart. of Agriculture, Northern Province, P. Bag X9487, 69 Biccard Street, Polokwane,
South Africa. E-mail: Mkhariii2, '.; il I'L. a.. i prov.gov.za
2ARC Grain Crops Institute, Private Bag X1251, Potchefstroom, 2520, South Africa
The community-based seed production scheme in the Limpopo Province, South Africa was
initiated in 2000, with the aim of addressing the seed security of smallholder farmers. The project
focused on Vhembe and Capricorn districts, and was supported by Gesellschaft fir Technische
Zusammenarbeit (GTZ). Before the start of the project, smallholder farmers in the two districts
had been exposed to field trials aimed at addressing their specific environmental and socio-
Farmer variety selection
In 2002/2003 farmers in Mashushu community of Capricorn district selected ZM 421, an
open pollinated variety. The main reasons for selecting this variety were its comparative
high and stable yield, drought tolerance, and early maturity compared to other varieties in
variety evaluation verification and observation (VEVO) trials. Its early maturity was especially
attractive, because it eased the burden of guarding the crop from destruction by baboons.
Farmers in Vhembe district selected two OPVs, ZM 521 and Grace, in 2000/2001. ZM 521
was high yielding, matured early and had good milling properties. Grace was selected for its
suitability for making green mealies (boiled green maize). Both varieties were released in 2001.
Farmers were trained in producing the seed of their preferred varieties, for local seed security
South African National Seed Organization (SANSOR) has been involved in the seed certification
of ZM 521 since 2002. SANSOR works closely with farmers producing this variety, to ensure that
the seed that they produce is certified.
Seed production involves registration of seed units within 28 days from the date of planting
(Table 3.1), inspection of the seed units at different plant physiological stages, and presentation
of seed samples for certification. Four officials from the Department of Agriculture Limpopo
Province (DALP) completed the SANSOR course for authorised seed inspectors.
The first certified seed was produced in 2003, and the project was officially launched on 7August
2003. Madzivhandila College of Agriculture was registered with the National Department of
Agriculture (NDA) as a seed establishment, authorized to carry out seed cleaning and packaging.
Seed growers associations have been formed in the two districts, and a seed growers'
cooperative, and seed collection and processing points identified. Farmers grow the seed for the
cooperative individually, but they obtain many services communally, including transportation
and processing of seed.
Department of Agriculture Limpopo Province (DALP) has made available R 100,000 for the
renovation and construction of a Service Cooperative, as well as seed collection and processing
points. The seed producers are now selling their excess seed to the outside market, after meeting
their communities' seed demand.
Three seed units planted in February 2004 were harvested and certified in September 2004. The yield
was lower than expected due to poor quality of basic seed received from Capstone Seed Company,
South Africa; prevalence of maize streak virus (MSV); and poor soil fertility management.
At the time of going to press, the seed units at Mbahela, planted in September 2004, had been
harvested, the seed cleaned at Madzivhandila College and presented for certification. Those
planted in January 2005 had been registered with SANSOR, and the seed inspections carried
out at different physiological stages.
Table 3.1. Registered seed units in Limpopo Province in 2004/2005 seasons.
Seed units Planting date Hectares Status
Maraxwe 6 Feb 2004 12 Certified
Tshiombo Block II 6 Feb 2004 10.5 Certified
Tshiombo Block III (Mathombo-Tshwuka) 30 Feb 2004 16 Certified
Mbahela 1 Sep 2004 18 Certified
Sptizkop 4 Dec 2004 15 Certified
Maraxwe 10 Jan 2005 18 Certified
Tshiombo II 24 Jan 2005 18 Certified
Mianzwi 24 Jan 2005 9 Certified
Tshiombo Block III 10Jan 2005 9 Certified
Provision of basic seed
Agricultural Research Council (ARC) has provided basic seed for multiplication, as shown
in Table 3.2. Seed units in Capricorn and Vhembe district have been registered and certified. At
the beginning of the project the farmers were given basic seed for free, but this was found to be
unsustainable. The farmers are currently buying basic seed at R8.00 (1US$ = R 6.20) per kilogram.
