objective 1b: groundnut, chickpea and pigeonpea for...
TRANSCRIPT
1
Objective 1: Groundnut, chickpea and pigeonpea for Eastern and Southern
Africa
Several activities have been undertaken in Eastern and Southern Africa (ESA) during the past two years
towards meeting the key milestones of the project. This report will provide highlights of the major
activities undertaken and the achievements in terms of the progress made and the key findings from the
research undertaken during this period.
1. Legume situation and outlook assessments
1.1. Mapping of target areas and crop production zones
Spatially disaggregated data for all the target crops has been collected from all the three countries –
chickpea (Ethiopia), groundnut and pigeonpea (Malawi) and groundnut and pigeonpea (Tanzania).
Whereas information is available at the local level in areas or districts where project activities have
started, such disaggregated data is not available at district and sub-district level nationally. The available
data at the zonal and district levels for all crops has been mapped to help delineate and define the major
production areas and target regions for the project in each country. Future work needs to determine the
suitability of selected varieties in the specific farming systems in the larger target environment. We have
also initiated discussions with Harvestchoice to see how the remaining gaps for the legume crops can be
filled.
1.2. Situation and outlook analyses for targeted legumes
The situation and outlook assessments have been completed for all the three legume crops in selected
countries: chickpea (Ethiopia), groundnut and pigeonpea (Malawi). There is an ongoing effort to
complete a market survey for pigeonpea in Tanzania which may also provide some of the required
information for a situation and outlook report on pigeonpea in this country. Data used for this analysis are
primarily obtained from existing data at ICRISAT and from secondary sources such as FAOSTAT,
published documents and various national reports. The secondary data comprised of aggregate data on
global, regional and national production output, data on export and import volumes of targeted crops,
global and national price trends over years, and data related to access to seeds. The time series data used
ranges from a minimum of 3 years to a maximum of 14 years. A number of statistical tools were
employed to analyze, summarize and present the data. For analyzing the historical trends over the years
and estimate the growth rate, descriptive statistics are used. The global food projection modeling
2
framework of IMPACT (the International Model for Policy analysis of Agricultural Commodities and
Trade) recently calibrated and adapted for policy analysis of dryland crops is applied to examine the
future situation for chickpea in Ethiopia and groundnuts and pigeonpea in Malawi. These country sub-
sector assessments have been already published in the TL-II website for a wider use. The reports are
expected to feed into regional situation and outlook reports that would outline the current production
conditions, key socioeconomic and technological constraints and key interventions for unlocking the
potential of the targeted legumes in the region.
In terms of the projected area and production trends over the 20–year horizon for which the projections
are made, the results suggests that chickpea area and production in Ethiopia will show significant growth
in the coming years. However in order to leverage the chickpea sub-sector for poverty reduction and food
and nutritional security in the country, results suggest that there is a need to design a more flexible and
sustainable seed systems that meet the needs of the resource poor farmers. This requires policy makers to
open up the seed sector and encourage and assist private seed companies and community seed producer
associations by improving access to agri-business development services and empowering cooperatives
and village agro-dealers. Results also suggest the need for improving the performance of chickpea value
chains by increasing farmer linkages with the industry and exporters, reducing transaction costs and
targeting the development and distribution of large-seeded kabuli varieties that offer price premiums in
international markets. Ultimately, the competitiveness of smallholder farmers in chickpea production will
depend on accessing, adopting and adapting promising varieties and production practices. This will
require large-scale demonstration efforts and participatory variety selection in key production
environments to identify locally adapted and profitable varieties.
The future outlooks for groundnuts in Malawi seem promising; however, there are a number of
constraints that negatively impact on the development of the groundnut sub-sector. The analysis has
revealed weaknesses in the current seed systems as well as in the enforcement of quality standards. The
technology delivery and the grain marketing systems are underdeveloped, leading to the low use of
improved technologies and the production of poor quality of nuts with high levels of aflatoxins that make
it unacceptable in the international markets. Although the volumes of groundnut exports remain lower
than the levels seen in the late 1980s, the review has shown that Malawi maintains a comparative
advantage in groundnut production and competitiveness in exports suggesting that there is scope for
increasing groundnut exports once the required quality standards are adhered to. Therefore the findings
suggest the need for faster productivity enhancement, strengthening seed delivery systems to reach more
farmers and the development of existing value chains.
3
For pigeonpea in Malawi, historical trends show a rise in harvested area, yield and production.
Furthermore, the outlook analysis based on production and exports simulations shows that area,
production as well as domestic demand will continue to rise. Nonetheless, there are a number of
constraints that negatively impact on the development of the pigeonpea sub-sector. The findings reveal
existing structural weaknesses in seed and technology delivery and grain marketing systems, which have
an effect on the diffusion and adoption of improved technologies and consequently the on-farm
productivity and profitability of this crop. Furthermore, while global demand for pigeonpea continues to
rise, there is an increasing pressure on African farmers to benefit from these markets due to intense
competition for export markets (mainly India) from Myanmar and other emerging producers, as well as
the surging demand for other substitutes. Unless productivity growth and market development help offset
these threats, it can severely affect the overall competitiveness for pigeonpea farmers. The findings
suggest the need for faster productivity enhancement, strengthening seed delivery systems to reach
farmers who continue to rely on low-yielding and disease-susceptible local varieties and development of
existing value chains and alternative pigeonpea export markets.
