The Impacts of Food Legume Research in the CGIAR:

A Scoping Study

Robert Tripp

7 February 2011

1

1. Introduction

The Standing Panel on Impact Assessment (SPIA) of the CGIAR has expressed concern

that there are few ex-post impact assessments available for the legume crops that are the

mandates of four CGIAR Centres (CIAT, IITA, ICARDA, and ICRISAT). In preparation

for a stripe impact assessment on legumes beginning in 2011, SPIA has commissioned an

initial study of the status of legume uptake and impact in the CGIAR. This paper reviews

the literature, summarizes available documentation on estimated levels of adoption,

discusses a framework for future legume impact studies, and provides suggestions and

possible research themes that could contribute to the wider review.

The paper is based on an analysis of literature found on the shelf, available electronically,

and supplied by correspondents in the CGIAR and elsewhere. The review is fairly

superficial and there is surely additional relevant documentation (e.g. in national

institutions) that would provide further insight. The review was done without in-depth

technical knowledge of many of the disciplines that impinge on legume impact, thus this

review must be considered as simply a first step.

With respect to technology uptake, the focus was on significant instances of change in the

past 15 years. The terms of reference envisaged a review of nine legume species, but this

study discusses information on the uptake of only seven (common bean, cowpea,

groundnut, chickpea, pigeonpea, lentil and faba bean). No relevant information could be

located for grass pea or soybean (despite significant earlier success for the latter crop in

central Nigeria). The terms of reference also contemplated a range of legume technology,

but the majority of the discussion and all of the uptake studies focus on the adoption of

modern varieties (MVs), although crop management implications are discussed.

Following this introduction, Section 2 provides a brief review of some of the factors that

make assessment of legume technology uptake and impact particularly difficult. Section 3

(and the Annex) summarizes the information that is available on the uptake of legume

MVs. Section 4 discusses the factors that should be considered in defining possible

impact pathways for legume technology. Section 5 is a discussion of possible cases and

themes that might be appropriate for impact studies or assessments.

2. Legume Impact

There are several factors that can help explain why the uptake and impact of legume

technology is less well documented than is the case for some other major staples. Some

are related to the relative importance of legumes and hence the absolute contribution of

changes in legume technology and the importance that farmers may accord to

opportunities for innovation. A second set of factors is related to the mechanisms for

promoting legume technology and particularly the limitations of national seed systems for

diffusing new varieties. A third set of factors relates to the way that statistics are collected

about legume technology use.

2

2.1 Legumes in cropping systems

Legumes are vital elements of many cropping systems but they occupy niches that are

varied and subject to modification, substitution or even abandonment. It is impossible to

provide a succinct summary of the various cropping patterns involving legumes around

the world. There are important instances where they are planted as a major sole crop,

often as a component of a rotation, but many more examples involve various types of

intercropping, mixed cropping and relay cropping, taking advantage of complementarities

in growth habits and farm labour profiles. Woolley et al (1991) describe six major classes

of cropping system involving common bean. Beans are mostly intercropped in sub-

Saharan Africa, with the major exception of southern Africa where nearly half are

monocropped (Kimani et al, 2005). Groundnut is often grown as an intercrop in West

Africa but there are parts of Mali where they are mostly grown as a sole crop (Ndjeunga

et al, 2008). Pigeonpea has been traditionally grown as an intercrop in India, but new

short-duration genotypes are included as sole crops in rotations (Ali, 1990). More than

half of the lentil grown in Bangladesh is planted under mixed cropping (Sarker et al,

2004).

Legumes are often seen as insurance or risk-reducing crops, in part because some of them

have deep rooting systems. But many legumes are prostrate growers and sensitive to

competition for light, and the relative abundance of N in the plants makes them

particularly susceptible to insects and diseases (Breman and van Reuler, 2002). Their

susceptibility to biotic and abiotic stresses leads one reviewer to characterize legume

production as ‘unpredictable and unreliable’ and to observe that if legume performance

were the subject of a school report the assessment might read ‘could do better’ (Smartt,

1990:9). Yield improvements in legumes in the past several decades have often been

much less than those for cereals. Pigeonpea and chickpea yields have increased only

modestly (Parthasarathy Rao et al, 2010), while groundnut yields in developing countries

have increased by about 2% per year (Birthal et al, 2010). In the past half-century the

yield of pulses in India has increased by 25%, compared to 211% for cereals (Srivastava

et al, 2010). Legumes’ ability to fix atmospheric nitrogen is an important attribute but it is

dependent on the appropriate Rhizobium bacteria and adequate environmental conditions.

Cowpea is a common intercrop in much of semi-arid West Africa, but it is susceptible to

drought and pests and its total production is only about 10% of the cereals it is grown

with (Mortimore et al, 1997). A study in India found that chickpea yields were more

variable than those of wheat (Kelly and Parthasarathy Rao, 1994, cited in Gowda et al,

2009).

The broad range of legume cropping systems provides valuable flexibility, but also makes

it difficult to target and develop new technology. Different cropping systems often

require different varieties and even within one system farmers may plant several varieties

with different agronomic or food preparation characteristics. Mortimore et al (1997)

found between 5 and 9 different (local) cowpea varieties in the villages they surveyed in

Nigeria; Rwandan bean farmers plant mixtures with an average of five varieties (CIAT,

2008). Legume variety performance will vary under different management regimes and

there has been some attention to the implications for breeding programmes in crops such

3

as cowpea (Blade et al, 1997) and common bean (Woolley and Rodriguez, 1985).

Because legumes are often a ‘secondary’ crop, farmers may not provide the management

attention and inputs accorded to cereals. Farmers are likely to manage intercrops in ways

that favour the cereal; fertilization of a cowpea-cereal intercrop will shift production to

the cereal, for instance (Blade et al, 1997). Much was made of the fact that the advent of

productive, early-maturing wheat and rice varieties led to the displacement of legume

acreage in the Green Revolution (Lipton with Longhurst, 1989).

Legumes make a significant contribution to diets, but are rarely the major focus of

attention. They are often referred to as ‘poor man’s meat’ and with few exceptions direct

legume consumption tends to drop or at best remain stable with increases in income.

(This does not necessarily apply to the use of soybean and other legumes in industrial

food products.) Statistics show bean consumption in Africa remaining relatively constant

over the past 15 years with the exception of significant declines in several countries

where beans are a more important element of the diet (Rwanda, Burundi and Kenya)

(Kalyebara et al 2007). A review found per-capita pulse consumption in developing

countries dropping by 6% from 1980 to 1995; in 1995 pulses contributed 4g (7%) of the

total 56g of protein consumed per capita in developing countries (Kelley et al, 2000).

Although legume protein provides a good balance to cereals in many traditional diets,

there is often a preference for alternatives, particularly animal products. It has been

argued that the traditional Mesoamerican diet based on maize, beans and squash was

brought about by the relative absence of domesticated animals (Albala, 2007).

In summary, while legumes make an important contribution to crop production and local

diets, there are a number of factors that make the development of productive technology,

and farmers’ interest in pursuing such technology, a considerable challenge. Legumes are

subject to various biotic and abiotic stresses and are often managed as complementary

crops. The wide range of cropping systems involving legumes is a testament to farmer

ingenuity, but new technology must either fit the conditions of these specific systems or

be developed by innovative researchers and farmers who are willing to look for new

management options for crops that often receive less attention than grain staples or other

cash crops.

2.2 Technology diffusion and seed systems

Despite the challenges to legume improvement, there have been many successes in

legume breeding. However, these have not always been matched by widespread diffusion

of the new varieties, and legume seed systems play a large role in the explanation.

Although there is some variation, legumes tend to share certain characteristics that make

formal seed production efforts less attractive. Most of the legumes we are examining are

essentially self-pollinating, meaning that farmers can maintain home-saved seed. (The

major exceptions are faba bean and pigeonpea which are partially insect-pollinated.) In

addition, legumes tend to have higher seeding rates and/or lower multiplication rates than

many major cereals, leading to higher seed prices and relatively higher seeding costs,

which again makes farmers favour home-saved seed. Added requirements of threshing

and conditioning some legume crops may further increase the seed cost. One seed

4

technology expert has called groundnut ‘the worst seed crop in the world’ (M. Turner,

personal communication).

These factors explain why the commercial (public or private) seed industry for legumes

has often been slow to emerge in both developing and industrialized countries. Bishaw et

al (2009) surveyed six CWANA countries and found that formal legume seed production

was little more than one percent of that for cereal seed. Our focus is on public legume

varieties and many public research systems are less aggressive than they could be at

promoting new varieties to seed producers. A study in three states of India in the mid-

1990s found that the weighted average age of varieties indented for breeder seed was 15.3

years for groundnut and 12.9 years for chickpea, compared to 9.3 years for wheat and

11.5 years for rice. Two groundnut varieties released in 1940 together accounted for 14%

of demand and the most popular chickpea variety had been released in1960 (Witcombe et

al,1998). (In the US, public bean breeding faces increased competition from the private

sector and is seeing its proportion of the seed market decline (Maredia et al, 2010)).

In addition, many public research systems in developing countries have to face variety

release procedures which do not prioritize legumes. For instance in Kenya the variety

release system charges a fixed fee for a crop trial; maize breeders can share the trial cost

among many varieties (and companies), while bean breeders, with fewer entries, face

higher per-variety costs. The length of the variety release process has led to cases where

the same groundnut variety has been released many years apart in two neighbouring sub-

Saharan African countries. In addition, the location-specificity of some legume varieties

means that a relatively large number of varieties, each with a limited market, may be

competing for attention in the release and promotion processes.

The near absence of formal seed production for many legumes has meant that new

varieties face greater challenges. The idea that ‘a good new variety sells itself’ is only

very partially true. New legume varieties do spread from farmer to farmer (Grisley and

Shamambo, 1993; Kormawa et al, 2004; Alene and Manyong 2006b), and grain markets,

which are a very important seed source for many legumes, can help diffuse a new variety

(Jones et al, 2001; Tripp et al, 1998; David, 1997). But there are limitations on the degree

to which informal diffusion of a new variety can be expected (Sperling and Loevinsohn,

1992; Masangano and Miles, 2004) and additional methods must be used to promote the

technology.

Some of the more successful instances of legume variety diffusion have taken place in the

context of an aggressive seed distribution programme (e.g. Mather et al, 2003). The most

effective modalities for promoting new varieties through seed interventions have yet to be

fully worked out. Some efforts are quite comprehensive, such as the programme to

promote bean seed in sub-Saharan Africa (Rubyogo et al, 2010), while others are more

dispersed. Beye and Margiotta (2008) describe a wide range of donor projects aimed at

promoting the seed of new groundnut varieties in West Africa.

The wide range of seed efforts that still characterize many legume variety promotion

efforts (seed loans, seed villages, seed banks, community seed enterprises, seed exchange

5

programmes, farmer research groups, seed fairs, small seed packs, etc) indicates that we

still need to learn what should be done to make seed available in the most efficient way

possible while promoting the eventual emergence of a sustainable commercial seed

system. This paper is not the place to explore that question, but certainly part of the

reason for lower than expected impact from legume research is the seed bottleneck. At

the very least it may be necessary to acknowledge that in many variety development

situations the budget must include a line for technology promotion through sensible seed

delivery if any reasonable cost/benefit ratio for legume research is to be expected, and

that the costs of seed delivery may need to be considered in impact assessments.

2.3 Legume statistics

A further impediment to the assessment of legume technology diffusion is the availability

of accurate statistics. Legumes often do not receive the same attention in farm surveys

that staple cereals attract, and because legumes are often intercropped or planted in

‘minor’ seasons they may not be accurately reported. Langyintuo et al (2003) review the

various ways that cowpea statistics are collected and reported and Johnson et al (2003)

discuss the problems in finding accurate statistics on beans (at one time FAO had India,

which grows little common bean, as the world’s largest bean producer).

In addition, the large number of legume varieties sometimes makes it difficult to identify

the movement of new varieties. Farmers in Malawi may use different names for the same

bean variety (CIAT, 2007) and in various countries it is not uncommon to find a legume

MV known by the name of its promoter, such as an NGO or extension agent, rather than

by its official name. A recent survey in northern Nigeria found MV names such as

‘World Bank’, ‘Liberia’, and ‘Kanada’ (Ellis-Jones, 2008). In Uganda farmers grew a

groundnut MV developed by a nearby research station but did not know its name or

origin (referring to it as ‘India’) (Longley et al, 2001). As with many other crops, the

distinction between ‘local’ and ‘modern’ varieties is not always recognized by farmers,

making it more difficult to assess variety uptake.

3. The Use of New Legume Technology

The focus of this review is to summarize what is known about the uptake of legume

technology and to use this information to identify opportunities for impact assessment. As

mentioned in the introduction, virtually the only legume technology described in any

detail (or assessed beyond very location-specific situations) is new crop germplasm,

hence the discussion in this section is related entirely to studies on legume MVs. (Section

4 returns to offer some information on crop management.)