Quality protein maize
Basic seed of Obatanpa, a quality protein maize (QPM) variety, has been made available to
farming communities in Vhembe and Capricorn, as shown in Table 3.2. Initially the basic seed
was offered free to farmers, but they are now paying R8.00 per kilogram. Analysis by the
University of Pretoria showed that on average, Obatanpa has amino acid levels 40 to 80% higher
than normal white maize (Table 3.3).
Mrs. Rosinah Lebago won first prize in the SANSOR Emerging Seed Entrepreneur of the Year 2004
contest in South Africa. The prize-giving ceremony was held in May 2005 in Durban (Fig. 13).
Figure 13. Mrs. Rosinah Lebago, SANSOR Emerging Seed Entrepreneur of 2004, with her award.
Table 3.2. Projects that received basic seeds from Agricultural Research Council, South Africa (ARC).
Projects District No. farmers Area planted (ha) Variety
Dzumauli Vhembe 106 25 Obatanpa
Mulodi Vhembe 10 6 Obatanpa
Nana Vhembe 19 2 Obatanpa
Mianzwi Vhembe 22 6 ZM 521
Maraxwe Vhembe 43 12 ZM 521
Mphambo Vhembe 32 12 SAM 39
Tshiombo Block III Vhembe 38 16 ZM 521
Hlalahala-Madonsi Vhembe 8 6 ZM 521
Strytskraal Sekhukhune 9 4 SAM 37
Mogalatsane Sekhukhune 3 1 SAM 43
Mashushu Capricorn 6 2 Obatanpa
Mashushu Capricorn 15 6 ZM 421
Jack Mafarane Capricorn 33 13 ZM 521
Monare Capricorn 5 2 ZM 521
GaThaba Capricorn 6 2 ZM 421
GaThaba Capricorn 6 2 ZM 521
GaMogano Capricorn 4 2 ZM 521
Total 365 119
Table 3.3. Nutritional analysis of Obatanpa, a quality protein maize (QPM) variety.
% increase of essential amino acids
Amino acid QPM White maize in QPM compared to normal maize
Aspartic acid 0.96 0.52 85%
Glutamic acid 2.02 1.49 36%
Serine 0.66 0.39 69%
Glycine 0.54 0.30 80%
Histidine 0.31 0.25 24%
Arginine 0.77 0.40 92%
Threonine 0.51 0.32 59%
Alanine 0.88 0.53 66%
Proline 1.17 0.77 52%
Tyrosine 0.50 0.24 100%
Valine 0.68 0.38 79%
Methionine 0.28 0.15 86%
Isoleucine 0.41 0.24 71%
Leucine 1.11 1.06 5%
Phenylalanine 0.51 0.39 30%
Tryphtophan 0.10 0.06 67%
Lysine 0.44 0.27 63%
Seed Growers Associations
Ten seed growers associations (SGAs), one in each village, were formed in Vhembe district
(Table 3.4). These SGAs have been trained by Yebo Cooperative on management of SGAs and
seed co-operatives. The Cooperative also assisted the SGAs to register a seed service cooperative
with the Registrar of Cooperatives in December 2004.
Seed Growers Service Cooperative
Currently, the farmers in Limpopo Province are selling seed at R 17.50/kg (1US$ = R 6.20).
However, economic analysis of the seed production schemes shows that the seed's total
production and processing costs are approximately R 21.60/kg, which includes both invisible
and visible costs. This high production cost is due to opportunity costs (invisible costs) which
make up about 38% of the total. These include organization of meetings, time, storage of the
seeds in the homestead, security, and support from DALP. The umbrella body organization
constitutes 4% of the invisible cost, which is paid in kind by farmers. The Limpopo agricultural
extension contributes 10% of its time and resources to the project. Visible costs make up 48% of
the actual cost, and include land preparation, purchase of basic seed, registration of units, seed
treatment/processing, seed inspection, and purchase of plastic bags. The farmers pay for these
goods and services in cash.