2. Baseline surveys
This is one of the most demanding activities and milestones undertaken during the first two years. This
started with extensive discussions with various partners, including the project breeders and seed systems
specialists, and visits to the project sites to identify the key issues to be captured in the baseline data. A
standardized survey instrument was then developed at two levels (household and village/community) for
all countries to gather comparable data for understanding limiting constraints, farmer and market
preferences and measuring adoption and multi-dimensional impacts of the project. Baseline data has been
collected for all the three crops in three target countries (pigeonpea and groundnut in Malawi and
Tanzania, and chickpea in Ethiopia) and personnel from local partners were trained in methods for survey
design, sampling and good practices in administering survey instruments. About 39 local staff and
supervisors were trained and the instruments were pre-tested before the surveys were administered using
personal interviews. The baseline studies also benefited from other complimentary projects (e.g. Treasure
Legumes project supported by IFAD) in terms of sharing fixed costs which allowed surveying a larger
sample of the target farmers.
In Ethiopia a combination of stratified and purposive sampling methods was used to select the three
districts included in the survey in the project target areas, namely Gimbichu, Lume-Ejere and Minjar-
Shenkora. These districts represent one of the major chickpea growing areas in the country where new
4
varieties are beginning to be adopted by farmers. Eight kebeles1 from each of Gimbichu and Lume-Ejere
districts and ten kebeles from Minjar-Shenkora district were randomly selected from where a random
sample of 700 households was selected for detailed household survey. This include 400 households in the
two districts (Minjar-Shenkora and Gimbichu) targeted under the TL-II project and 300 households in the
one district (Lume-Ejere) selected for IFAD project. Village level data was also collected from about 40
villages that fall within the 26 kebeles. EIAR/DZARC took the lead for the baseline survey in Ethiopia.
While the larger datasets help understand the broader production and market constraints for chickpea in
Ethiopia, the data from the 400 households will be relevant in evaluating the adoption and impact of the
TL-II project.
In Malawi a total of 594 randomly selected households were surveyed in four purposively selected
districts. The multi-stage sampling process included selection of 16 Sections (four per district) and 48
villages (four randomly selected among the groundnut and pigeonpea growing villages in each Section).
The sample included 298 households in two districts (Michinji and Balala) targeted under the TL-II
project and 296 households in two other districts (Thyolo and Chiradzulu) under the sister project
supported by IFAD. The village level data was also collected from 47 villages in the four districts. About
22% of the sampled households were female-headed, while about 50% of the respondents were also
females. The Malawi baseline was coordinated by ICRISAT-Malawi with support from the University of
Malawi.
In Tanzania the sampling framework is based on a random sample of villages in four districts in the
Northern zone, representing the main pigeonpea and groundnut growing division and wards in Babati and
Kondoa and pigeonpea growing areas in the two districts of Arumeru and Karatu. In each of the four
districts three major divisions were selected giving rise to a total of 12 divisions. Subsequently, two wards
were sampled in each of the selected divisions resulting into a total of 24 wards. Twenty five farmers
were then randomly sampled from a list of farm families in each village and ward. A total of 600 farm
households (of which 50% were supported under the IFAD project) in four districts were surveyed using
the standardized survey instrument. In addition to household surveys, a total of 24 communities/villages
(6 from each of the survey districts) were also surveyed using key informants to gather data on broader
village/community level socioeconomic factors. The survey was undertaken in collaboration with the
Selian Agricultural Research Institute (SARI) and was administered from October to December 2008.
1 It is usually named peasant association and is the lowest administrative unit or local council in the country.
5
The data was subsequently transferred into STATA econometric software package for the analysis. The
data was further checked for consistency and completeness and analyzed. Descriptive statistics such as
frequencies, cross-tabulations, means and ratio were employed to analyze, summarize and present the data
in the baseline report. Analysis was conducted by disaggregating important relevant information by
districts and crops so that a snapshot comparison can be made as needed. In order to improve data quality
and comparability across countries, data entry, analysis and write up of the report is led by ICRISAT
Nairobi. Lack of skilled socioeconomics support within the collaborating national programs also made it
difficult to enter, code and manage the data and lead the analysis at the local level. At the time of
reporting analysis and write up of baseline reports for Ethiopia and Malawi have been completed while
the analysis and report writing for the Tanzania dataset is underway.
3. Major findings
3.1 Household assets
In Ethiopia, the average farm size for the whole sample in the study areas are about 2.24 ha. The average
total land holding for the upper 25% of the sampled households is about 4.1 ha of which 3.8 ha is
cultivated land and 0.5 ha is a fallow land. For the lowest 25% of the sampled households, the average
land holding is 0.6 ha of which the share of fallow land is nil. Average number of plots owned by
sampled households’ amounts to 7.4, with an average size of 0.38 ha. The average farmland under rainfed
conditions is 1.99 ha whereas average irrigated land is negligible (0.02 ha).