There is virtually no recent national-level data on the uptake of legume MVs, nor was it

possible in the course of this review to organize any expert surveys that might estimate

national adoption levels. The most recent multi-country, multi-crop estimates of MV

adoption and impact are found in the studies from the mid- and late-1990s reported in

Evenson and Gollin (2003). This data deficiency (which is not confined to legumes) is

6

one of the reasons for new BMGF projects (DIIVA and TRIVSA) that will develop such

data.

The studies that were available and relevant are summarized in tables in the Annex to this

paper. Many of these were suggested by correspondents at the Centres and they were

supplemented by further literature search. It was only possible to find studies on seven

target legume species. The tables report the most recent studies available and do not try to

follow historical trends. In the cases of common beans, groundnut and lentil some

comparative figures from the earlier Evenson and Gollin (2003) studies are also

presented.

The studies were done for a variety of reasons, under various types of sponsorship. It is

well to be aware of this context, especially when considering further studies, because it is

important to seek synergies and opportunities for collaboration among sources of funding

to promote legume technology. Some of the major past and current efforts that help

document legume technology uptake (project, Centres, crops/regions, donor) include:

� PABRA (Pan-African Bean Research Alliance): CIAT; beans in E and S Africa;

multiple donors.

� PRONAF (Projet Niébé pour l’Afrique): IITA; cowpea in W. Africa ; IFAD.

� PROFRIJOL(Programa Cooperativo Regional de Frijol para Centro América,

México y El Caribe): CIAT; beans in Central America and the Caribbean; SDC.

� Treasure Legumes: ICRISAT; legumes in Ethiopia, Kenya, Malawi, Tanzania;

IFAD.

� Nile Valley Project: ICARDA; legumes in Egypt and Sudan; IFAD.

� Tropical Legumes 2: ICRISAT, CIAT, IITA; beans, cowpea, groundnut, chickpea,

soybean, pigeonpea in Africa and India; BMGF.

� Pulse CRSP: various collaborators; beans and cowpea in Africa and Latin

America; USAID.

� Peanut CRSP: Various collaborators; groundnut in Africa, Latin America and

Asia; USAID.

� DIIVA (Diffusion and Impact of Improved Varieties in Africa): New project to

estimate grain, legume and root and tuber variety uptake and, in some cases,

impact in Africa; BMGF.

� TRIVSA (Tracking Improved Varieties in South Asia): Similar to DIIVA, for

South Asia; BMGF.

The last five of these all have plans for future assessment of variety uptake and are

possible sources of collaboration for some impact studies, although specific funding

commitments are not yet available.

The patterns of legume MV uptake described in the Annex tables are quite varied and no

attempt will be made here to summarize them, although relevant cases are discussed in

subsequent sections of this paper. But some comments are in order on the nature of the

studies that produced the uptake estimates.

7

The studies described in the Annex do not represent a complete accounting of the

adoption data available for legume varieties. They are only the studies suggested by

Centre staff or found in the course of a brief literature search. There are, for instance,

surely a number of other studies available with national research entities. The lack of

country-level, or even more local-level adoption data may not be too surprising. The

organization of a competent adoption study takes time and resources and Centres are

currently pulled in many directions. On the other hand, basic adoption data are necessary

for Centres to monitor how their primary products, new crop varieties, are performing in

the market.

The studies that are reported were carried out for a variety of reasons, but in many cases

they were baselines or assessments required by particular projects. In some cases the

adoption figures reported in the Annex are drawn from documents that summarize several

such studies that are apparently unpublished, and the summaries give relatively little

detail on the nature of the study. Few of the studies provide much information that would

be useful to biological or social scientists who might be interested in refining the

technology or addressing specific bottlenecks, the type of monitoring that should be part

of technology development. Many of the detailed adoption studies appear to see journal

publications as the priority audience and feature analysis that concentrates on parameters

that are found in traditional, academic technology diffusion studies rather than focusing

on a more practical monitoring role. In a few cases the adoption data was used in an

impact assessment.

Many of the studies report relatively high levels of MV uptake and the sample is

obviously biased by the fact that such studies are rarely undertaken in situations where

low adoption rates are expected. Almost none of the studies attempt nationwide estimates

although many cover substantial areas of a country where the crop is most important. No

attempt is made in the tables to specify the details of the sampling, but in many cases the

studies have been done in areas that have profited from substantial technology promotion,

often through the projects sponsoring the study, so some degree of over-estimation is

likely in these cases. The nature of ‘adoption’ is rarely specified, so it is often unclear

what proportion of an adopter’s field is planted to MVs; the area in MVs is also reported

when available. The yield advantage of the MVs is recorded in the tables only when it is

reported in the study, and as explained in Section 4 such estimates may be of limited

relevance in some cases of legume MV uptake.

4. Possible Impact Pathways

4.1 Types of impact

In trying to delineate distinct hypotheses related to the impact of legume technology

(‘impact pathways’) it is useful to consider both the type of technology being examined

and the breadth and nature of its effects.

With respect to technology type, we have already seen that most of the likely current

candidates are based on new germplasm, although there are often important crop

8

management implications. Byerlee (1994) used the distinction between Type A variety

change, when MVs replace older or traditional varieties (typical of the Green Revolution

for rice and wheat); and Type B changes where further generations of MVs are adopted

as part of a more conventional variety improvement and delivery system. This division is

useful and has been employed by a number of other authors, but it has some limitations.

There can be some confusion between chronology and size of impact (earlier

introductions of MVs are not necessarily the ones that have the largest effect). Perhaps

more important, impacts may derive not only from the fact that an MV is high-yielding

but also from its ability to change production patterns (e.g. double cropping or other types

of intensification). Thus the following discussion will not attempt any strict classification

of the legume technology but will instead focus on a range of relatively recent varietal

introductions, some of which also have important crop management consequences.

With respect to the nature of impact, there are many possibilities for classification

(Maredia, 2009). Walker et al (2008) usefully distinguish between Stage 1 and Stage 2

for ex-post impact assessment. Stage 1 is concerned with the economic returns for a given

technology, focusing on the impact on adopters, while Stage 2 extends to broader

‘community-level’ impacts that may include parameters such as consumer prices, market

efficiency, or environmental externalities. The two stages are perhaps best seen as a

continuum rather than separated by an unambiguous dividing line; the scope, depth and

time-scale of an intervention do not necessarily offer opportunities for a simple bipartite

division. In addition, there is the question of whether broader ‘multi-dimensional’

impacts can be legitimately assessed in cases where they were not part of the technology

planning process.

The following discussion will begin by examining the role of new legume varieties in

bringing about changes in crop production and income. It then examines the often

important crop management changes associated with legume variety introduction. The

rest of the section looks at broader impacts, including effects on poverty, market

development, soil quality, other environmental factors, and human nutrition.

4.2 Crop production

Probably the most obvious cases of impact from agricultural technology are those

instances where a more productive variety replaces an older one. The process for legumes

is subject to varying histories. In India and several other Asian countries there is a long

history of public plant breeding. In parts of southern Africa there are instances of well-

established but quite old groundnut MVs from public research (particularly from South

Africa) (Bantilan et al, 2003), while most lentil variety development is fairly recent (Aw-

Hassan and Shideed, 2003).

Although variety replacement is relatively easy to measure, assessing the precise nature

of a new variety’s superiority and impact is not necessarily straightforward. In some

cases the variety may simply be more efficient and higher yielding than the one it

replaces. The groundnut variety released as CG 7 in Malawi and MGV 4 in Zambia is

significantly higher yielding than its competitors although it is no less susceptible to the

major groundnut diseases (ICRISAT, 1994). But there are often other factors that must be

9

examined in assessing yield advantage. In the case of legumes some of the more

important factors are disease or insect resistance, consumer qualities, and the range of

variety use. In addition, crop management changes associated with new varieties may be

important (and are discussed in section 4.4).

Biotic stresses are among the most important yield constraints for many legume species

and are often the major justification for variety change. MVs are generally selected for

their resistance or tolerance to disease and this factor is at times crucial to variety uptake.

Development of resistance to bean golden mosaic virus (BGMV) was responsible for

much of early success in bean breeding in Latin America (Johnson et al 2003); resistance

to the root rots that were causing devastating yield losses in western Kenya encouraged

farmers to adopt new bean MVs (CIAT, 2004); partial resistance to rosette virus

encouraged adoption of groundnut MVs in Uganda (Moyo et al, 2007); several diseases

caused chickpea production to plummet in Nepal and led to an integrated crop

management effort (Pande et al, 2005); disease resistance was a major key to the success

of lentil MVs in Bangladesh (Sarker et al, 2004); developing resistance to ascochyta

blight was one of the key factors in promoting winter chickpea in a number of countries

(Singh et al, 1997). If a disease is endemic and prevalent, the MV’s yield advantage over

a susceptible variety can be estimated with some confidence, but if disease incidence is

sporadic or seasonal the estimation of yield advantage is more complex. In Honduras,

there was significant uptake of BGYMV-resistant MVs although they showed a yield

advantage only in one of the two major bean seasons (Mather et al 2003).

There are also examples where legume breeding has addressed abiotic stress such as

drought. A recently released chickpea variety with drought tolerance (Kusumenglou et al,

2006) has been reported to have withstood a serious drought in Turkey (SeedQuest,

2007). A recently released bean variety in Nicaragua is being promoted on the basis of its

drought tolerance (S. Beebe, personal communication).

Home consumption and market preferences are often important determinants of MV

acceptance. The disease-resistant bean varieties introduced in Honduras were of a size

and colour that resulted in a 10-15% price discount, and hence farmers faced a yield-price

trade-off (Mather et al 2003). Beans are particularly challenging in this regard (David,

1997; Mooney, 2007) and variety preferences may vary within a limited geographical

area. Some haricot bean MVs achieved limited acceptance in Ethiopia because they were

not the preferred colour (Negash, 2007).Cowpea preferences vary across West Africa

(Langyintuo et al, 2003).

Because legumes are often not the major crop in a farming system, they may be managed

in a variety of ways (intercrops, rotations, relay crops, etc) that mean farmers may be

looking for various qualities in MVs and thus a range of variety types needs to be

considered. This is the case for lentil in Bangladesh, for instance (Sarker et al, 2004). In

Rwanda farmers plant a range of bean types, partly as a risk aversion strategy in the face

of a range of biotic and abiotic stress and partly for different consumption goals (Sperling

et al, 1993). The fact that legumes may be managed in various ways, including a

10

significant degree of intercropping, adds to the challenges of estimating the yield

advantage of an MV.

Another important factor in estimating the yield advantage of a legume MV is the

multiple uses for which a legume may be grown. Not only may legumes be harvested and

consumed as dry grain or immature pods, but the leaves of several legume species are

important vegetables. In Kenya, early maturing pigeonpea varieties are mostly grown as

vegetables, while late-maturing types are for dry grain use (Shiferaw et al, 2008). Several

legumes are also important sources of oil and other industrial products.

In addition, legume haulms are often used as fodder, and crop management for a new MV

may be conditioned by the relative importance of food and feed in the farming system.

Some of the semi-erect cowpea varieties that have achieved widespread adoption in

Nigeria are an example of such a dual-purpose crop (Kristjanson et al, 2005). Stalks are

an important by-product of pigeonpea production and are used as fuel.

4.3 Income

The assessment of income changes due to increased legume production needs to take

account of the multiple uses for the crop. It is important to understand the destination of

production when assessing impact. In some cases a legume may be almost completely a

cash crop, such as pigeonpea in northern Tanzania, where an insignificant fraction of the

harvest is retained for household consumption (Shiferaw et al, 2007). In other instances

there is more of an even division between home use and market sale; slightly over half of

pigeonpea production in Kenya goes to home consumption (Shiferaw et al, 2008) while

cowpea farmers in northern Nigeria sell about 60% of their production (Lowenberg-

DeBoer and Ibro, 2008). In other cases a legume is an essential element in the farm

family’s food basket; chickpea farmers in Ethiopia and bean farmers in Rwanda sell less

than one-third of their output (Asfaw et al 2010a, CIAT, 2008a). In the latter cases the

marketed proportion may of course vary significantly among different classes of farm

households in the same production system.

If one is seeking the economic value of the increased production market prices can be

used regardless of end use, but it is important to remember that in many cases legume

prices differ significantly by variety (e.g. CIAT, 2007). Thus market prices may not tell

the whole story in cases where the increased production is particularly important for

household food consumption, and it may be that some measure of increases in available

calories and protein, for instance, may be appropriate.

Even in cases where home consumption of grain (or other products) may be valued at

market prices, an impact assessment may also need to make appropriate assumptions

about the elasticity of demand and its effect on the price of the increased supply. There

may be cases of production increase of a sufficient magnitude to examine consumer price

effects, although it would be important to identify the appropriate counterfactual.