Three villages in Tshiombo produced 13.33 tons of seed during the last season of 2004. Seed
processing was done at the Madzivhandila College of Agriculture, where the farmers spent
1-2 weeks at a time treating and packaging the seed. A simple concrete mixer is being used for
treating the seed with Gaucho. As there are no alternative chemicals or generic forms of Gaucho
in the market at present, seed treatment has been expensive, making up more than 40% of the
total seed production costs. It is expected that generic forms of Gaucho will soon be available in
the market, which will reduce seed treatment costs.
If the above-mentioned challenges (especially the
high cost of seed dressing and processing) cannot
be addressed properly, the current community
based seed production model of Limpopo
Province will not be sustainable.
Functions of the Seed Growers Service
The formation of a Seed Growers Service
Cooperative and seed collection/processing
points will help reduce the cost of seed
production by undertaking complementary
functions. The Cooperative will be located in
the center of the participating villages, so it is
accessible to all. The SGA will be registered with
NDA Genetic Resources as a seed establishment,
and it will have the following functions:
* Serve as a seed storage, processing and
packaging establishment. This will reduce costs
of transport incurred by farmers.
Table 3.4. Seed Growers Associations (SGAs)
in the two study districts.
Number of members/
District/SGA Seed growers
Lekgothoane Ga Thaba 42
Mashushu 3-4TON 20
Ga -Mogano 18
* Trade and negotiate prices on behalf of the farmers, including negotiating discounts on Gaucho
(seed treatment insecticide), and other products.
* Liaise with DALP and SANSOR inspectors for inspection and certification of seed. Qualified
seed inspectors from Limpopo Department of Agriculture (LDA) will assist in seed inspection
and submission of seed samples from the farmers to the National Department of Agriculture
for germination and purity tests.
* Organise for marketing of seed surpluses from the local seed production scheme.
* Provide the seed growers with basic seed; the cooperative may source the basic seeds from
ARC and other suppliers.
Collection Processing Point (CPP)
This will be managed by local seed growers associations, and have the following functions.
* Storage of seed after harvesting. Storage is currently a serious problem. Farmers normally
store their seeds in their kitchen and bedrooms after harvesting, which makes fumigation
difficult. During the harvesting season children have to vacate their bedrooms and stay with
relatives until all the seed has been threshed, fumigated, and sent to the College for treatment.
* Ease transportation. Seed is currently treated in Madzivhandila Agricultural College.
Farmers living far from Madzvhandila have difficulty transporting their seed to and from the
college. At the moment the farmers around Vhembe village (about 60 km from the college)
are transported to and from the college daily, but some stay on for 2-3 days to do the seed
treatment and packaging. The costs incurred in this exercise include transportation, meals and
* Host equipment and supplies, including a weighing scale, seed treatment equipment and a
moisture meter for the farmers' use, as well as packaging materials such plastic bags, labels
and seals bought from the Seed Services Co-operative. The above equipment and materials
will facilitate seed cleaning, dressing, packaging and labelling. It will be possible for farmers to
receive maize flour or fertilizer in exchange for their seed through the CPP.
Scaling Up Strategies in Reference
J. Mulila-MittiP, C. Jacquet de Haveskercke', G. Kabwe2 and R. Katanga2
'World Agroforestry Centre Southern Africa Regional Program, P. O. Box MP 128, Mount Pleasant, Harare,
Zimbabwe E-mail: email@example.com
1 World Agroforestry Centre Zambia National Programme, PO Box 510046, Chipata, Zambia
Following more than 15 years of successful demonstration of the potential of agroforestry
technologies to make a positive impact on the livelihoods of smallholder farmers in the region,
the World Agroforestry Centre-ICRAF Southern Africa has been focusing its efforts on scaling
up these technologies to reach as many farmers as possible in the five countries (Malawi,
Mozambique, Tanzania, Zambia and Zimbabwe) where the project operates. The project has
focused on a process of institutionalizing agroforestry in the research, extension and development
policies of these countries.
Strategies employed by ICRAF to enhance this institutionalization in the program countries are:
capacity building, developing effective partnerships and networking, and promoting policies
more conducive to adoption, with the central focus being strengthening the capacity of farmers
for agroforestry innovations that meet their needs. Among the key interventions of the strategies
are the following fundamentals: farmer-centred research and extension approaches; establishment
of strategic partnerships; knowledge and information sharing; establishment of viable seed
systems; developing market options; local institutional capacity strengthening; diversification of
agroforestry options; and influencing policy at different levels.