Looking at the average livestock holding, sheep and oxen seem to be the highest with mean holding of 2.9
and 2.8, respectively. About 87% of the sampled respondents own house with iron roof whereas only
2.4% own house with stone wall. Only 2% own bicycle which suggest the inferior use of the item for
transporting purposes. About 77% of the sampled households own at least one radio whereas the
ownership of television is very minimal (1.3%). Only 6.1% of sampled households own mobile phone in
the survey regions.
In Malawi, the mean land holding size for the study area is very low: 1.1 hectares. The average land
holding for the upper 25% is 2.1 ha of which 1.9 ha is cultivated. Households in the lower two quartiles
do not practice fallowing due to smaller land sizes. Overall, about 90% of the land owned is cultivated
while about 10% of it is left under fallow.
Looking at livestock ownership, chickens (52%), goats (29%) and pigs (7%) are the predominant types of
livestock in the survey area. Using tropical livestock unit (TLU) as a unit of measurement, households in
6
Malawi have an average TLU per household of 0.37. About 46% of the farm households own a bicycle
while 54% of the households own a radio. Bicycles are one of the major means of transportation in
Malawi. They are used to transport goods as well as people. Very few own a mobile phone (5%) and
Television (3%). About 70% of the households have at least a house made from bricks. Only 26% of the
households dwell in houses with iron sheet roofs.
In Tanzania, the average total owned and operated land size amounts to 4.7 ha and 4.9 ha, respectively.
About 0.8 ha is used as fallow land. The average total own land under rainfed amounts to 4.69 ha whereas
the average own irrigated land only 0.13 ha suggesting the need to promote and invest in irrigation
technologies.
3.2 Area cultivated for legumes
In Ethiopia, the predominant legume crops grown include chickpea, field pea, faba bean, lentil and grass
pea. On average about 0.68 ha (which is about 28% of the total cultivated land) has been allocated to
legumes during 2006/07 cropping season of which chickpea take the lead with about 0.38 ha. Area
devoted to desi chickpea is on average about 0.22 ha which is slightly higher compared to kabuli types
(0.16 ha). The average area devoted to field pea and faba bean is the lowest among the legumes with
about 0.05 ha each. In terms of the proportion of farmers planting legumes during 2006/07 cropping
season, desi chickpea take the lead with about 54% followed by lentil (37%) and faba bean (34%). Kabuli
chickpea is grown by about 30% of the sample farmers during the same season. Looking at chickpea desi
and kabuli together, results suggest that about 65.7% of sampled farmers grow chickpea as a crop.
In Malawi, the prevalent legume crops grown include groundnuts, pigeonpea, chickpea and beans. The
average land allocated to groundnuts and pigeonpea is about 0.3 ha each but it is important to note that
these are often intercropped with other crops like maize. Chickpea and beans have the lowest area
coverage with 0.03 and 0.01 ha respectively. When it comes to the share of area allocated to different
crops, 17% and 15% of the total cultivated land is allocated to groundnuts and pigeonpea, respectively.
The proportion of farmers planting groundnuts during 2006/07 cropping season is the highest among
legume farmers followed by pigeonpea. For the sample as a whole, on average about 35% and 15% of
the sampled farmers planted groundnuts and pigeonpea during the same season, respectively.
However, the percentage of farmers growing groundnut or pigeonpea is much higher when one
considers the results by the growing regions.
7
3.3. Cropping pattern, input use, yield and returns
In Ethiopia, chickpea is grown in wheat and teff-based farming systems mainly through crop rotations
using the residual moisture at the end of the rainy season mainly on black soils. Hence, bread wheat and
teff are the first and second most common cereal crops produced among the sampled households in
survey regions. Desi chickpea is the third most popular crop produced by 53.6% of sampled households.
Bread wheat, kabuli chickpea and durum wheat are the top three crops.
Generally the average amount of fertilizer use on legume crops is low and only few farmers apply any
commercial fertilizers. From the total seed planted during 2006/07 cropping season the share of seed
purchased from the market is about 48.9% for kabuli chickpea while it is only 3.1% for desi chickpea.
The average seed rate for kabuli and desi chickpea amounts to 186.5 and 149.8 kg per ha which is
relatively higher than the recommended rate of maximum 140 kg per ha. About 70% of households use at
least some chemicals in crop production to control pests and diseases on chickpea. The average number
of ploughing for kabuli and desi chickpea amounts to 5.4 and 4.5, respectively whereas the frequency of
weeding for the same crops is about 2.4 and 1.6, respectively. The average labor demand is about 97
person days per ha for kabuli and 83 person days per ha for desi chickpea. The average yield for kabuli
chickpea is about 2.5 tons per ha whereas for desi chickpea it is about 1.9 tons per ha. Kabuli and desi
chickpea have the third and fourth highest gross margin after barley and teff in terms of returns to land
and management among all cultivated crops. When comparing against legumes, chickpea generally stands
out to be the crop with the highest gross margin. Kabuli and desi chickpea have a net-return of about ETB
75322 and ETB 7088 per ha while other legumes such as faba bean and lentil have a net-return of about
ETB 2681 and 4843, respectively.