Farmers’ income gains from a new technology must of course include assessment of any

changes in production costs. Even if a variety is simply more responsive, there may be

11

incentives for using additional inputs. Disease- or pest-resistant varieties may allow

savings in pesticides (where they were employed before). On the other hand, some

legume MVs may require additional insecticide, as is the case for the determinate

varieties of cowpea in West Africa (Abatania et al, n.d.; Kristjanson, 2005) and early-

maturing pigeonpea in East Africa (Jones et al, 2002), or the susceptibility of chickpea to

pod borer when it is grown in warmer, non-traditional areas of India (Gowda et al 2009).

Other management changes may also need to be considered, such as the differences

between spreading and erect groundnut varieties for weeding and harvesting labour.

4.4 New cropping systems

The preceding discussion assumed the case of a MV simply replacing an older variety

under more or less the same management regime. Changes in input and labour use are

commonly found in the adoption of MVs for many crops but there are a number of

important instances of legume MVs that are associated with more radical changes in

cropping systems that need to be accounted for in assessing technology impact. The

importance of such cases may be partly due to the fact that legumes are often

accommodated in various niches in farming systems and hence elicit a very broad range

of management practices.

Most of the legume species we are examining have expanded significantly from their

centres of origin, but there are relatively few contemporary cases of the introduction of a

new species as a result of a plant breeding programme. Probably the most important

example is soybean and the rapid growth of that crop in Nigeria the 1990s is well

documented (Sanginga et al, 1999). There are attempts to emulate this success in other

parts of Africa (Chianu et al, 2008a, b),but there are relatively few other examples of

widespread adoption of new species. Pigeonpea production has increased in some parts of

Eastern Africa, mostly as an export crop (Jones et al, 2002). Smartt (1990) discusses the

problem of matching new legume species of high biological potential to consumer

preferences in Africa.

But some of the most important examples of the uptake of legume MVs are associated

with significant changes in production systems.

• In northern Nigeria, the availability of medium-maturity cowpea MVs has brought

significant cropping system changes to a region that has traditionally planted (as

intercrops with cereals) short-maturing cowpea varieties for food and late-

maturing varieties principally for fodder (Mortimore et al, 1997). The MVs have

been planted after the main cropping season on residual moisture as a dual-

purpose crop (Kristjanson et al, 2002) and as part of intensively managed strip-

intercrops with cereals (Ajeigbe at al, 2010; Alene and Manyong, 2007).

• Climbing beans have become increasingly important in Rwanda (Sperling and

Muyaneza, 1995) where they offer significantly higher yields than bush beans (in

part by escaping some of the moisture problems that cause disease) and provide a

way of taking advantage of vertical space on very small farms. But they require

more intensive management, the provision of poles, and better soil fertility.

12

Although this is not a new cropping system (many farmers had previous

experience with climbers), it has expanded in a way that surely requires some

examination of the management correlates.

• There are several examples of significant change in chickpea technology in India,

as the crop shifts away from its traditional production areas in the north of the

country (Gowda et al 2009). The availability of earlier-maturity, wilt-resistant

MVs has allowed the crop to be grown in warmer climates with shorter growing

seasons. Chickpea production has increased significantly in Madhya Pradesh,

often grown on previously fallow land, and as a relatively new crop in Andhra

Pradesh, where it may replace cotton, tobacco or other crops.

• The development of blight-resistant, cold-tolerant chickpea MVs has allowed a

significant expansion of winter-sown chickpea in Syria (Mazid et al, 2009; Singh

et al, 1997). Farmers have adopted the new varieties specifically for winter

sowing and tend to follow somewhat different sowing and plant protection

practices than they use in spring-sown chickpea.

The analysis of the impact of MV adoption associated with significant cropping system

changes presents additional challenges in establishing counterfactuals, but these cases

may offer some of the most outstanding potential examples of recent impact.

4.5 Other management changes

As discussed earlier, this analysis of potential cases of legume impact focuses almost

entirely on germplasm introduction (and often the accompanying crop management

change). This is not because other aspects of legume management have not received

attention from researchers but because it is difficult to find examples of management

change that cover a significant area or number of farmers.

One of the major examples is cowpea storage. Cowpeas are particularly subject to post-

harvest loss due to bruchid damage (Murdock et al, 1997). Various crop storage

techniques have been tested and promoted. Major success appears to have been achieved

in Senegal with the introduction of ‘triple bag storage’, using several layers of thick

plastic bags (Boys et al 2007). The method is currently being promoted in other West

African countries (Moussa et al, 2009).

Insect control for legumes has been the subject of various IPM efforts, but the extent of

uptake has so far been modest. Examples of legume IPM programmes include those for

cowpea in Benin (Nathaniels, 2005), groundnut and cowpea in Uganda (Bonabana-

Wabbi, 2002), pigeonpea in India (Tripp and Ali, 2001), chickpea in Nepal (Pande et al,

2005) and common beans in Nicaragua (Labarta and Swinton, 2005).

There have also been crop management efforts for disease control in chickpea in

Bangladesh (Johansen et al, 2008) and weed control in legumes in WANA (Abang et al,

2007)

13

4.6 Poverty reduction and equity

When we turn to consider broader impacts of agricultural technology uptake probably the

most important task is to examine opportunities for assessing impact on poverty and the

equity of the distribution of benefits. Measures of poverty can be complex (Maredia,

2009) and are not examined here, but presumably most major examples of legume

technology uptake that are the subject of an impact analysis assessing changes in

production and income could include some examination of the poverty reduction or

equity implications of the technology change. Alwang and Siegel (2003) provide a

proposed methodology for assessing the poverty impact of agricultural technology that

uses data from Malawi and show that increases in groundnut production would favour

wealthier households. Moyo et al (2007) do an ex-post impact assessment for groundnut

in Uganda showing a positive impact on poverty indicators, and a study using adoption

from Uganda comes to positive conclusions regarding poverty impact (Kassie et al,

2010). Johnson and Klass (1999) estimated the poverty impact of bean MVs in Honduras

using GIS data. Other instances where legume uptake has been examined with respect to

poverty impact include a study relating households’ assessment of food security to

cowpea technology adoption in northern Nigeria (Alene and Manyong, 2006a), a study

relating food security and nutritional status to soybean uptake in Benue State, Nigeria

(Sanginga et al, 1999) and a preliminary examination of the relation between poverty

status and the use of faba bean technology in Sudan, Ethiopia and Egypt (Amegbeto et al,

2008)..

Gender-related issues are particularly important, especially given the fact that many

legume crops are women’s responsibility. Women’s priorities for variety characteristics

may or may not be addressed in plant breeding programmes (Ashby et al, 1987) and the

results will be reflected in the gender equity of the subsequent variety adoption. Changes

in crop management may have considerable consequences for women. An examination of

adoption of climbing bean MVs in Rwanda showed that women-headed households were

as likely to adopt as male-headed households and that adopters were more likely to be on

smaller farms and among poorer households (Sperling and Muyaneza, 1995). Tipilda et al

(2008) looked at the effect of cowpea MVs in northern Nigeria and concluded that some

of the extra income was passed to women, and that some of the extra yield contributed to

women’s ability to produce and market bean cakes and other cowpea preparations. In

some areas of southern Mali, groundnuts are women’s most important source of income

(Diallo, 2009).

The discussion above on income effects mentioned the possibility, in selected cases, of

looking at the impact on consumer prices and its effect on poorer households. The

importance of this factor would depend on the extent to which legumes are an important

item in the food baskets of non-producing or net-purchasing households.

4.7 Market changes

There may be cases where increased legume production has contributed to market

development. Whether increased trade is a cause or consequence of increased production

would need to be examined, but there are instances where the behaviour of these markets

should be included in an impact assessment. Increased market demand is credited with

14

stimulating the growth in lentil yield and area in Nepal (Neupane et al, n.d.). There is

considerable effort at promoting the domestic cowpea market in Nigeria, including the

development of new commercial food enterprises (Lowenberg-DeBoer and Ibro, 2008).

Examples of important legume export markets include the export of pigeonpea and

chickpea from Africa to India (Jones et al, 2002; Simtowe et al, n.d.), haricot bean export

from Ethiopia (Ferris and Kaganzi, 2008), and regional West African cowpea trade

(Langyintuo, 2003).

4.8 Soil quality contributions

A distinguishing feature of legumes is their ability to fix atmospheric nitrogen through

symbiosis with otherwise free-living soil bacteria (rhizobia). A large proportion of the

plant’s nitrogen needs are provided in this way, although the exact amount depends on

environmental factors and species; for instance common beans are particularly deficient

and may only fix half of their nitrogen requirement (Smartt, 1990). Whatever nitrogen is

fixed represents a ‘saving’ from externally provided nutrient sources (Graham and Vance,

2003), but such savings are not necessarily the basis for impact assessment.

The contribution of a legume to soil fertility depends on the species and on crop

management. Although there have been some studies that propose the transfer of fixed

nitrogen to intercropped cereals, the evidence is thin (Giller, 2001). Most of the nitrogen

contribution of a legume will derive from the amount of plant matter returned to the

system, including roots and nodules, incorporation of crop residues, and the recycling of

nutrients by grazing animals or insects (ibid). The precise contributions are obviously

difficult to measure but there has been a great deal of experimental work (ibid); some

relevant reviews include those for cowpea (Carsky, 2002) and pigeonpea (Rao, 1990).

The key to assessing residual effects is the balance between the nitrogen harvested in

grain (and other plant parts removed from the system) and that left behind. Species such

as soybeans or short-season varieties of cowpea or common bean may remove more

nitrogen than they provide to the soil (Giller, 2001; Smartt, 1990). The impact of a

higher-yielding legume MV on soil nitrogen status will thus depend on its harvest index

and the farmer’s crop residue management and these are subject to various trade-offs. In

the case of pigeonpea in Africa, for instance, the best prospects for soil fertility

contributions come from long-season, indeterminate varieties, while farmers often favour

faster-maturing and higher (grain) yielding varieties (Freeman and Coe, 2002). Other

work has examined the various economic and crop management factors that condition

farmers’ possible use of legumes in rotations or intercrops (Kamanga et al, 2009; Snapp

et al, 2002).

Another factor related to assessing the possible impact of a legume is the cost of the

nitrogen fixation, which is an energy-intensive process (Leigh, 2004), although the

assimilation of externally applied nitrate also requires considerable energy (Giller, 2001;

Stoskopf, 1991). Such energy costs contribute to low yields for legumes compared to

cereals, although it must be remembered that legumes have much higher protein (and

often lipid) content than cereals, requiring extra energy for synthesis (Smartt, 1990). It

must also be remembered that although the legume may provide most of its nitrogen

needs and can make some of the nitrogen available for the succeeding crop, an

15

unfertilized legume crop may remove other nutrients or compete with intercrops (Kimaro

et al, 2009).

So while the nitrogen fixation of legumes makes a valuable contribution to cropping

systems, the extent to which changes in legume germplasm or management would imply

a significant positive impact on soil quality depends on various factors and assumptions.

The simple substitution of a new (e.g. higher yielding or disease resistant) variety would

be fairly unlikely to qualify. If there are significant management changes (new intercrops,

rotations, etc) then there is more possibility of finding a soil-related impact. The case for

seeking impact is greatest where legumes are specifically introduced to a system to

improve soil quality (Mekuria and Waddington, 2002; Ojiem, 2006).

The introduction of legumes to a cropping system may of course make other

contributions to soil quality, such as controlling erosion, reducing water or nutrient loss,

and increasing access to nutrients such as P from deep soil horizons (Giller, 2001;

Shapiro and Sanders, 2002; Adu-Gyamfi et al, 2007). Legume rotations may also play an

important role in weed control (DeGroote et al, 2010). But the measurement of such

impacts related to technology change is most defensible when there has been conscious

design and introduction of the new cropping patterns.

4.9 Other environmental factors

There are other instances of potential environmental impact from changes in legume

technology. These include pesticide management and the fuel wood and related demands

of legumes.

We have already discussed that new disease- or pest-resistant legume varieties may save

on the use of pesticides while some legume MVs may imply additional pesticide (e.g.

when a determinate variety replaces an indeterminate one). In cases where a legume is a

new introduction (as with chickpea in southern India), the pesticide balance depends on

which crops it replaces (e.g. chilli or sorghum in this case).

Many legumes require considerable cooking time, which has significant implications for

fuel wood demands. The importance of this factor varies among species; lentil, for

instance, is valued partly because of its relatively short cooking time (Aw-Hassan and

Shideed, 2003). Cooking time is indeed one of the important factors that farmers consider

in choosing new varieties, although there is not much documentation on the introduction

of MVs specifically for firewood saving. One of the important constraints to the spread of

climbing beans, which are generally much more productive than bush beans, is the

requirement for staking material and the potential environmental consequences.

4.10 Human nutrition

Legumes are widely recognized for their contribution to human diets. The assessment of

protein quality is fraught with difficulties, and while it has been recognized that the lysine

deficiency of most cereal proteins is usefully balanced by legume protein (itself deficient

in sulphur-containing amino acids), legume protein has often been considered of

relatively low quality. But recent revisions in protein quality measures have seen legume

16

protein achieve a higher rating (Messina, 1999). Legumes are also a good source of fibre

and some other nutrients; legumes are important sources of zinc and (sometimes)

calcium, but relatively poor sources of iron. One exception is mungbean, and a study

estimates the nutritional impact of the widespread adoption of mungbean MVs, which

have a higher iron content than local varieties, in Pakistan (Weinberger, 2005). The

presence of oilgosaccharides in many legumes can lead to some digestibility problems

and in some cases there are anti-nutritional factors that may be important (Messina,

1999).