ICRAF is building the capacity of farmers in management and problem-solving skills through a
four-pronged scaling up strategy that involves:
1. Direct training of farmer trainers and local change agent teams (group of agroforestry trainers
selected by the community;
2. Training of partners to undertake farmer training and other forms of extension;
3. Facilitating farmer-to-farmer extension and exchange;
4. Supporting national extension initiatives on sustainable agriculture.
The four-pronged approach outlined above is supported by other activities aimed at creating
awareness of agroforestry options through farmer field days, and development of extension
materials. We discuss the approach in greater detail here, and examine further refinements that
would improve agroforestry adoption and its positive impacts.
Prong 1. Direct training of farmer trainers
and local change agent teams
Through this approach, ICRAF staff members go directly to the community and train the locally
formed groups to become trainers in agroforestry technology establishment and management.
The communities select the individuals that should receive the training. The trainers are trained
over two years through a practical, field-based modular agroforestry training plan (year 1 on
food security and year 2 on income generation). Other critical development issues such as HIV/
Aids, community mobilization and farmer organization are integrated. The training is modular,
and allows farmers to implement one stage of training outputs before starting the next module.
This 'Farmer first' approach to capacity building has been found to be efficient (Boehringer,
2002) as it ensures that farmers enhance their skills and knowledge through training.
Farmer trainers have been very effective in promoting the adoption of agroforestry technologies
when the trainers themselves have experienced the benefits of practicing agroforestry on their
own farms. The project's experience has also shown that the multiplier effect of this approach
can be significant when the farmer trainer is highly motivated and committed.
Prong 1 has been implemented as the major approach in the pilot scaling up areas (PSUAs)
of the USAID-supported TARGET project in Malawi, Mozambique, Zambia and Zimbabwe.
PSUA change agent teams (each comprising 24 farmers, two traditional leaders and four local
extension officers) were selected. In the project, ICRAF facilitated group formation. During
the sensitization meetings, the community would be advised on the importance of selecting
good representatives that would ensure the communities benefited from their involvement
as trainers. Alternatively, selection was done through partners who mobilized and assisted
communities to form training groups.
By training the local change agents, the project has built partnerships that are beneficial to the
sustainability of agroforestry dissemination at community level, as well as implementation of
other development activities in the area. The change agent teams have the potential to become
effective community based organizations (CBOs) for development of other enterprises in their
communities. Continued collaboration of the change agents from the different PSUAs also
allows for the establishment of community-based networks for agroforestry development work.
Training farmer trainers
The training could either take 2 years, or be done as an intensive week-long course. The
modules presented in both approaches are the same and include:
* Introduction to agroforestry (all agroforestry technologies described);
* Nursery establishment and management;
* Transplanting, tree management, and horticulture (budding and grafting);
* Leadership, team building, marketing, farmer organization;
* Enterprise development.
Other activities such as farmer competitions, review meetings, exchange visits and field days
are conducted with farmers to complement their activities in the field.
The farmer trainers are also exposed to skills and practice on preparing and presenting lessons.
At the end of the training the farmer trainers make an action plan for their area, which includes a
program for training, distribution of seeds, setting up of a nursery and monitoring of activities.
Roles and responsibilities of the trainers
Farmer trainers are expected to undertake the following tasks: create awareness among their
fellow farmers on agroforestry; train fellow farmers; source and distribute seed; establish
nurseries; and monitor field activities. In addition, the trainers establish their own fallows, so as
to lead by example. Trainers are also expected to conduct field days and host farmers' tours.
While it stands to reason that capacity building at the local level is an effective way of creating
sustainable dissemination of agroforestry technologies, experience from the field, especially
in the Chinyanja Triangle (TARGET) Project, covering Malawi, Mozambique and Zambia,
has shown that the approach requires enormous levels of continuous technical and logistical
support. It has also emerged that ICRAF does not have the capacity (human and otherwise)
to offer direct support to farmers. Major modifications to the approach include a significantly
reduced role of community mobilization by ICRAF. It is expected that this shift will contribute
to enhanced local ownership of community dissemination work.