Malawi is predominantly a maize country and, therefore, not surprising to note that over 90% of the
households planted maize in the 2006/07 cropping season. Groundnut is the second most frequently
cultivated crop (55%), while pigeonpea comes third and it is cultivated by 40% of the households in the
sample. Consistent with the distribution of households growing each crop, maize is allocated more land
(0.70 ha) followed by groundnuts (0.3 ha) and pigeonpea (0.3 ha). Yield results indicate that improved
varieties consistently exhibit higher yields than local varieties. The yield of maize is about 1129 kg/ha,
while the yield of groundnut and pigeonpea averaged 622 kg/ha and 355 kg/ha, respectively. About 85%
of the sampled households apply some chemical fertilizer to some of the crops they grow. Interestingly,
farmers also applied some fertilizer to legume crops (pigeonpea and groundnuts); however, these
quantities are likely to have come from fertilizer applied on maize intercrops. The average seeding density
2 During the time of the survey 1US$ = 9 Ethiopia Birr.
8
for groundnuts is 47 kg/ha (range from 14 kg/ha in Thyolo to 63 kg/ha in Mchinji). About 22% of the
groundnut seed is purchased. The average seed rate for pigeonpea is 13 kg/ha of which 19% is purchased
from local markets and the rest is recycled own saved seed. The results on gross margin analysis indicate
that tobacco is the most profitable crop with the highest average return of about MK 120,000/ha3.
Groundnut and maize have the second (about MK 30,000/ha) and third highest (MK 20,000/ha) average
return, respectively. Interestingly, pigeonpea remains one of the least profitable crops being cultivated.
3.4 Poverty profile
In Ethiopia, the average per capita household expenditure which includes food grains, livestock produce,
vegetables, beverages, clothing and social activities amount to about ETB 2164 for 2006/07 cropping
year. Out of the per capita expenditure, cash expenditure is ETB 819, while the rest, ETB 1346, is
consumption of commodities produced on the farm. The implication is that most food in rural areas of
Ethiopia is derived from own sources. The average annual household income is about ETB 16,797.
Income sources include crop income, livestock income, non-farm activities and wage income. Crop
income (including amount consumed at home) accounts for the largest share of 78%, followed by
livestock incomes of 15% and non-farm business activities of 6% of the total household income. This
indicates that livelihoods are predominantly dependent on crop production while the role of off-farm
income is very limited. The share of wage income is insignificant compared to other sources (0.8%).
When stratified by income quartiles, households in lower income quartiles have lower income shares
coming from crop production than households in higher income quartiles and this result is consistent
across all the three districts. Surprisingly no significant income disparity between male and female headed
households is observed in the study area. Among all the quartiles, the income of female headed
households is a bit higher than the male headed households. Using the international poverty line of 1 US$
per day, results show that about 77% of the sampled households are below the poverty line. The national
average poverty incidence using national poverty line of about US$ 0.45 per day is about 47% for rural
population. The poverty gap and poverty severity for the entire sample is estimated at about 0.12 and
0.04, respectively.
In Malawi, the average annual per capita expenditure for the sample is MK 9068, ranging from MK 8200
for Chiradzulu to MK 9971 for Mchinji district. A larger proportion of the cash expenditure is on
clothing, bedding and energy (25%). Subsequently, the expenditure on livestock products (22%),
vegetables and other food items (21%) are equally important. Interestingly, expenditure on food grains
only accounts for 18% of the cash expenditure. This can be attributed to the fact that most rural
3 During the time of the survey 1US$ = 140 Malawian Kwacha.
9
households consume what they produce. The average annual income for the sample households is MK
65582. Like in Ethiopia, crop income accounts for the largest share of 71% of the total household income
while non-farm income accounts for 16% of the household income. Wage income is the third major non-
farm income source for the study area accounting for 11.3% of the total household income while livestock
income is negligible, accounting for only 2.6%. Crop income accounts for 97% of the total income in the
households belonging to the lowest income quartile but declines to 59% in the highest income group.
Using a national poverty line of 0.3 US$ per day, results on poverty incidence indicate that about 70% of
the sampled households live below the poverty line. These poverty levels are much higher than the
national poverty rate of 55% reported by the National Statistics Office. Overall, more female-headed
households (77%) are poor compared to male- headed households (70%).
3.5. Technologies, production and marketing
3.5.1 Variety preference and adoption
In Ethiopia, the largest proportion of households receive information about kabuli varieties from
neighboring farmers (46.6%) while about 45% receive such information from government extension
services. The third most important source of information is farmer cooperative (26.1%). Neighboring
farmers also remain the major source of information (72.4%) for desi varieties. The improved kabuli
varieties arerti and shasho are known to 43.9% and 48% of farmers, respectively. While majority of the
farmers (98%) knew the local desi varieties, less than 5% were aware of the improved desi varieties. The
most widely grown improved variety among chickpea farmers remains shasho (20.6%) followed by ejere
(11.7%) and arerti (10%). The proportion of chickpea farmers who planted improved desi during 2007 is
less than 3% while about 76% planted the local desi. About 54.5% of the chickpea area is allocated to
local desi followed by shasho (21%) and ejere (11.9%). The potential demand for planting new varieties
is much higher than those who currently plant them – indicating problems in accessing seed. The major
reasons why some farmers never adopted the improved varieties was lack of access to seed and fear of
theft during greens stage. The third and fourth major reasons are related to shortage of land and lack of
cash to buy seed and/or lack of credit. Generally kabuli varieties are highly preferred by chickpea farmers
for their high economic return in addition to their grain color and size. Kabuli varieties perform superior
in terms of yield per ha compared to the desi types.