While the focus is on the protein in legume grain, it is well to remember that some

legumes may be eaten green or provide leaf protein. In some cases the leaves may make a

nutritional contribution that rivals that of the grain; it has been estimated that cowpea leaf

production may contribute up to 15 times more protein than the grain from the same field

(Bittenbender et al,1984). The exact balance of the contribution in this case would depend

on how the leaves are harvested (excessive leaf harvest will affect grain yield) and dietary

customs (e.g. cowpea leaves are a more important part of the diet in northern Ghana than

in northern Nigeria). The timing of legume’s contribution to the diet is also important.

For instance, cowpeas and leaves may fill the hunger gap before cereal harvest in West

Africa.

Similar to the case of legumes’ impact on soil nitrogen, the significant contribution of

legumes to dietary protein does not necessarily imply opportunities for assessing the

nutritional impact of changes in legume technology. An increase in legume production

may or may not find its way into the producers’ diets. If it does, then overall dietary

changes during the intervention period would need to be assessed. The best opportunities

would seem to be in situations where legumes are an important part of the diet, or a new

legume (such as soybean) has been introduced. A fairly complete analysis of cropping

and consumption trends would be required. In a contrary example, the Green Revolution

in northern India saw farmers shift away from growing legumes and there was

considerable concern about dietary implications, but Ryan and Asokan (1977) show that

there was a net gain in both dietary calories and protein. Ideally, the demonstration of

nutritional impact of a crop variety intervention should be part of a research programme

in which nutrition was explicitly addressed (e.g. Low et al, 2007).

Another opportunity for examining the nutritional implications of changes in legume

technology is where the shift brings higher income from legume crop sales, assuming the

legume component of income could be clearly identified. Demonstrating relationships

between diet or nutritional status and relatively modest changes in income is often

challenging and must take account of household characteristics as well (Kennedy and

Peters, 1992) so any candidate cases would need to be carefully screened. There may also

be cases where significant increase in legume production has affected the prices paid by

poorer consumers who may be more dependent on legumes as a protein source.

4.11 Summary

Box 1 presents a summary of the factors that have been discussed in this section. The

factors are in only a rough sequence of feasible assessment and complexity of analysis.

17

Box 1. Summary of Potential Impacts of Introduction of New Legume Variety

Yield and yield stability (simple variety substitution) Assessment may need to account for:

� Cases where multiple varieties, uses, cropping patterns � Cases where yield advantage (e.g. from disease resistance) may vary by season,

year � Uses of the harvest (grain, vegetable, fodder, etc)

Income

� Income from occasional sale of excess production or as cash crop � Home-consumed product may be valued at market price, but often considerable

variation by variety (grain, vegetable, fodder) � For fodder value additional animal products may be considered � Varieties of higher grain quality may increase income � Even simple variety substitution may imply changes in crop management costs

(labour, pesticides, etc) Changes related to significant modification in crop management

� Requires a more complete assessment of alternative cropping systems for yield, income

� May imply major changes in labour, input use � May require analysis of use of previously fallow land, a replacement in a rotation,

or entirely new rotation Poverty status

Various parameters might be considered, including: � Measures of household poverty � Distribution of benefits by wealth or social class � Distribution of benefits by gender (within or between households) � Consumer price changes

Agricultural markets

Relation of increased legume supply to: � Domestic markets � Export markets � Food processing industry

Environment

� Soil quality (various parameters) may be affected by increased legume production or (more likely) introduction of a legume to a crop system

� New legume varieties or systems may increase or decrease pesticide use � Other potential impacts (e.g. fuel wood, staking material)

Human nutrition Direct consumption effects may include estimation or measurement of differences in:

� Supply of protein (and perhaps other nutrients) � Dietary consumption � Nutritional status

Indirect effects (related to increased income): � Impact of additional income on household food supply, diet, nutrition

18

The importance of any of these factors will depend on the particular case and the nature

of the cropping system, consumption and marketing changes entailed.

Given that there are relatively few impact assessments of legume technology, and

legumes are rarely the dominant component in a cropping system, it is possible to argue

that initial emphasis should be given to the basic production and income impacts of

legume technology. There may, however, be particular cases where there are

opportunities to examine broader impacts, but the primary impacts would still need to be

carefully measured. Assessment of many of the broader impacts would require fairly

clever specification and estimation of counterfactuals, even when there might be clear

‘adopter’ and ‘non-adopter’ populations available.

Another factor related to the breadth of impact assessed is the extent to which the

outcomes being measured were contemplated in the original research programme. While

it is legitimate to consider unintended consequences of a research programme, when

impact assessment is related to the accountability of the research enterprise it is best to

concentrate on telling a coherent story of planned technological change. The choice of

factors should of course be related to the interests of the audience and hence the

evaluator’s knowledge of the ‘impact of impact studies’. But the choice of documenting

unintended consequences should probably concentrate on the examples where lessons can

be learned to guide future technology development, rather than on presenting some

serendipitous, unplanned outcome as an example of impact.

5. Possible cases for impact assessment

The information presented in this report only provides some preliminary ideas for the

organization of a stripe review on legume impact. Such a review will need to take as

broad a view as possible, assessing major trends in legume technology change; but it may

have to be bolstered by some original research on uptake and impact in specific localities.

These latter cases may be at a regional, national, or sub-national level, depending on the

resources available, the possibilities of confidently sampling from wider trends, and the

value of illustrative case material. Detailed decisions would have to be taken in

consultation with the appropriate Centre and national researchers.

Given the significant resource implications for any extensive impact assessment, it is very

important that the review be planned in conjunction with the various legume-related

projects that are also interested in developing uptake and impact data. DIIVA, for

instance, will be organizing data collection of great value in Africa. For legumes in a

number of countries it will be seeking expert opinion on the uptake of MVs; such data

may provide further guidance on deciding which countries or crops might offer

appropriate case studies and, on the other hand, it may be that a stripe review study could

help to ground-truth such analyses. In a smaller number of cases, DIIVA will support

country-level farm surveys that will assess farmer variety use, and the prospects for using

such data (or even contributing to the questions asked on the survey) for impact

assessment are obvious. It appears that there will only be one case of a DIIVA- sponsored

19

impact assessment on legumes (beans in Rwanda and Uganda), and the stripe review

would of course want to pay close attention to this. There may be similar opportunities

for South Asia with TRIVSA, but plans are less far along in that project. Several other

legume projects at least expect uptake data to be developed on the technologies they

sponsor and collaborative efforts with a stripe review may be feasible in some cases.

The following discussion is organized by legume species (although in some cases

assessments of several legumes in the same environment may be appropriate). Most of

the examples were developed in correspondence with Centre staff but their inclusion here

does not signify any kind of official endorsement. Given the very modest amount of data

currently available on legume impact, and considering the challenges of measuring some

of the broader outcomes sometimes attributed to legume production, it is probably best to

concentrate most effort on the more basic impact parameters related to production,

income and the distribution of that income (i.e. the top rather than the bottom of Box 1).

The choice of research areas and specific cases for a stripe review on legume technology

impact will have to pay attention to a number of themes, including:

• The spread of new varieties. There are many cases where MVs are either

replacing traditional varieties or older MVs in existing production systems. In

some cases the new MVs owe their superiority to a combination of factors, while

in other cases they represent the outcome of research aimed specifically at a

particular biotic threat that had severely reduced legume production.

• Crop management changes. Because legumes are often accommodated in niches

in the principal staple production systems, new legume technology may lead to

significant crop management modification. In most of the cases reviewed, such

change is led by the introduction of a new variety rather than by crop management

research, but there may well be significant instances of the latter. Assessing the

impact of such change is both important and challenging, and it would need to

assign appropriate credit for innovation to research, extension and farmers.

• Subsistence production. In a number of farming systems legumes are an

indispensible item in subsistence production and the description of major

instances of technology change in these cases is important.

• Legume cash crops. There are several cases where market demand (domestic or

export) has contributed to the movement of new legume technology, and

representatives of these cases should be described.

• Delivery mechanisms. Given the exceptional challenges in delivering seed of

legume varieties, an impact assessment should at least keep its eye on cases of

successful, cost-effective legume seed provision.

5.1 Common beans

Annex Table 1a indicates that there has been significant progress made in the diffusion of

bean MVs since the mid-1990s in a number of countries in sub-Saharan Africa. Given the

high levels of poverty and dependence on beans in Rwanda, technological progress there

would be an important example of impact, and it is appropriate that DIIVA is supporting

a full impact assessment in Rwanda (and in Uganda). A stripe review should take account

20

of the results of that analysis, but there is probably not much additional to be done on that

case.

The data in Table 1a show quite high adoption rates for other countries (Kenya, Tanzania,

DR Congo, and especially Malawi), but these figures come from summary documents

and the nature of the surveys that provided this data is not clear. Given that DIIVA will

support expert surveys on variety uptake in all of these countries, it might be best to wait

for those results before deciding whether one or more of these countries deserves

particular attention for impact assessment.

One possible case of interest is beans in Western Kenya. In the late 1980s bean

production in Kenya began to decline because of the spread of root rot. Many farmers

abandoned planting beans. By the mid-1990s five varieties each of bush and climbing

beans resistant to root rot had been developed and introduced (Odendo et al, 2005). Beans

are grown as a sole crop and intercropped with maize. The major change was simply the

substitution of disease-resistant MVs. Estimates for levels of adoption range from 35-

80% in one summary (CIAT, 2004) to 76-100% in another paper (Kalyebara et al, 2008).

The area is apparently confined to Vihiga and Kakamega districts in Western Kenya (W.

Kenya comprises several other districts and grows about 125,000 ha of beans). Although

the area is relatively small, it is an example of efficient research response to a particular

threat. A broader look at beans in Kenya (over 900,000 ha) might also be possible.

Another very specific case that may deserve more attention is the uptake of white beans

in Ethiopia for export markets. If there were a significant interaction between new

varieties and market growth (and data to confirm this are not immediately available) then

it would be an interesting case in a country that is struggling to support market-oriented

agriculture (Ferris and Kaganzi, 2008).

In various African bean cases, impact assessment could look at the relative effectiveness

of seed provision mechanisms for new varieties, a subject that CIAT has thought about

quite carefully (Rubyogo et al, 2010).

Annex Table 1b also shows progress for bean varieties in parts of Latin America. The

Pulse CRSP is currently planning to do expert surveys for a number of countries in LAC

to update MV adoption data and as a basis for some impact assessments. Communication

between the stripe review and CRSP could provide useful synergy for describing the

impact of bean research in the region. It would appear that much of the progress in

Central America is the steady replacement of older MVs by more recent ones.

The instance that is best described is Honduras, which grows approximately 100,000 ha

of beans. All of Central America was subject to the spread of Bean Golden Mosaic Virus

(BGMV) in the 1970s, seriously affecting yields. New BGMV-resistant varieties were

developed and introduced by CIAT, which won the King Badouin Prize for this work in

1984. Studies in the early 1990s indicated increasing adoption rates. An impact study was

done in 2001 (Mather et al, 2003). Since that time newer varieties have been released and

promoted and there is evidence of widespread adoption of some of these (B. Reyes,

21

personal communication). Further analysis would be needed to identify the yield

advantages of the new generation of MVs and to decide how to estimate this. It is not

known if the market price discount of MVs over traditional varieties has been overcome.

The Mather (2003) study reports that a little over half of bean production is marketed. (In

the early 2000s Honduras was essentially self-sufficient in beans and exported some to its

neighbours). There is a large government programme that subsidizes the distribution of

MV bean seed and the costs (and benefits) of this could be included in an impact study.

5.2 Cowpea

Cowpea is largely an African crop and 85% of Africa’s cowpea area is found in Nigeria

and Niger. The annual production of those two countries is over 2 million mt (Langyintuo

et al, 2003). IITA has been involved in developing and promoting new cowpea varieties

with various traits (Singh et al, 2002); erect and semi-erect varieties of medium and early

maturity, some of which are dual purpose (food-feed), seem to be the most important.

The only evidence available on significant adoption (Table 2) comes from Nigeria, and

given that this is the leading producer it would make sense that an impact assessment

should concentrate here (and perhaps across the border in southern Niger). (There are

also instances in West Africa where cowpea is not a traditional crop but cowpea MVs

have been taken up as a cash crop; no recent adoption data were available.)

A recent, informal analysis estimates that 38% of the cowpea in Nigeria is MVs (Alene,

personal communication). Despite a number of adoption studies, there has been no

concerted effort at estimating the impact of cowpea MVs in this region. A precise

inventory of the most important cowpea varieties for northern Nigeria and southern Niger

is not immediately available. There has been a ‘pipeline’ of new varieties, although

apparently some of the MVs promoted since the mid-1990s have only been officially

released in Nigeria in the past few years.