Prong 2. Training of partner staff
Prong 2 is designed to build the capacity of partner organizations (largely NGOs) that work
at the grassroots level. Ideally, partner NGOs should be able to contribute financially to the
training of their staff/ farmer trainers, alongside their core business. This approach has led to
mutual learning amongst partners by directly sharing information and experiences, and an
increase in joint planning and implementation of agroforestry and other activities.
In this approach partner organizations select some of their staff to be trained, and invite ICRAF
to provide the training. These partner staff are expected to, in turn, train their community
groups. However, staffing limitations within partner organizations have stood in the way of this
approach; so far, partners have only been able to mobilize community groups they wish trained,
and provide resources for the training.
The project has established partnerships with a wide range of development agencies that
include NGOs, CBOs, and faith-based organizations (FBOs) in all five countries. It has been
found that partners with a strong grassroots presence are best able to reach large numbers of
farmers, and monitor field activities at the farm level. Where such partners have long-term
commitment in the community, sustainability of the adoption of agroforestry technologies is
more secure, The most effective partners are also those that consider agroforestry a priority
in their development agendas. Such partners typically commit substantial resources towards
promotion of agroforestry in their areas. For Prong 2 to be effective therefore, ICRAF needs to
identify and work with partner NGOs/CBOs that use appropriate participatory approaches
and are genuinely committed to the empowerment of local community development structures,
including the establishment of viable community based seed production systems.
Prong 3. Farmer-to-farmer exchange
This approach involves exposing new farmers to agroforestry by facilitating their visits to
farmers who have been practicing agroforestry for some time and have started to reap its
benefits. For farmers, "Seeing is believing." (Pretty, 1995). The process of farmer-to-farmer
training starts with the host farmers being approached on the willingness to host their fellow
farmers. After agreement is reached, the host farmers meet as a group in their own area and
assign roles to different individuals. These roles include training, guiding the field visits,
and appointing a moderator who is responsible for logistics. As the benefits accruing from
agroforestry technologies (especially the soil fertility improvement options) take time, showing
new farmers the benefits realized by fellow farmers from the technologies is a very powerful
way of promoting adoption.
Farmer-to-farmer exchange visits were a major approach for enhancing adoption of
agroforestry technologies in the TARGET project. In this project ICRAF facilitated cross-border
visits of farmers between Zambia, Malawi and Mozambique. Local change agent teams were
the target audience for the exchange visits. The visits helped team members hear first hand, the
experiences and benefits of agroforestry from practicing farmers. Furthermore, other farmers
within the locality of host farmers, joined in the visits, and learnt about agroforestry.
Exchange visits are, however, expensive and therefore should not be used on their own, but to
complement other scaling up pathways. Where ICRAF works through partners, exchange visits
could be encouraged and costs shared with partner organizations. It is expected that partners
will be willing to pay for exchange visits because of their effectiveness in enhancing adoption.
Ultimately, however, the most sustainable way to support exchange visits, is by the farmers
themselves contributing towards their costs.
Prong 4. Support to existing initiatives
In all five countries there has been substantial institutionalization of agroforestry in the national
research and extension system. However, the level of government support for promotion of
agroforestry technologies varies from country to country. Each of the countries has a National
Agroforestry Steering Committee (NASCO) responsible for facilitating the institutionalization
of agroforestry in research, extension and development.
Specifically the NASCOs' roles include
* identifying priority research areas and guiding donor support accordingly;
* reviewing the research proposed by various departments and advising them on areas for
* developing policies on the development of agroforestry;
* liaising with extension services to facilitate technology transfer; and
* organising seminars and workshops on agroforestry. Hence the NASCOs are critical for
facilitating vertical processes of promoting the adoption of agroforestry.