In Malawi, about 74% of the households are aware of at least one pigeonpea variety. The awareness rate
for improved pigeonpea varieties (ICP 9145 (released in 1987) and ICEAP 00040) is much lower. Of the
two improved varieties, ICEAP 00040 is the most widely known by 20% of the farmers while ICP9145 is
only known by 8% of the farmers. With regards to adoption, although more farmers expressed awareness
10
of pigeonpea, fewer indicated that they have ever grown the crop and much fewer grew the crop in
2007/2008 season. For local pigeonpea although 71% of the farmers know the crop, only 57% have ever
grown it and only 31% grew the crop in 2007/08 season. For Mthawajuni variety, 53% knew the crop,
48% indicated that they ever grew the crop, while 44% actually grew it in 2007/08. As for the improved
varieties, farmers are more aware of ICEAP 00040 (20%) and 18% have ever grown it but only 13% grew
it in 2007/08. In general the sample adoption rate of improved pigeonpea variety in Malawi is about 10%
however if the technology would have been universally known within the population, the potential
adoption rate is about 45%. There is about 35% adoption gap due to lack of awareness of improved
pigeonpea varieties that would justify investment in its further dissemination. As for groundnuts,
chalimbana is the most widely known variety (84%) followed by CG7 (53%) and manipintar (11%).
While 84% of the sample farmers are aware of chalimbana, only 69% have ever grown the variety and
only 43% grew chalimbana in 2007/08.
Farmers were also asked to give their reasons for not planting some of the varieties they knew. Aside
from the lack of awareness for some of the legume varieties seed is a major constraint to adoption. In
Malawi, about 60% of the farmers reported that they lacked seed for some of the groundnut and
pigeonpea varieties they knew but never planted. The second major reason for non-adoption of some of
the groundnut and pigeonpea varieties, reported by about one-fifth of the farmers was that the varieties
were low yielding. Related to the seed problem is the lack of cash to buy seed which was reported by
about 10% of the respondents. Among pigeonpea varieties, mthawajuni and ICEAP040 are the most
preferred varieties with overall rankings of 4.3 and 4.1, respectively. ICPL945 and local pigeonpea are
ranked last with an overall ranking of 3.8 each. The findings further indicate that most highly preferred
varieties are liked for the three key traits they exhibit as follows; high yielding, early maturity and short
time of cooking. Interestingly, mthawajuni, considered as a local variety, is highly preferred for its high
yield, as well as its early maturity and its shorter cooking time.
In Tanzania, although about 88% of the sample respondents grew pigeonpea during 2006/07 season, only
19% planted improved varieties. Area allocated under pigeonpea is about 1.54 ha whereas the average
area for improved ones is only 0.27 ha. The share of improved pigeonpea in total pigeonpea area in the
study areas is about 18%. Groundnut is grown by about 18% of the sampled farmers during 2006/07
cropping season and the average land allocated is about 0.97 ha.
3.5.2 Market participation and marketed surplus
In Ethiopia, about 37% and 64% of chickpea farmers have participated in the marketing of kabuli and desi
chickpea, respectively. Within the kabuli category, the proportion of chickpea farmers involved in
11
marketing of shasho variety is the highest followed by ejere type. Generally the proportion of chickpea
farmers involved in marketing is relatively high (about 82%) indicating the role of this crop as a source of
cash in the study area. The marketed surplus for kabuli chickpea is about 293.7 kg which is higher than
desi types (217 kg) and overall chickpea is the fourth in terms of quantity sold in markets. Durum wheat,
bread wheat and white teff are the first three crops with the highest market surplus during the 2006/07
cropping year. About 74% of the chickpea marketed surplus is sold in the main market which is about 5-
10 km from the farm. About 89% of chickpea producers used donkeys for transporting while about 15%
used public means to ship chickpea to the market. Urban grain traders are the first major buyers of
chickpea in all the three districts followed by rural traders and rural assemblers.
In Malawi, about 73% of the groundnut farmers participated in groundnuts marketing. Each of the
farmers sold an average amount of 137 kg of groundnuts. The degree of market participations appear to
vary with the type of variety grown. In general more than three-quarters of the farmers that planted
chalimbana2005 (improved variety), sold some of the produce while marketing of produce for farmers
that grew other varieties is slightly lower. Results for pigeonpea indicate that 91% of the pigeonpea
farmers sold some pigeonpea, and that only 29% of the pigeonpea produced is marketed. On average,
farmers sold about 71 kg of pigeonpea per year. The major mode of transport to market pigeonpea and
groundnuts is carrying on head or head-load and walking on foot (38%). However, the use of head load is
particularly more prevalent among pigeonpea farmers (50%) than among the groundnut producers (27%).
About a quarter of farmers that marketed their crop used the bicycle to transport their produce.
3.5.3 Seed supply system
Analysis of the baseline data from the target countries show the following key features in relation to the
basic characteristic and performance of seed system.