In many cases the new varieties imply changes in management. In some areas there has

been significant promotion of MV use as part of intensive strip intercropping with cereals

(with variations in the details of intercropping by agro-climatic zone), although this is

apparently not relevant in other areas, where relay cropping is practiced. Some of the

varieties are also promoted for dry-season cultivation. There is a tendency for higher

insecticide use on the MVs and many farmers have to contract people to do the spraying.

Cowpea is important both as a cash crop and for home consumption, so impact

assessment might include market parameters or an examination of the growth of cowpea

markets. Northern Nigeria has been the focus of most of the studies, but southern Niger is

also important and exports a large amount of cowpea to Nigeria and other countries.

Studies have shown MV uptake negatively correlated with distance from markets (Alene

and Manyong, 2006a; Kristjanson et al, 2005). There has been one study correlating self-

assessed household food security with MV uptake (Alene and Manyong, 2006a) but the

exact contributions (consumption vs income) are not specified. Fodder use should be

considered as a parameter, but needs further clarification.

22

IITA will try to estimate the use of MVs in Nigeria and Niger under DIIVA (using expert

interviews) and the Pulse CRSP is also interested in this type of assessment. There might

be possibilities to attract support for a more comprehensive study from other cowpea

projects, including Tropical Legumes 2, PRONAF, and the Gatsby Charitable Foundation

(which has been promoting the technology in northern Nigeria and southern Niger).

5.3 Groundnut

Annex Table 3 does not show a large amount of data on which to base decisions for

impact assessment of groundnut. For East Africa, Uganda shows a significant increase in

MV use and the adoption data was used as the basis for an impact study (Kassie et al,

2010). In Southern Africa there has been considerable effort at promoting groundnut

MVs, but no recent uptake data was available for this paper. In West Africa, preliminary

discussion with ICRISAT staff indicated that Mali might be a particularly interesting

example. Groundnut is an important cash crop in Mali and in some areas is women’s

most important income source (Diallo, 2009). New varieties have been tested and

promoted, including some seed production and distribution activities. It is also important

to note that there will be a farm-level groundnut MV adoption study carried out in Mali

under DIIVA. A farm-level survey in Tanzania on several crops for DIVA will include

groundnut.

Groundnut is also an important crop in certain parts of south Asia. Table 3 indicates that

there were a number of adoption studies done in the 1990s, but little recent information is

immediately available. An ICRISAT paper on groundnut in India (Deb et al, 2005) does

not seem to go far beyond the earlier data. The ICRISAT website gives a brief summary

of progress in groundnut in Asia, but little specific detail. Recent baseline studies for the

Tropical Legumes 2 project in Karnataka and Tamil Nadu indicate that older groundnut

varieties still predominate (Lokesha et al, n.d.; Karunakaran and Bantilan, n.d.). ILRI is

currently conducting an impact study on dual purpose groundnut in southern India (N.

Teufel, personal communication).

5.4 Pigeonpea

Pigeonpea is traditionally important in Asia (which grows nearly 90% of the crop), but

there are also several important instances of pigeonpea cultivation in Eastern and

Southern Africa (which accounts for most of the rest of world area).

The adoption studies for pigeonpea (Table 4) are all from Africa. The adoption data do

not show particularly high rates of MV uptake (there is a considerable gap between the

two figures provided for Kenya), but ICRISAT will be estimating pigeonpea MV use in

Kenya, Malawi and Tanzania for DIIVA. Based on that assessment, further decisions can

be made about the feasibility of looking further at impact.

If it appears that there is significant technology uptake, the cash crop aspect in SSA may

deserve specific attention. Pigeonpea is valued for home consumption in many of these

African cases, but perhaps a more interesting focus is the interaction with commercial

markets. A very specific case that might be interesting is northern Tanzania. In Babati

District farmers rank pigeonpea as their most important cash crop. Tanzania exports

23

pigeonpea to India and there has recently been considerable activity to supply a European

export market (R.Jones, personal communication). Tanzania has also exported across the

border to Kenya, which may be more important under the new East African Community.

The major MVs in Tanzania are indeterminate, long-duration varieties (similar to local

varieties) that are resistant to fusarium wilt. They have been introduced and promoted

under various projects and do not imply any significant modification in crop

management. ICRISAT will do a farm-level survey under DIIVA to estimate pigeonpea

MV use in Tanzania (also chickpea, groundnut and sorghum). Economic impact in

northern Tanzania was estimated about 7 years ago (Shiferaw et al, 2007; 2008a).

For Asia, it would probably be best to focus on India, which produces more than 80% of

Asia’s pigeonpea. It was not possible to locate any recent adoption studies on pigeonpea

MVs in India, although there had been earlier success in short-cycle pigeonpea (Bantilan

and Parthasarathy, 1999) and recent innovations such as very short-duration varieties and

the development of hybrid pigeonpea.

5.5 Chickpea

An obvious focus for looking at chickpea impact would be India, which accounts for

nearly two-thirds of the world’s production (5.65 million mt) and imports an additional

0.23 million mt. Two-thirds of it is produced under rainfed conditions. In the past several

decades there has been a significant shift in production from the northern states to central

and southern semi-arid tropic regions. Between 1985 and 2004 chickpea area, production

and yield grew at annual rates of 2.13%, 3.94% and 1.77%, respectively in central and

southern India (Gowda et al, 2009). Much of this shift has been made possible by the

availability of wilt-resistant, earlier- maturing varieties. The increased cultivation of

chickpea has occasioned significant modifications in some cropping systems.

There are a number of MVs (of different grain types and maturities) available for

southern India although a variety released in 1999 seems to maintain dominance

(Parthasarathy Rao, personal communication). Data on current chickpea variety use in

Andhra Pradesh and Karnataka is available in baseline studies done for the Tropical

Legumes 2 project (Suhasini et al, n.d.; Kiresur et al, n.d.) although more interpretation

would be needed to capture the adoption of recent MVs . The adoption of MVs and the

expansion of chickpea area take place in the context of various cropping systems. In some

cases chickpea is planted in previously fallow land but it is also replacing other crops,

including tobacco, cotton and post-rainy season sorghum. In some areas chickpea

cultivation is entirely mechanized. It would take additional enquiry to identify the most

appropriate cropping systems and areas to investigate chickpea impact in India. There is

the possibility that TRIVSA might support an assessment of chickpea MV adoption in

India and chickpea in India is also a focus for the Tropical Legumes 2 Project.

There are several other instances that may deserve attention in considering chickpea

impact. ICARDA has helped promote winter-sown chickpea in Syria, largely through the

availability of disease-resistant, cold-tolerant varieties (Mazid et al, 2009), and the winter

sowing system is also promoted in other Mediterranean countries. Recent reports indicate

very significant uptake of a drought-tolerant chickpea variety in Turkey (SeedQuest,

2007).

24

Although chickpea is not a major crop in Bangladesh, it is one of half-dozen legumes that

have been promoted through a well-organized extension effort demonstrating new

varieties (Afzal, et al (2004); Uddin etal (2005)). It may make sense to look at this effort

across several legumes; lentil (see below) and mungbean are among the most important,

but black gram, grass pea and chickpea have also been promoted.

5.6 Lentil

The major example of documented uptake of lentil MVs is Bangladesh, which grows

about 150,000 ha of lentil and has traditionally imported more than half of its

consumption. Lentil research developed a number of new lentil varieties and evidence

from 2002 showed quite high adoption rates (60% among farmers in blocks where

demonstrations were carried out). The new varieties are generally higher yielding, in part

because of rust resistance and (in one case) blight resistance. Lentil is grown in the post-

rainy season, about half as a sole crop and half intercropped with wheat, oilseeds or other

crops. It does not appear that the MVs require any change in crop management, but they

are found (and preferences among them vary) in a range of cropping systems. There is no

information to date on TRIVSA activities in Bangladesh. An adoption study was carried

out in 2002 (Sarker et al, 2004) and there is an impact analysis based on this data (Aw-

Hassan et al, 2009). As mentioned in section 5.5, Bangladesh has actively promoted a

range of legume species and might be a focus of investigation for several of these.

There may be other cases for lentil. There is a study from Ethiopia (Aw-Hassan et al,

2009) and there has been interest in using it to assess impact, but no further information is

immediately available on the scope or breadth of the technological change. Lentil is the

most important pulse in Nepal and yields and production have increased significantly; no

statistics are available to document the contribution of new lentil technology to that

change but ICARDA varieties are said to be popular (Neupane et al, n.d.).

5.7 Faba bean

Available information on the uptake of faba bean technology is limited to a brief

summary of three country studies done in Egypt, Ethiopia and Sudan examining the

outcome of a regional faba bean programme (Amegbeto et al, 2008). The results show

quite high use of recommended varieties in Egypt and Ethiopia and generally lower

adoption of some of the crop management recommendations. Further information would

be required before identifying any possible cases for additional study.

25

Acknowledgements

It would not have been possible to write this paper without the extensive help provided by

many people who are familiar with legume research. While they are not responsible for

any of the conclusions or interpretations in the paper, they deserve thanks for their

patience in the face of intrusive and often uninformed questioning: Hakeem Ajeigbe,

Dawit Alemu, Arega Alene, Robert Andrade, Aden Aw-Hassan, Cynthia Bantilan, Steve

Beebe, Boris Bravo-Ureta, David Cummins, C.L. Laxmipathi Gowda, Richard Jones,

Alpha Kamara, Genti Kostandini, Mywish Maredia, Jupiter Ndjeunga, Bonny Ntare,

George Norton, P. Parthasarathy Rao, Byron Reyes, Said Silim, Franklin Simtowe,

Louise Sperling, Nils Teufel, Tom Walker, Tim Williams, and Jonathan Woolley. Several

Centres also provided useful comments on an earlier draft of this paper.

26

Annex. Uptake of Legume MVs

The following tables summarize the results of recent studies documenting the uptake of

legume MVs. The tables concentrate on the most recent studies available. Where

possible, comparative data from earlier studies (reported in Evenson and Gollin, 2003)

are provided. (These comparative data are shown in italics.)

1a. Common Beans, sub-Saharan Africa

Location, year Data collection MV adoption Yield

advantage

Source

Ethiopia, Alaba

District 2005

160 farmers ~15% of bean

area

21% (survey

data)

Negash (2007)

Ethiopia, Melkassa and

Awassa

260 farmers 27% Kalyebara et al

(2008)

Ethiopia,

nationwide,1998

Expert survey 8% CIAT-

related

varieties

Johnson et al

(2003)

Kenya, Kakamega

and Vihiga

Districts, 2001

383 farmers 35-80%

(depending on

location,

variety)

81% (survey

data)

CIAT

Highlights 18

(2004);

Kalyebara et al

(2008)

Tanzania, northwestern and

northeastern, 2004

306 farmers, 3

cropping systems

MVs 54% “of

all farmers’

seed”

13% (survey

data)

CIAT

Highlights 42

(2008c);

Kalyebara et al

(2008)

Tanzania,

nationwide,1998

Expert survey 4% CIAT-

related

varieties

Johnson et al

(2003)

Uganda, six

districts, 2003

529 farmers Average 51%

(20%-80%)

41% (survey

data)

CIAT

Highlights 43

(2008b);

Kalyebara et al

(2008)

Uganda, nationwide,1998

Expert survey 15% CIAT-

related

varieties

Johnson et al

(2003)

Rwanda,

nationwide, 2004

383 farmers 94% for

climbing beans

(26% for bush

beans)

10% (climbing

beans)

30% (bush

beans)

(survey data)

CIAT

Highlights 41

(2008a);

Kalyebara et al

(2008)

Rwanda,

nationwide,1998

Expert survey 16% CIAT-

related

varieties

Johnson et al

(2003)

27

DR Congo, Kivu

Province

240 farmers 82% 35% (survey

data)

Kalyebara et al

(2008)

DR Congo, nationwide,1998

Expert survey 48% CIAT-

related

varieties

Johnson et al

(2003)

Malawi,

nationwide

? 93% 2% (survey data) Kalyebara et al

(2008)

Malawi,

nationwide,1998

Expert survey 1% CIAT-

related

varieties

Johnson et al

(2003)

1b. Common Beans, Latin America

Location, year Data collection MV adoption Yield

advantage

Source

Honduras, Mideast and

Northeast Regions,

2001

210 farmers About 40% of

area in MVs

41-50%

(experimental)

Mather et al

(2003)

Honduras, nationwide, 1998

Expert survey 45% (CIAT-

related

varieties)

Johnson et al

(2003)

Central America (6 countries),1998

Expert survey ~39% Johnson et al

(2003)

Mexico, 3 states,

2001

455 farmer

interviews and

postal survey

71%, 42%, 8% Gonzalez-

Ramirez et al

(2005)

Ecudaor,

Imbabura and

Carchi Provinces,

2006

132 farmers For red mottled

beans (which

account for

80% of bean

area), 45%

MVs

40%

(experimental,

under disease

pressure)

Mooney (2007)

Ecuador,

nationwide, 1998

Expert survey 20% (CIAT-

related

varieties)

Johnson et al

(2003)

Colombia,

nationwide, 1998

Expert survey 10% (CIAT-

related

varieties)

Johnson et al

(2003)

Peru, nationwide,

1998

Expert survey 16% (CIAT-

related

varieties)

Johnson et al

(2003)

28

2. Cowpea

Location, year Data collection MV adoption Yield

advantage

Source

Nigeria, Kano and

Jigawa States,

1999

462 farmers in

villages with known

adoption

38% of cowpea

area

~60% grain

yield (crop

models)

Kristjanson et

al (2002, 2005)

Nigeria, Kano and

Kaduna States,

2003/04

480 farmers 72% of farmers

“adopters”;

~80% cowpea

area in MVs

Alene and

Manyong

(2006 a,b;

2007);

Tipilda et al,

n.d.