However, the operations of the NASCOs have been severely hampered by funding constraints,
and as such they have not been very effective in attaining their mission, particularly at the
development level. Furthermore, the NASCOs have not been as effective in influencing
government policies as expected, likely because the committees are external to government
NASCOs' memberships have typically been skewed towards the national research system,
extension, and to a lesser extent institutions of higher learning. It has been recognized that
such a composition is not ideal. Inclusion of NGOs and the private sector is essential. This fact
is appreciated by some of the NASCOs and reconstitution of membership has been initiated
The role of public extension agents in scaling up
It is important to note that the ultimate goal of scaling up, regardless of which prong is used
in dissemination, is to reach more farmers, through trained farmer trainers. However, the
sustainability of this local farmer-to-farmer extension depends on continued support from the
local public extension system.
In areas where extension officers are present, it becomes easy to organise the farming community
into groups for training and other events. Farmers and the local farmer trainers require
continuous technical backstopping, which can best be provided by personnel that live within the
community, such as agricultural extension officers. Many NGOs have high staff turnovers and
sometimes, as a result of their short durations in the farming communities, are not able to connect
effectively with them. Hence, prong 4 becomes important as it aims to develop and enhance the
capacities of these agricultural extension staff to a level that enables them to confidently provide
technical backstopping to the farmer trainers and farmers, as well as partner staff.
While Prong 4 has great potential for facilitating horizontal processes of promoting adoption
of agroforestry technologies (scaling out), a major challenge has been under-investment in the
public research and extension systems.
In reviewing the different approaches and documenting lessons learnt in scaling up strategies,
four additional prongs have been proposed:
* Prong 5. Training of partner staff/farmer trainers by established trained local partners or
consultants. This prong is similar to Prong 2, except that instead of ICRAF, a local institution
or consultant does the agroforestry training. This approach can be used to train partners based
far away from where ICRAF operates. The CBO, NGO, National Agroforestry and Education
Training Teams (NAFTs) carrying out the training will have gone through agroforestry training
by ICRAF (prong 1 or 2). This prong ensures that more training requests from interested
organizations can be accommodated, eases the burden on ICRAF staff, and builds local capacity.
* Prong 6. Support to agroforestry networking. This involves support to agroforestry networks,
as a means to cater for the less strategic organizations, as well as updating existing well
established partners. Grouping all interested organizations and stakeholders into agroforestry
networks can cut costs and increase outreach. These networks should primarily be seen
as fora for the exchange of knowledge and experience, and fostering collaboration among
participants. Network members are encouraged to meet frequently on an individual basis,
share reports and materials, organise exchange visits and synchronise their activities more.
* Prong 7. Establishment and strengthening of school community links. Reaches farmers
through agroforestry activities in schools.
* Prong 8. Sensitising policy makers about the benefits of agroforestry. By producing
policy briefs and use of public media channels and events (local radio, TV programmes,
documentaries, field days, agricultural shows etc.), it may be possible to reach policy
shapers including parliamentarians, provincial-, district- and village-level administrators,
and traditional authorities. This can catalyse the adoption of agroforestry in their respective
The paper has reviewed the scaling up concept for ICRAF Southern Africa by analysing the
initial four prongs that have characterized dissemination work, and highlighted lessons learnt
so far. Recommendations of new approaches have then been made. There is no single recipe
to scaling up, and different approaches can be successful depending on the innovation, the
environment, and the resources at hand (Franzel et al., 2004). The paper has also highlighted
the important success factors necessary for scaling up: farmer-centred research and extension
approaches, knowledge and information sharing and facilitation, and building of local capacity
Boehringer A. 2002. Facilitating the wider use of agroforestry in Southern Africa. In: Franzel, S., Cooper
P, Denning GL and Eade D (eds) Development and Agroforestry: Scaling-up the Impacts of Research.
ICRAF /Oxfam, UK.
Pretty, J. N. 1995. Regenerating agriculture: policies and practice for sustainability and self-reliance. Earthscan
Publication Ltd., London UK.
Franzel, S.; Denning, G.L.; Lilleso, J.P.B.; Mercado, A.R. World Agroforestry Centre (ICRAF), Nairobi (Kenya)
2004. Scaling up the impact of agroforestry: lessons from three sites in Africa and Asia. Agroforestry
Systems 61 p. 329-344. ICRAFP