Ethiopia –chickpea: Farmers who planted chickpea varieties during 2006/07 cropping season were asked
where they source the seed and the corresponding quantity. The first major source of seed for arerti and
shasho varieties is own farm-saved recycled seed followed by producers groups. About 47% of those who
planted arerti and 50% of those who planted shasho used their own saved recycled seed whereas about
33% and 26% of those who planted the same variety sourced from producer marketing groups or
cooperatives. Own saved recycled seed again played a vital source of seed for chefe (77%), worku (77%)
and local desi (84%) varieties while producer marketing groups also contribute for ejere types (33%). The
third and fourth important source of seed for most kabuli and desi varieties is local seed producer and
local trader, respectively. A relatively small share of sampled farmers also sourced seed through farmer-
to-farmer exchange and extension demonstration plots. The quantity of seed sourced from own saved is
12
the highest for most varieties of kabuli and desi chickpea. On average farmers obtained about 48 kg of
shasho, 46 kg of ejere, 39 kg of worku and 45 kg of local desi from own saved seeds during the 2006/07
cropping season. Producer marketing groups and/or cooperatives are the second major sources of seed
especially for kabuli varieties.
Malawi-groundnut: for groundnut about 63% of the seed used comes from own farm-saved recycled
seed. Some 57% of the farmers depend on this source. About 16% of the seed is bought from local seed
markets while about 18% comes from informal farmer-to-farmer seed exchange. About 24% of the
farmers bought seed from local markets. The average amount bought from different sources ranges from 7
to 8 kg. The amount received through informal seed exchanges range from 1 kg to 40 kg with an average
of 14.8 kg. About 16.7% of the farmers rely on these informal sources.
Malawi-pigeonpea: for pigeonpea about 70% of the seed used comes from own farm-saved recycled
seed. About 74% of the farmers depend on this source. About 20% of the seed is bough from local seed
markets while about 11% comes from informal farmer-to-farmer seed exchange. About 17% of the
farmers bought seed from local markets. The average amount bought from different sources ranges
between 3 to 6 kg. The amount received through informal seed exchanges range from 1 kg to 10 kg with
an average of 4.7 kg.
Generally the survey results from all study countries reveal two main seed supply systems for the three
target legumes (pigeonpea, groundnut and chickpea). They are the informal non-market based seed supply
systems and the quasi-formal, mainly market-based seed supply systems. The informal seed supply
sources included own saved seed; gifts from family and friends; farmer-to-farmer seed exchanges and
others (e.g. donations by NGOs, government agencies, farmer groups/cooperatives, research demo/PVS
plots etc.). Across both the target crops and countries, most of smallholder producers got seed from the
informal sources especially the own recycled seed normally referred to as own saved seed. Similarly, by
extension, these informal seed sources accounted again for the highest proportion of the total amount of
seed planted by the farmers in the survey sample. In general these findings are consistent with earlier
expectation that informal seed sources are the most important sources of legume seeds in the surveyed
communities. The importance of quasi-formal or market based channels increases with the availability of
new farmer-preferred varieties which creates incentives for the emergence of markets and trade in the
supply of the new seeds.
The observed low private sector participation in the seed systems may indicate the need for stronger
public support for legume seed production at least in the early stage until demand for is high to attract
private sector seed companies. The high cost of exclusion (low excludability) resulting from the potential
13
for using saved or recycled seed, renders investments in legumes seed systems unprofitable and hence
discouraging the private sector from investing in legume seed production.
4. Monitoring and evaluation
With the completion of the baseline report, it has become important to develop mechanisms for project
monitoring and evaluation. Each of the different Objective components of the project have already
development impact pathways outlining the roadmap for translating activities into measureable outputs,
which in turn lead to desirable outcomes and ultimately into impacts of the project over the long term.
Enhancing attribution and project effectiveness in meting desired goals would require steps for process
evaluation and impact assessment. Process evaluation will aim to determine to what extent the project has
been implemented as planned and identify operation and strategic lessons for smooth implementation.
Monitoring will involve description and measurement of observed outcomes along defined metrics agreed
a priori. The baseline data collected is an important monitoring tool for establishing benchmarks and
measuring progress against selected impact indicators. The key indicators for impact – both qualitative
and quantitative- are being developed.
Impact evaluation will aim to measure the project’s success in achieving stated objectives using a
counterfactual. While process evaluation requires careful description and monitoring of activities and
outcomes, impact evaluation requires measurement of tangible indicators affected by the program and
how this differs from the situation without interventions (counterfactual). The propensity score matching
(PSM) method is being developed to identify a matched sample of non-adopters (a comparator group
having similar characteristics as adopters) among the baseline households which will serve as a
counterfactual. Using the baseline survey from the same target areas, this approach allows identifying
valid comparator group of households (counterfactuals) sharing the same characteristics and probability
of adoption of new technologies. This methodology is being developed and has been tested with data
from Uganda for groundnuts.
14
List of tables showing major finding for Malawi and Ethiopia baseline survey
A. Malawi
Table 1. Fragmentation of plots and land ownership
Chiradzulu Thyolo Balaka Mchinji Total
Number of plots 3 3.6 4.4 2.9 3.5
Average plot size (ha) 0.32 0.27 0.26 0.50 0.34
Tenure status
- Total owned (ha) 0.82 0.86 1.06 1.44 1.05
- Total rented in (ha) 0.07 0.05 0.04 0.06 0.06
- Total own irrigated (ha) 0.04 0.01 0.00 0.03 0.02
- Total own rainfed (ha) 0.79 0.83 1.05 1.41 1.03
15
Table 2.