Nigeria, Bauchi

and Gombe States

130 farmers 56% of farmers

plant an MV

Agwu (2004)

Nigeria, Borno

State, 2007

150 farmers 40% of cowpea

land in MVs

Gadbo and

Amaza (2010)

Senegal, 2004 90 farmers, some

chosen from project

villages

<4% of

farmers

Boys et al

(2007)

Ghana, Northern

and Upper West

Regions, 2007(?)

169 farmers from

project villages

16% of cowpea

area in MVs

Abatania et al

(n.d.)

3. Groundnut

Location, year Data collection MV adoption Yield

advantage

Source

Uganda, 7

districts, 2006

945 farmers, 4

farming systems

53% of

groundnut area

in MVs

Kassie et al

(2010)

Uganda, nationwide, 1999

Expert survey 0% (ICRISAT)

10% (other) Bantilan et al

(2003)

Kenya, Nyanza

and Western

Province, 2006

463 farmers (21%

grew groundnut)

<2% of

farmers use

MV

Nambiro, et al

(2010)

Malawi, 4

districts, 2006/07

594 farmers 26% of farmers

at least 1 MV;

~33% area

Simtowe et al

(2010)

Malawi, nationwide, 1999

Expert survey 10% (ICRISAT)

Bantilan et al

(2003)

Zambia, nationwide, 1999

Expert survey 15% (ICRISAT)

5% (other) Bantilan et al

(2003)

Zimbabwe, nationwide, 1999

Expert survey 2% (ICRISAT)

50% (other)

Bantilan et al

(2003)

29

Mali, Koulikoro

and Kayes

Regions, 2007(?)

343 farmers (in

areas of seed

project)

4 MVs occupy

43% of

groundnut area

23%* Ndjeunga et al

(2008) ; *ICRISAT website

Nigeria, Kano,

Jigawa and

Katsina States,

2007 ?)

470 farmers (in

areas of seed

project)

32% of

groundnut area

in MVs

31%* Ndjeunga et al

(2008) ; *ICRISAT website

Niger, Dosso,

Zinder and Maradi

Regions, 2007( ?)

370 farmers (in

areas of seed

project)

13% of

groundnut area

in MVs

43%* Ndjeunga et al

(2008) ; *ICRISAT website

India, AP and

Maharashtra, 1997

485 farmers,

various localities

4-98% depending

on locality, variety Deb et al, 2005

Nanded (Rabi) 0% (ICRISAT)

94% (other)

Nanded (Kharif) 32% (ICRISAT)

49% (other)

India, 1997

(selected

examples)

Guntur (Kharif) 98% (ICRISAT)

0% (other)

Bantilan et al

(2003)

4. Pigeonpea

Location, year Data collection MV adoption Yield

advantage

Source

Malawi, 4

districts, 2006/07

594 farmers Under 15%

adoption of

MVs ; ~15% of

area

Shiferaw et al

(2009)

Tanzania,

nationwide, 2007

600 farmers 17% adoption

of one MV

Shiferaw et al

(2009)

Tanzania, Babati

District, 2003

240 farmers , but

with some bias

towards those

participating in

promotion

35% farmers

growing new

varieties.

(Estimate 8%

district-wide.)

Shiferaw et al

(2007, 2008a)

Kenya, Mbeere

and Makueni

Districts, 2005

400 farmers 55% of farmers

plant MVs;

51% of

pigeonpea area

in MVs

50-80% Cited in

Shiferaw et al

(2008b)

Kenya, Mbeere

and Makueni

Districts, 2006/07

411 farmers 24% of farmers

plant at least

one MV

Simtowe (n.d.,

draft)

30

5. Chickpea

Location, year Data collection MV adoption Yield

advantage

Source

Ethiopia, 2002 National (CSA)

statistics

<1% Kassie et al

(2009)

Ethiopia, four

districts, 2003

323 farmers 18% of farmers

plant an MV

Dadi et al

(2005)

Ethiopia, three

districts, 2006/07

700 farmers 32% of farmers

planted an

MV; ~33% of

area

Asfaw et al

(2010a,b)

Myanmar, 2007 National statistics

(?)

82% area in

new (<6 years

old) varieties

Than et al

(2007)

Syria, 2004/05 470 farmers

growing winter

chickpea

66% using new

MVs

Mazid et al

(2009)

India, Andhra

Pradesh, 3

districts, 1995

310 farmers 27% of

chickpea area

Joshi, et al

(2005)

India, Andhra

Pradesh, 2

districts, 2007

? 60% of

chickpea area

SAT Trends

No.87 (2008)

6. Lentil

Location, year Data collection MV adoption Yield

advantage

Source

Bangladesh, 13

districts, 2002

250 farmers 63% of block

farmer area

and 24% of

non-block area

in MVs

30% Aw-Hassan et

al, 2009; Sarker

et al 2004

Bangladesh, nationwide,1997

Expert survey 12% 30% Aw-Hassan and

Shideed (2003)

Egypt, nationwide,1997

Expert survey 50% 40% Aw-Hassan and

Shideed (2003)

Pakistan, nationwide,1997

Expert survey 30% 32% Aw-Hassan and

Shideed (2003)

Ethiopia, 4

districts, 2004

289 farmers (10%

of land in lentil)

19% of farmers

adopt MVs

37% Aw-Hassan et

al, 2009

31

7. Faba Bean

Location, year Data collection MV adoption Yield

advantage

Source

Sudan, Dongola

Governorate,

2007(?)

Project participants

and non-

participants, 312

farmers

8% of farmers

plant an MV

8% Amegbeto et al,

2008 ;

ICARDA, 2008

Ethiopia, Arsi

Region, 2007(?)

Project participants

and non-

participants, 198

farmers

71% of farmers

plant an MV

42% Amegbeto et al,

2008 ;

ICARDA, 2008

Egypt, Beni Mazar

District, 2007(?)

Project participants

and non-

participants, 80

farmers

100% of

farmers plant

an MV

18% Amegbeto et al,

2008 ;

ICARDA, 2008

32

References

Abang, M, Bayaa, B, Abu-Irmaileh, B and Yahyaoui, A (2007) ‘A participatory farming

system approach for sustainable broomrape (Orobanche spp.) management in the

Near East and North Africa’, Crop Protection 26: 1723-1732.

Abatania, L, Gyasi, K, Coulibaly, O, Adeoti, R and Salifu, A (n.d.) ‘Adoption and impact

of improved cowpea technologies in Ghana’, Ibadan: IITA/PRONAF.

Adu-Gyamfi, J, Myaka, F, Sakala, W, Odgaard, R, Versterager, J and Høgh-Jensen, H

(2007) ‘Biological nitrogen fixation and nitrogen and phosphorus budgets in

farmer-managed intercrops of maize-pigeonpea in semi-arid southern and eastern

Africa’, Plant and Soil 295: 127-136.

Afzal, M, Hamid, A, Bakr, M, Sarker, A, Erskine, W, Haque, M and Aktar, M (2004)

‘Technology dissemination to boost pulse production and human nutrition in

Bangladesh’, paper presented at the 4th

International Crop Science Congress,

Brisbane, September, 2004.

Agwu, A (2004) ‘Factors influencing adoption of improved cowpea production

technologies in Nigeria’, Journal of International Agricultural and Extension

Education 11: 81-88.

Ajeigbe, H, Singh, B, Ezeaku, I and Adeosun, J (2010) ‘On-farm evaluation of improved

cowpea-cereals cropping systems for crop-livestock farmers: Cereals-cowpea

systems in Sudan savanna zone of Nigeria’, African Journal of Agricultural

Research 5(17): 2297-2304.

Albala, K (2007) Beans. A History, Oxford: Berg.

Alene, A and Manyong, V (2006a) ‘Endogenous technology adoption and household

food security: the case of improved cowpea varieties in northern Nigeria’

Quarterly Journal of International Agriculture 45: 211-230.

Alene, A. and Manyong, V (2006b) ‘Farmer-to-farmer technology diffusion and yield

variation among adopters: the case of improved cowpea in northern Nigeria’,

Agricultural Economics 35: 203-211.

Alene, A and Manyong, V (2007) ‘Gains from high yielding varieties with and without

complimentary technology: The case of improved cowpea in northern Nigeria’,

Journal of Agricultural and Food Economics 2:1-14.

Ali, M (1990) ‘Pigeonpea: Cropping systems’, in Y. Nene, S. Hall and V. Sheila (eds.)

The Pigeonpea, Wallingford, UK: CABI.

33

Alwang, J and Siegel, P (2003) ‘Measuring the impacts of agricultural research on

poverty reduction’, Agricultural Economics 29: 1-14.

Amegbeto, K, Mazid, A, Shideed, K, Legese, G, Bedier, M and Hashim, A (2008)

‘Impacts of crop technologies on productivity, food security and poverty in Nile

Valley Region’ Paper presented at the Ninth International Conference on Dryland

Development, Alexandria, Egypt, November 2008.

Asfaw, S, Shiferaw, B and Simtowe, F (2010a) ‘Does technology adoption promote

commercialization? Evidence from chickpea technologies in Ethiopia’, paper

presented at the CSAE conference, Economic Development in Africa, Oxford,

March 2010.

Asfaw, S, Shiferaw, B, Simtowe, F, Muricho, G, Abate, T and Ferede, S (2010b) ‘Socio-

economic assessment of legume production, farmer technology choice, market

linkages, institutions and poverty in rural Ethiopia’, Nairobi: ICRISAT.

Ashby, J, Quiros, C and Rivera,Y (1987) ‘Farmer participation in on-farm variety trials’,

AgREN Network Paper 22, London: ODI.

Aw-Hassan, A and Shideed, K with Sarker, A, Tutwiler, R and Erskine, W (2003)

‘Economic impact of international and national lentil improvement research in

developing countries’, in R. Evenson and D. Gollin (eds.) Crop Variety

Improvement and its Effect on Productivity, Wallingford, UK: CABI.

Aw-Hassan, A, Regassa, S, Islam, Q and Sarker, A (2009) ‘The impact of improvement

research. The case of Bangladesh and Ethiopia’, in in W. Erskine, F. Muehlbauer,

A. Sarker and B. Sharma (eds.) The Lentil. Botany Production and Uses,

Wallingford, UK: CABI.

Bantilan, M and Parthasarathy, D (1999) Efficiency and Sustainability Gains from

Adoption of Short-Duration Pigeonpea in Nonlegume-Based Cropping Systems,

Patancheru: CRISAT.

Bantilan, M, Deb, U and Nigam, S (2003) ‘Impacts of genetic improvement in

groundnut’, in R. Evenson and D. Gollin (eds.) Crop Variety Improvement and its

Effect on Productivity, Wallingford, UK: CABI.

Beye, A and Margiotta, M (2008) ‘Impact evaluation of the CFC funded project

Groundnuts Seed Systems in West Africa (Phases I and II)’, London: Triple Line

Consulting.

Birthal, P, Parthasarathy Rao, P, Nigam, S, Bantilan, M and Bhagavatula, S (2010)

Groundnut and Soybean Economies in Asia. Facts, Trends, and Outlook,

Patancheru: ICRISAT.

34

Bishaw, Z Makkawi, M and Niane, A (2009) ‘Seed quality and alternative seed delivery

systems’, in W. Erskine, F. Muehlbauer, A. Sarker and B. Sharma (eds.) The

Lentil. Botany Production and Uses, Wallingford, UK: CABI.

Bittenbender, H, Barrett, R and Indire-Lavusa, B (1984) ‘Beans and cowpeas as leaf

vegetables and grain legumes’, Bean/Cowpea CRSP Monograph No. 1, Michigan

State University.

Blade, S, Shetty, S, Terao, T Singh, B(1997) ‘Recent developments in cowpea cropping

research’, in B. Singh, D. Mohan Raj, K. Dashiell, L. Jackai (eds.) Advances in

Cowpea Research, Ibadan: IITA.

Bonabana-Wabbi, J (2002) ‘Assessing factors affecting adoption of agricultural

technologies: The case of integrated pest management (IPM) in Kumi District,

eastern Uganda., unpublished MS thesis, Virginia Polytechnic Institute and State

University.

Boys, K, Fulton, J, Faye, M and Lowenberg-DeBoer, J (2004) ‘Adoption and the

economic impact implications of storage technology and improved cowpea

varieties in the North Central Peanut Basin of Senegal’, Purdue University,

Bean/Cowpea CRSP.

Breman, H and van Reuler, H. (2002) ‘Legumes: when and where an option? (No

panacea for poor tropical West African soils and expensive fertilizers)’, in B.

Vanlauwe, J. Diels, N. Sanginga and R. Merckx (eds) Integrated Plant Nutrient

Management in Sub-Saharan Africa, Wallingford, UK: CABI.

Byerlee, D (1994) Modern Varieties, Productivity, and Sustainability, Mexico, DF:

CIMMYT.

Carsky, R, Vanlauwe, B and Lyasse, O (2002) ‘Cowpea rotation as a resource

management technology for cereals-based systems in the savannas of West

Africa’, in C. Fatokun, S. Tarawali, B. Singh, B. Kormawa and M. Tamó (eds.),

Challenges and Opportunities for Enhancing Sustainable Cowpea Production,

Ibadan: IITA.

Chianu, J, Vanlauwe, B, Mahasi, J, Katungi, E, Akech, C, Mairura, F, Chianu, J and

Sanginga, N (2008 a) ‘Soybean situation and outlook analysis: The case of

Kenya’, Draft report, Tropical Legumes 2.

Chianu, J, Vanlauwe, B, Mayaka, F, Katungi, E, Akech, C, Mairura, F, Chianu, J and

Sanginga, N (2008 a) ‘Soybean situation and outlook analysis: The case of

Tanzania’, Draft report, Tropical Legumes 2.

35

CIAT (2004) ‘Impact of improved bean varieties in Western Kenya’, Highlights. CIAT in

Africa, no.18, Kampala: CIAT.

CIAT (2007) ‘Consumer preferences for beans in Malawi and its influence on price’

Highlights. CIAT in Africa no. 37, Kampala: CIAT.

CIAT (2008a) ‘Farm-level impacts of improved bean varieties and agronomic

technologies in Rwanda’ Highlights. CIAT in Africa no. 41, Kampala: CIAT.

CIAT (2008b) ‘The impact of improved bush bean varieties in Uganda’, Highlights.

CIAT in Africa, no. 43, Kampala: CIAT.

CIAT (2008c) ‘The impact of improved bean production technologies in northern

Tanzania’, Highlights. CIAT in Africa , no. 42, Kampala: CIAT.

Dadi, L, Regassa, S, Fikre, A, Mitiku, D, Gaur, P, Gowda, C and Bantilan, M (2005)

Adoption Studies on Improved Chickpea Varieties in Ethiopia, Addis Ababa and

Patancheru: EIAR and ICRISAT.

David, S (1997) ‘Dissemination and adoption of new technology: A review of

experiences in bean research in Eastern and Central Africa, 1992-1996’ Network

of Bean Research in Africa Occasional Publications Series No. 21, CIAT:

Kampala.

Deb, U, Bantilan, M, and Nigam, S (2005) ‘Impacts of improved groundnut varieties in

India’, in P. Joshi, S. Pal, P. Birthal and M. Bantilan (eds) Impact of Agricultural

Research: Post-Green Revolution Evidence from India, Delhi: NCAP.

DeGroote, H, Vanlauwe, B, Rutto, E, Odhiambo, G, Kanampiu, F and Khan, Z (2010)

‘Economic analysis of different options in integrated pest and soil fertility

management in maize systems of Western Kenya’, Agricultural Economics 41:

471-482.

Diallo, M (2009) ‘Diagnostic study of farmers’ seed quality and variety preferences in

southern Mali’ Unpublished PhD dissertation, Michigan State University.

Ellis-Jones J (2008) ‘A participatory adoption and impact assessment survey of

PROSAB’s activities (2004-2008)’ unpublished draft paper.

Evenson, R and Gollin, D (eds.) (2003) Crop Variety Improvement and its Effect on

Productivity, Wallingford, UK: CABI.

Ferris, S and Kaganzi, E (2008) Evaluating Marketing Opportunities for Haricot Beans

in Ethiopia, Addis Ababa: IPMS/ILRI.

36

Freeman, A and Coe, R (2002) ‘Smallholder farmers’ use of integrated nutrient-

management strategies: Patterns and possibilities in Machakos District of Eastern

Kenya’, in C. Barrett, F. Place and A. Aboud (eds.) Natural Resources

Management in African Agriculture, Wallingford, UK: CABI.

Gadbo, B and Amaza, P (2010) ‘Tobit analysis of improved dual-purpose cowpea in

Damboa, Borno State, North-Eastern Nigeria’, Agricultural Journal 5(4): 248-

252.

Giller, K (2001) Nitrogen Fixation in Tropical Cropping Systems, 2nd

Edition,

Wallingford, UK: CABI.

Gonzalez-Ramirez, H, Bernsten, R, Oehmke, J and Acosta-Gallegos, J. (2005). ‘Adoption

of improved bean varieties in Chihuahua, Durango, and Zacatecas, Mexico’,

Agricultura Técnica en México 31(1): 73-88.

Gowda, C, Parthasarathy Rao, P, Tripathi, S, Gaur, P, and Deshmukh, R (2009)

‘Regional shift in chickpea production in India’, in M. Ali and S. Kumar (eds),

Milestones in Food Legumes Research, Kanpur: IIPR.

Graham, P and Vance, C (2003) ‘Legumes: Importance and constraints to greater use’

Plant Physiology 131: 872-877.

Grisley, W and Shamambo, M (1993) ‘An analysis of the adoption and diffusion of

carioca beans in Zambia resulting from an experimental distribution of seed’,

Experimental Agriculture 29: 379-386.

ICARDA (2008) ‘Fighting poverty: Impact of improved faba bean technologies in

Africa’, ICARDA Impact Brief No.2, Aleppo: ICARDA.

ICRISAT (1994) ‘Groundnut Elite Germplasm ICGV-SM 83708’, ICRISAT Plant

Material Description No. 51, Patancheru: ICRISAT.

ICRISAT (2008) ‘One better than the other’, SATrends Issue 87, ICRISAT.

Johansen, C, Bakr, M, Islam, M, Mondal, N, Afzal, A, MacLeod, W, Pande, S and

Siddique, K (2008) ‘Integrated crop management of chickpea in environments of

Bangladesh prone to Botrytis grey mould’, Field Crops Research 108: 238-249.

Johnson, N and Klass, J (1999) ‘The impact of crop improvement research on rural

poverty. A spatial analysis of BGMV-resistant bean varieties in Honduras’, paper

prepared for the workshop ‘Assessing the Impact of Agricultural Research on

Poverty Alleviation’, San Jose, Costa Rica, September 1999.

37

Johnson, N, Pachico, D and Wortmann, C (2003) ‘The Impact of CIAT’s genetic

improvement research on beans’, in R. Evenson and D. Gollin (eds.) Crop Variety

Improvement and its Effect on Productivity, Wallingford, UK: CABI.

Jones, R, Audi, P and Tripp, R (2001) ‘The role of informal seed systems in

disseminating modern varieties. The example of pigeonpea from a semi-arid area

of Kenya’ Experimental Agriculture 37: 539-548.

Jones, R, Freeman, H and Lo Monaco, G (2002) ‘Improving the access of small farmers

in eastern and southern Africa to global pigeonpea markets’ AgREN Network

Paper No. 120, London: ODI.

Joshi, P, Asokan, M and Bantilan, M (2005) ‘Chickpea in non-traditional areas. Evidence

from Andhra Pradesh’, in P. Joshi, S. Pal, P. Birthal and M. Bantilan (eds) Impact

of Agricultural Research: Post-Green Revolution Evidence from India, Delhi:

NCAP.

Kalyebara, R, Andima, D, Xavery, P, Mugabo, J, Njingulula, P, Ngulu, F, Chirwa, R,

Kirkby, R and Buruchara, R (2007) ‘Impact of bean R&D technologies in sub-

Saharan Africa – evidence from seven countries’ PABRA, Kampala.

Kamanga, B, Waddington, S, Robertson, M and Giller, K (2010) ‘Risk analysis of maize-

legume crop combinations with smallholder farmers varying in respource

endowment in central Malawi’, Experimental Agriculture 46: 1-21.

Karunakaran Rao, G and Bantilan, M (n.d.) ‘Baseline assessment of groundnut for

Tamilnadu State in India’, Draft report, Tropical Legumes 2.

Kassie, M, Shiferaw, B and Murico, G (2010) ‘Adoption and impact of groundnut

varieties on rural poverty. Evidence from rural Uganda’, Discussion Paper Series

10-11, Gothenburg: Environment for Development.

Kelley, T, Parthasarathy Rao, P and Grisko-Kelley, H (2000) ‘The pulse economy in the

mid-1990s: A review of global and regional developments’, in R. Knight (ed.)

Linking Research and Marketing Opportunities for Pulses in the 21st Century,

Dordrecht: Kluwer.

Kennedy, E and Peters P (1992) ‘Household food security and child nutrition: The

interaction of income and gender of household head’, World Development 20:

1077-1085.

Kimani, P, Buruchara, R, Ampofo, K, Pyndji, M, Chirwa, R and Kirkby, R (2005)

‘Breeding beans for smallholder farmers in Eastern, Central and Southern Africa:

Constraints, achievements and potential’, in R. Kirkby (ed.) PABRA Millennium

Workshop, Arusha, May 2001, PABRA.

38

Kimaro, A, Timmer, V, Chamshama, S, Ngaga, Y and Kimaro, D (2009) ‘Competition

between maize and pigeonpea in semi-arid Tanzania: Effect on yields and

nutrition of crops’, Agriculture, Ecosystems and Environment 134: 115-125.

Kiresur, V, Kulkarni, G, Bantilan, M and Kulkarni, V (n.d.) ‘Adoption of chickpea

cultivars in Karnataka: Patterns, trends and constraints’, Draft report, Tropical

Legumes 2.

Kormawa, P, Ezendinma, C and Singh, B (2004) ‘Factors influencing farmer-to-farmer

transfer of an improved cowpea variety in Kano State, Nigeria’, Journal of

Agriculture and Rural Development in the Tropics and Subtropics 105 (1): 1-13.

Kristjanson, P, Tarawali, S, Okike, I, Singh, B, Thornton, P, Manyong, V, Kruska, R and

Hoogenboom, G (2002) Genetically Improved Dual-Purpose Cowpea:

Assessment of Adoption and Impact in the Dry Savannah Region of West Africa,

Nairobi: ILRI.

Kristjanson, P, Okike, O, Tarawali, S, Singh, B and Manyong, V (2005) ‘Farmers’

perceptions of benefits and factors affecting the adoption of improved dual-

purpose cowpea in the dry savannas of Nigeria’ Agricultural Economics 32: 195-

210.

Kusumenoglu, I, Malhotra, R, Aydin, N and Singh, K (2006) ‘Registration of “Gokce” a

kabuli chickpea cultivar’, Crop Science 46: 2703-2704.

Labarta, R and Swinton, S (2005) ‘Do pesticide hazards to human health and beneficial

insects cause or result from IPM adoption? Mixed messages from farmer field

schools in Nicaragua’, paper presented at American Agricultural Economics

Association Annual Meeting, Providence, RI, July 2005.

Langyintuo, A, Lowenberg-DeBoer, J, Faye, M, Lambert, D, Ibro, G, Moussa, B, Kergna,

A, Kushwaha, S, Musa, S and Ntoukam, G (2003) ‘Cowpea supply and demand in

West and Central Africa’, Field Crops Research 82: 215-231.

Leigh, G (2004) The World’s Greatest Fix. A History of Nitrogen and Agriculture,

Oxford: Oxford University Press

Lipton, M with Longhurst, R (1989) New Seeds and Poor People, London: Unwin

Hyman

Lokesha, H, Parthasarathy Rao, P, Nageswara Rao, G and Bantilan, M (n.d.) ‘Baseline

assessment and market survey analysis of the groundnut in Raichur district,

Karnataka’, Draft report, Tropical Legumes 2.

39

Longley, C, Kayobyo, G and Tripp, R (2001) ‘Guidelines for seed production and the

dissemination of improved varieties’ NARO/DFID Client-Oriented Agricultural

Research and Dissemination Project. Draft, London: ODI.

Low, J, Arimond, M, Osman, N, Cunguara, B, Zano, F and Tschirley, D. (2007) ‘A food-

based approach introducing orange-fleshed sweet potatoes increased vitamin A

and serum retinol concentrations in young children in rural Mozambique’ Journal

of Nutrition 137: 1320-1327.

Lowenberg-DeBoer, J and Ibro, G (2008) ‘A study of the cowpea value chain in Kano

State, Nigeria from a pro-poor and gender perspective’, Washington, DC: USAID.

Maredia, M (2009) ‘Improving the proof. Evolution of and emerging trends in impact

assessment methods and approaches in agricultural development’, IFPRI

Discussion Paper 00929, Washington, DC: IFPRI.