Crop
choice,
area
planted
and
technolo
gy
adoption
Crops
Hholds growing
crop (%)
Area
planted (ha)
% total crop area
allocated to improved
varieties
Maize 93.3 0.69 40.2
Groundnut 54.7 0.30 35.3
Pigeonpea 40.4 0.29 14.5
Tobacco 14.3 0.06 96.5
Cotton 7.6 0.05 93.3
Cassava 7.2 0.05 29.5
Sorghum 6.4 0.05 15.8
Sweet potatoes 4.0 0.02 56.3
Chickpea 3.9 0.03 13.8
Beans 2.0 0.01 58.3
Rice 2.0 0.00 66.7
16
Table 3. Crop yields (kg/ha), 2006/07
Crops Chiradzulu Thyolo Balaka Mchinji Total
Maize 949 1205 1093 1432 1177
Groundnut 443 258 671 722 610
Pigeonpea 368 468 372 - 389
Tobacco 796 465 1087 855 844
Cotton - - 673 - 673
Cassava 3721 330 1611 1094 846
Sorghum 190 59 - - 117
Sweet potatoes 486 1533 2597 - 1475
Chickpea 255 199 - - 232
Beans 222 638 - 1328 1079
17
Rice 1597 247 887 494 979
18
Table 4. Groundnut and pigeonpea area planted and technology adoption
Crops Chiradzulu Thyolo Balaka Mchinji Total
Groundnuts
Hholds growing crop (%) 28.9 29.2 70.1 89.6 54.7
Area planted (ha) 0.12 0.19 0.39 0.51 0.30
% area allocated to improved
varieties 13.6 58.7 26.7 41.3 35.3
Pigeonpea
Hholds growing crop (%) 44.1 34 86.1 0 40.4
Area planted (ha) 0.28 0.25 0.69 NA 0.29
% area allocated to improved
varieties 3.0 15.5 20 NA 14.5
19
Fig. 1 Relative gross income and net returns (Kwacha/ha)- accounting for family
labour
0
20
40
60
80
100
120
140
Mai
ze
Gro
un
dn
ut
Pig
eon
pea
To
bac
co
Co
tto
n
Cas
sava
So
rgh
um
Sw
eet
po
tato
es
Bea
ns
Ric
e
Th
ou
san
ds
Gross income Variable cost Return to land and management
Fig 2. Sources of variety information (%)
74
86
0
10
20
30
40
50
60
70
80
90
100
Go
ve
rnm
en
t
Fa
rme
r c
lub
NG
O
Re
se
arc
h
Sto
ck
ist
An
oth
er
farm
er
Ra
dio
/TV
Pa
ren
ts
Groundnut Pigeonpea
20
Table 5. Adoption of pigeonpea and groundnuts varieties (%)
Know the variety
(awareness % of
all farmers)
Ever Planted (%
of all hhs)
Ever Planted
(% hh aware)
Planted in 06/07
(% of all)
Pigeonpea 40
ICEAP 040 (Kachangu) 20 18 86 8
ICPL 9145 (Sauma) 8 6 76 4
Nthawa June 54 48 90 33
Local pigeonpea 72 57 79 21
Groundnuts 55
CG 7 53 38 72 21
Chalimbana 84 69 82 37
Manipintar 11 9 83 2
Chalimbana 2005 9 8 91 5
Kalisere 5 5 100 2
ICGV-90704 (Nsinjiro) 3 3 79 1
ICG 12991 (Baka) 3 2 50 0
JL 24 (Kakoma) 2 1 55 1
21
Table 6. Cropping pattern of Pigeonpea varieties (%)
Varieties
Growers (% of total
sample farmers)
Pigeonpea farmers
growing the variety (%)
Area allocated (% total
pigeonpea area)
ICEAP 040 (Kachangu) 8 19 9
ICPL 9145 (Sauma) 4 10 6
Nthawa June 33 82 52
Local pigeonpea 21 53 33
All pigeonpea 40 100 100
Table 7. Cropping pattern of groundnut varieties (%)
Varieties Growers (% of total
sample farmers)
Groundnut farmers
growing the variety
(%)
Area allocated
(% total
groundnut area)
CG 7 21 38 27
Chalimbana 37 68 53
Manipintar 2 7 7
Chalimbana 2005 5 9 7
Kalisere 2 4 3
ICGV-90704 (Nsinjiro) 1 2 1
JL 24 (Kakoma) 1 2 1
22
All groundnut 55 100 100
Table 8. Market participation and marketed surplus for pigeonpea and
groundnuts
Crops / varieties
Market participation
(% of growers)
Marketed
surplus (kg/hh)
% sold
Pigeonpea
ICEAP 040 (Kachangu) 39 29 16
ICPL 9145 (Sauma) 46 34 23
Nthawa June 55 44 26
Local pigeon pea 55 47 52
Groundnuts
CG 7 56 80 29
Chalimbana 2005 77 139 60
Chalimbana 58 87 35
ICGV-90704 (Nsinjiro) 67 162 33
ICG 12991 (Baka) 67 93 12
Kalisere 60 154 45
Manipintar 61 83 28
23
B. Ethiopia
Table 1. Fragmentation of plots and land ownership