Maredia, M, Bernsten, R and Ragasa, C (2010) ‘Returns to public sector plant breeding in

the presence of spill-ins and private goods: the case of bean research in

Michigan’, Agricultural Economics 41: 425-442.

Masangano, C and Miles, C (2004) ‘Factors influencing farmers’ adoption of Kalima

bean (Phaeseolus vulgaris L.) variety in Malawi’, Sustainable Agriculture 24:

117-129.

Mather, D, Bernsten, R, Rosas, J, Ruano, A and Escoto, D (2003) ‘The economic impact

of bean disease resistance research in Honduras’ Agricultural Economics 29: 343-

352.

Mazid, A, Amegbeto, K, Shideed, K and Malhotra, R (2009) Impact of Crop

Improvement and Management. Winter-Sown Chickpea in Syria, Aleppo:

ICARDA.

Mekuria, M and Waddington, S (2002) ‘Initiatives to encourage farmer adoption of soil-

fertility technologies for maize-based cropping systems in Southern Africa’, in C.

Barrett, F. Place and A. Aboud (eds.) Natural Resources Management in African

Agriculture, Wallingford, UK: CABI.

Messina, M (1999) ‘Legumes and soybeans: overview of their nutritional profiles and

health effects’ American Journal of Clinical Nutrition 70 (suppl): 439S-450S.

Mooney, D (2007) ‘The economic impact of disease-resistant bean breeding research in

northern Ecuador’ unpublished MS thesis, Michigan State University.

Mortimore, M, Singh, B, Harris, F and Blade, S (1997) ‘Cowpea in traditional cropping

systems’ in B. Singh, D. Mohan Raj, K. Dashiell, L. Jackai (eds.) Advances in

Cowpea Research, Ibadan: IITA.

40

Moussa, B, Otoo, M, Fulton, J and Lowenberg DeBoer, J (2009) ‘Evaluating the

effectiveness of alternative extension methods. Triple-bag storage of cowpeas by

small-scale farmers in West Africa’, paper presented at the Agricultural and

Applied Economics Association annual meeting, Milwaukee, July 2009.

Moyo, S, Norton, G, Alwang, J, Rhinehart, I and Deom, C (2007) ‘Peanut research and

poverty reduction: Impacts of variety research to control peanut viruses in

Uganda’, American Journal of Agricultural Economics 89: 448-460.

Murdock, L, Shade, R, Kitch, L, Ntoukam, G, Lowenburg-DeBoer, J, Heusing, J, Moar,

W, Chambliss, O, Endondo, C and Wolfson, J (1985) Postharvest storage of

cowpea in sub-Saharan Africa’, in B. Singh, D. Mohan Raj, K. Dashiell, L. Jackai

(eds.) Advances in Cowpea Research, Ibadan: IITA.

Musoni, A, Buruchara, R, and Kimani, P (2005) ‘Climbing beans in Rwanda:

Development, impact, and challenges’, in R. Kirkby (ed.) PABRA Millennium

Workshop, Arusha, May 2001, PABRA.

Nambiro, E, Mulwa, R, Njuguna, E, Murithi, F, Naranja, T, Kamau, M and Simtowe, F

(2010) ‘Socioeconomic assessment of groundnut production conditions, farmer

technology choice, market linkages, institutions and poverty in Nyanza and

Western provinces of Kenya’, Nairobi: ICRISAT.

Nathaniels, N (2005) ‘Cowpea, farmer field schools and farmer-to-farmer extension: A

Benin case study’ AgREN Network Paper No. 148, London: ODI.

Ndjeunga, J, Ntare, B, Waliyar, F, Echekwu, C, Kodio, O, Kapran, I, Diallo, A, Amadou,

A, Bissala, H and Da Sylva, A (2008) ‘Early adoption of modern groundnut

varieties in West Africa’, Socioeconomics and Policy Program Working Paper

24, Patancheru: ICRISAT.

Negash, R (2007) ‘Determinants of adoption of haricot bean production package in Alaba

Special Woreda, southern Ethiopia’, unpublished MSc thesis, Haramaya

University.

Neupane, R, Sarker, A, Shrestra, R, Yadav, N, Erskine, W and Francis, C (n.d.) ‘Lentil

production in Nepal’ Bangkok: APAARI.

Odendo, M, David, S and Otsyula, R (2005) ‘Impact of root-rot resistant bean varieties

in Western Kenya: Application of impact diagramming’, in R. Kirkby (ed.)

PABRA Millennium Workshop, Arusha, May 2001, PABRA.

Ojiem, J (2006) ‘Exploring socio-ecological niches for legumes in western Kenya

smallholder farming systems’ unpublished PhD thesis, Wageningen University.

41

Pande, S, Neupane, R, Stevenson, P, Grzywacz, D, Bourai, V, Rao, J and Kishore, G

(2005) ‘Integrated crop management of chickpea in Nepal: past, present and

future’ in S. Pande, P.Stevenson, R. Neupane and D. Grzywacz (eds) Policy and

strategy for Increasing Income and Food Security through Improved Crop

Management of Chickpea in Rice Fallows in South Asia, Proceedings of NARC-

ICRISAT-NRI Workshop, Patancheru: ICRISAT.

Parthasarathy Rao, P, Birthal, P, Bhagavatula, S, and Bantilan, M (2010) Chickpea and

Pigeonpea Economies in Asia. Facts, Trends and Outlook, Patancheru: ICRISAT.

Rao, J (1990) ‘Pigeonpea: Nitrogen fixation’, in Y. Nene, S. Hall and V. Sheila (eds.) The

Pigeonpea, Wallingford, UK: CABI

Rubyogo, J-C, Sperling, L, Muthoni, R, and Buruchara, R (2010) ‘Bean seed delivery for

small farmers in sub-Saharan Africa. The power of partnerships’, Society and

Natural Resources 23: 285-302.

Ryan, J and Asokan, M(1977) ‘Effect of Green Revolution in wheat on production of

pulses and nutrients in India’ Economics Program Occasional Paper 18,

ICRISAT: Hyderabad.

Sanginga, P, Adesina, A, Manyong, V, Ofife, O and Dashiell, K (1999) Social Impact of

Soybean in Nigeria’s Southern Guinea Savannah, Ibadan: IITA.

Sarker, A, Bakr, M, Afzal, M, Erskine, W, Rahman, M and Saxena, M (2004) Lentil

Improvement in Bangladesh, Bangkok: APAARI.

SeedQuest (2007) ‘New kabuli chickpea variety survives drought in Turkey’ SeedQuest

http://www.seedquest.com/News/releases/2007/august/20146.htm.

Shapiro, B and Sanders, J (2002) ‘Natural resource technologies for semi-arid regions of

sub-Saharan Africa’, in C. Barrett, F. Place and A. Aboud (eds.) Natural

Resources Management in African Agriculture, Wallingford, UK: CABI.

Shiferaw, B Silim, S, Muricho, G, Audi, P, Mligo, J, Lyimo, S, You, L and Christiansen,

J (2007) ‘Assessment of the adoption and impact of improved pigeonpea varieties

in Tanzania’, SAT eJournal 5(1): 1-27.

Shiferaw, B, Kebede, T and You, L (2008a) ‘Technology adoption under seed access

constraints and the economic impacts of improved pigeonpea varieties in

Tanzania’ Agricultural Economics 39:309-323.

Shiferaw, B, Okello, J, Muricho, G, Omiti, J, Silim, S and Jones, R (2008b) Unlocking

the Potential of High-Value Legumes in the Semi-Arid Regions. Analyses of the

Pigeonpea Value Chain in Kenya, Nairobi: ICRISAT.

42

Shiferaw, B, Asfaw, S, Simtowe, F, Lyimo, S, Silim, S, Purnachandra Rao, K,

Parthasarathy Rao, P, Bantilan, C, Chopde, V, Padmaja, R, Kavitha, K,

Deshmukh, R, Marawar, S and Sunandini, G (2009) ‘Objective 1: Targeting

pigeonpea breeding and seed delivery efforts to enhance the impact on the

livelihoods of the poor in drought-prone areas of eastern and southern Africa and

South Asia’, in Lessons from Two Years of Tropical Legumes 2, Nairobi:

ICRISAT.

Simtowe, F, Asfaw, S, Diagne, A and Shiferaw, B (2010) ‘Determinants of agricultural

technology adoption: the case of improved groundnut varieties in Malawi’,

contributed paper at the meeting of African Association of Agricultural

Economists, Capetown, September 2010.

Simtowe, F, Shiferaw, B, Kassie, M, Abate, T, Silim, S, Siambi, M, Madzonga, O,

Muricho, G and Kananji, G (n.d.) ‘Assessment of the current situation and future

outlooks for the pigeonpea sub-sector in Malawi’, Nairobi: ICRISAT.

Simtowe, F (n.d.) ‘Determinants of agricultural technology adoption: The case of

improved pigeonpea varieties in Kenya’, draft paper, Nairobi.

Singh, K, Malhotra, R, Saxena, M and Bejiga, G (1997) ‘Superiority of winter sowing

over traditional spring sowing of chickpea in the Mediterranean region’,

Agronomy Journal 89: 112-118.

Smartt, J (1990) Grain Legumes. Evolution and Genetic Resources, Cambridge:

Cambridge University Press.

Snapp, S, Rohrbach, D, Simtowe, F and Freeman H (2002) ‘Sustainable soil management

options for Malawi: can smallholder farmers grow more legumes?’, Agriculture,

Ecosystems and Environment 91: 159-174.

Sperling, L and Loevinsohn, M (1993) ‘The dynamics of adoption: Distribution and

mortality of bean varieties among small farmers in Rwanda’ Agricultural Systems

41: 441-453.

Sperling, L, Loevinsohn, M and Ntabomvura, B (1993) ‘Rethinking the farmer’s role in

plant breeding: Local bean experts and on-station selection in Rwanda’,

Experimental Agriculture 29: 509-519.

Sperling, L and Muyaneza, S (1995) ‘Intensifying production among smallholder

farmers: the impact of improved climbing beans in Rwanda’, African Crop

Science Journal 3: 117-125.

Srivastava, S, Sivaramane, N, Mathur, V (2010) ‘Diagnosis of pulses performance in

India’, Agricultural Economics Research Review 23: 137-148.

43

Stoskopf, N (1991) Understanding Crop Production, Reston, VA: Prentice-Hall.

Suhasini, P, Kiresur, V, Rao, G and Bantilan, M (2009) ‘Adoption of chickpea cultivars

in Andhra Pradesh: Patterns, trends and constraints’ Draft report, Tropical

Legumes 2.

Than, A, Maw, J, Aung, T, Gaur, P and Gowda, C (2007) ‘Development and adoption of

improved chickpea varieties in Myanmar’ SAT eJournal 5 (1):1-3.

Tipilda, A, Alelne, A andManyong, V (2008) ‘Engaging with cultural practices in ways

that benefit women in northern Nigeria’, Development in Practice 18: 551-563.

Tripp, R, Walker, D, Opoku-Apau, A, Dankyi, A and Delimini, L (1998) Seed

Management by Small-Scale Farmers in Ghana, Wallingford, UK: Natural

Resources Institute.

Tripp, R and Ali, A (2001) ‘Farmers’ access to natural pest control products: Experience

from an IPM project in India’, AgREN Network Paper No. 113, London: ODI.

Uddin, M, Ali, M, Rahman, M (2005) Prospects of chickpea in rice-based cropping

systems of Bangladesh’ in S. Pande, P.Stevenson, R. Neupane and D. Grzywacz

(eds) Policy and strategy for Increasing Income and Food Security through

Improved Crop Management of Chickpea in Rice Fallows in South Asia,

Proceedings of NARC-ICRISAT-NRI Workshop, Patancheru: ICRISAT.

Walker, T, Maredia, M, Kelly, T, La Rovere, R, Templeton, D, Thiele, G and

Douthwaite, B (2008) Strategic Guidance for Ex-Post Impact Assessment of

Agricultural Research, Report prepared for SPIA, CGIAR Science Council,

Rome.

Weinberger, K (2005) ‘Assessment of the nutritional impact of agricultural research: The

case of mungbean in Pakistan’, Food and Nutrition Bulletin 26(3): 287-294.

Witcombe, J, Packwood, A, Raj, A. and Virk, D (1998) ‘The extent and rate of adoption

of modern cultivars in India’ in J. Witcombe, D. Virk and J. Farrington (eds)

Seeds of Choice. Making the Most of New Varieties for Small Farmers, New

Delhi: Oxford and IBH.

Woolley, J, Ildefonso, R, Portes e Castro, T and Voss, J (1991) ‘Bean cropping systems

in the tropics and subtropics and their determinants’, in A. van Schoonhoven and

O. Voysest (eds.) Common Beans. Research for Crop Improvement, Wallingford,

UK: CABI.

Woolley, J and Rodríguez, W (1987) ‘Cultivar x cropping system interactions in relay

and row intercropping of bush beans with different maize plant types’,

Experimental Agriculture 23: 181-192.