Gimbichu Lume-Ejere Minjar-
Shenkora Total
Number of plots 7.30 8.39 6.22 7.38
Average plot size (ha) 0.38 0.36 0.39 0.38
Tenure status
- Total owned (ha) 2.33 2.56 1.79 2.24
- Total operated (ha) 2.73 2.67 2.03 2.45
- Total own irrigated (ha) 0.01 0.04 0.00 0.02
- Total own rainfed (ha) 1.93 2.20 1.76 1.99
24
Table 2. Crop choice, area planted and technology adoption
Crops Hholds growing crop (%)
Area planted (ha) % total crop area allocated to improved varieties
Kabuli chickpea 30.43 0.16 42.55
Desi chickpea 53.57 0.22 3.11
Field pea 17.57 0.04 2.26
Faba bean 33.86 0.06 0.43
Lentil 36.71 0.14 30.83
Grass pea (guaya) 22.00 0.06 3.03
White teff 66.86 0.54 1.31
Red teff 38.43 0.16 2.30
Bread wheat 94.00 0.79 36.02
Barley 38.29 0.08 3.43
Maize 25.29 0.03 1.49
25
Fig 1. Relative gross income and net returns (Birr/ha)
0
2000
4000
6000
8000
10000
12000
14000
Ka
bu
li
De
si
Fie
ld p
ea
Fa
ba
be
an
Le
ntil
Gg
ua
ya
Wh
ite te
ff
Re
d te
ff
Bre
ad
wh
ea
t
Gross income Variable cost Return to land and management
26
Fig. 2. Household income by income quartile (Jan – Dec 2007)
Household income (Birr/year) by quartile
0 10,000 20,000 30,000 40,000 50,000
Lowest
Second
Third
Highest
Lowest
Second
Third
Highest
Mal
e-he
aded
hous
ehol
ds
(N=
651)
Fem
ale-
head
ed
hous
ehol
ds (
N=
49)
27
Fig 3. Sources of variety information
0
10
20
30
40
50
60
70
80
Gov
ernm
ent e
xten
sion
Farmer c
oope
rativ
e
NG
O
Rese
arch c
entre
See
d/gr
ain
stoc
kies
t
Ano
ther
farm
er/ ne
ighb
our
Radi
o/ne
wsp
aper/T
V
Pro
ducer m
arketin
g gro
ups (P
MG
)
Family
Rela
tives
PercentKabuli varieties
Desi varieties
28
Table 3. Adoption of chickpea varieties, % of growing households
Know the variety
(awareness)
Ever planted Planted in
Aug/Sep 07
Kabuli
- Areti 43.8 16.71 10.09
- Shasho 48 27.57 20.56
- Chefe 6.43 2.29 1.31
- Ejeri 25 14.29 15.33
Desi
- Marye 3 1 0.94
- Worku 2.71 1.57 1.31
- Akaki 0.71 0.43 0.37
- Dubie 1.86 0.86 0.37
- Local 97.86 92 76.03
29
Table 4. Cropping pattern of chickpea varieties (%)
Varieties Growers (% of total sample farmers)
Chickpea farmers growing the variety
Area allocated (% total chickpea area)
Kabuli
- Areti 7.71 10.09 8.77
- Shasho 15.71 20.56 21.15
- Chefe 1.00 1.31 1.00
- Ejere 11.71 15.33 11.93
Desi
- Improved 3.29 4.30 2.73
- Local 51.29 67.10 54.42
Table 5. On-farm crop yields (kg/ha) by variety, 2006/07
Varieties Gimbichu Lume-Ejere Minjar-Shenkora Total
Areti 2710.0 3538.0 3055.5 3112.6
Shasho 2390.5 2943.3 1666.7 2873.3
Chefe 3175.0 1266.7 2880.0 2272.9
Ejere 1688.9 2702.7 4824.2 2950.2
Improved desi 1814.3 2169.2 1400.0 1960.9
Local 2236.7 2137.1 2065.0 2142.9
30
Fig 4. Farmers' Scoring of Chickpea Varieties (5=Very Good)
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
Yield
Dro
ught
Disea
sePes
t
Ear
lines
s
Pro
d co
st
Uni
form
ity
Col
or
Size
Pric
e
Coo
king
tim
e
Taste
Ove
rall sc
ore
Sc
ore
Areti Shasho Chefe Ejerie Marye Worku Local Desi
31
Table 6. Market participation and marketed surplus
Crops / varieties
Market participation (%)
Marketed surplus (kg)
% sold in main market
Areti 6.29 47.7 7.49
Shasho 16.71 177.31 21.92
Chefe 1 3.91 0.49
Ejeri 7.71 64.79 9.07
Improved Desi 1.29 3.29 0.63
All improved 33 297 39.6
Local Desi 47.43 214.28 35.18
All chickpea 80.43 511.28 74.78
Field pea 44.72 114.48 82.74
Faba bean 45.99 90.67 72.99
Lentil 83.28 426.77 76.46
Grass pea (guaya) 70.78 206.94 68.15