Research and development for sorghum and millets in Sub-Saharan Africa: What have we learned?

Abstract

The past two decades of R & D for sorghum and millets in SSA have generated a wealth of new evidence. A synthesis of this evidence identified six strategic lessons. These were that demand is not being driven by ‘new uses’, that plant breeding programmes need to respond to African farmer's contexts and objectives to increase impact, that availability is a bigger constraint on the supply of certified seed than physical or economic access, that higher adoption of improved varieties does not generally result in higher yields, that technology and varietal diversity can compensate for climate change, and that commercialisation can be gender – and socially inclusive. R & D in Africa differs from the Indian model because of the low level of commercialisation and the continued dominance of own consumption. Consequently R & D gives a dominant role to the public sector in plant breeding, to farmer organisations and groups in seed supply, and to prioritising household food security over commercial uses.

Keywords

Sorghum millets Sub-Saharan Africa research and development learning

Introduction

Our knowledge of smallholder agriculture in Sub-Saharan Africa (SSA) is rife with myths and misconceptions. Recent examples include the myths that women provide most of the labour for crop production (Palacios-Lopez et al., 2017), that most farmers suffer high post-harvest losses (Christiaensen and Demery, 2018), or that women take better care of the environment (Doss et al., 2018). Myths legitimise ‘development narratives’ that frame the design of agricultural development programmes (Leach and Mearns, 1996). Myths arise for many reasons but one reason they persist is the lack of evidence to refute them (Chambers, 2017). Some long-standing myths have been overthrown only with the advent of nationally - representative household surveys (Christiaensen and Demery, 2018). In this context, learning assumes a double significance: learning is not just about discovering what was previously unknown but also about un-learning what was previously accepted as knowledge.

A similar process of myth – making is evident with sorghum and millets. Commercialisation has become the reigning orthodoxy. In SSA, this approach has a long history. Market liberalisation from the 1980s onwards created new opportunities for the commercialisation of staple food crops, by weakening state control over pricing and marketing. ICRISAT's experience with the Sorghum and Millets improvement Programme (SMIP) in Eastern and Southern Africa (ESA) identified ‘market demand’ as the key driver for the adoption of fertiliser and improved varieties (Alumira and Heinrich, 2003). Finally, the fruits of commercialisation are highlighted by experience in India. Here, plant breeding operates through the Hybrid Parents Research Consortium, a public-private partnership where private companies use plant materials from the public sector to develop their own varieties. Three quarters of certified seed¹ is produced and sold by private firms (Pray and Nagarajan, 2009). Sorghum and millets have become cash crops (Kumara Charyulu et al., 2016a,b). Pearl millet grain now feeds the poultry industry, while fodder from rainy season sorghum feeds the dairy industry. Alongside the growth of these new markets, though, there has been a decline in demand for human consumption (Parthasarthy Rao and Basavaraj, 2015). New markets are therefore needed to ‘save the crop’ (Walker, 2010: 70). In ICRISAT's latest Strategic Plan, the future of sorghum and millets is seen to lie with market-led development (ICRISAT, 2010).²

Yet India's importance for sorghum and millets is fading. Africa has overtaken Asia as the world's’ biggest producer of both crops. Sorghum and millets are grown in both West Central Africa (WCA) and ESA.³ Production is concentrated in WCA, which accounts for 90% of total production of both crops (Nedumaran et al., 2017). Since 1980, Africa's production of sorghum has tripled (Dalton and Hodjo, 2020) while the production of millet has doubled (Meena et al., 2021). This gives R & D a strong foundation. The business case for R & D is compelling, with high returns on investment (Orr et al., 2020; Zereyesus and Dalton, 2017). The only question to answer is, what sort of R & D programme should it be?

We argue that Africa does not generally fit this model of commercialisation. True, there is potential for new markets. But demand is overwhelmingly for home consumption. Plant breeding still depends on the public sector. In ESA, improved varieties are not hybrids but open-pollinated varieties (OPVs).⁴ Commercial companies play a minimal role in the supply of certified seed. This calls for a different kind of R & D programme, based not on a universal model of market-led development but designed to fit the local context.

The general objective of this article is to analyse six strategic lessons from this programme. The specific objectives are to:

Identify some key lessons;

Synthesise the relevant evidence; and

Outline some strategic implications for R & D.

The aim of the article is not to present new research findings but to synthesise existing knowledge. The article is intended to provide a tour d’horizon for the general reader and guidance for future R & D programmes.

The article focuses on the period 2000–2021. This period saw the Harnessing Opportunities for Productivity Enhancement (HOPE) project, funded by the Bill & Melinda Gates Foundation. This 12-year project (2010–22) was led by the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT). The period also saw the Research Program on Dryland Cereals (2012–2076), later the Research Program on Grain Legumes and Dryland Cereals Agri-food Systems (2018–2022), both led by ICRISAT and managed by the Consultative Group on International Agricultural Research (CGIAR). ICRISAT's R & D in SSA is conducted with numerous partners, including National Agricultural Research and Extension Systems (NARES) Non-Governmental Organisations (NGOs), farmer organisations and private firms. Without the complementary resources and skills provided by these partners, ICRISAT's R & D programme would not be possible. Similarly, almost all the publications from this programme were written jointly by scientists from the NARES, ICRISAT and, or other international research institutions. For ‘ICRISAT’, therefore, read ‘ICRISAT and partners’. While relying primarily on the evidence from these programmes, the article also uses evidence from other sources. Because the article covers a wide range of topics, only the most pertinent evidence was selected for each lesson. Detailed statistical evidence has been deliberately excluded in order to highlight the general analysis.

Two caveats are needed. Africa has great diversity in agro-ecologies, farming systems, and state policies. This makes it unwise to generalise across the continent, or even the region. In presenting evidence, therefore, the country or countries to which it relates has been specified. This diversity means that there is varying potential for the intensification of crop production and the commercialisation of agriculture (Jayne et al., 2019). However, exploring how this diversity determines the constraints and opportunities for sorghum and millets is beyond the scope of this article. Although the article notes variations in the results of R & D between countries and regions, it does not attempt to explain them.

Lessons

Table 1 summarises the six lessons. For each lesson we list the original assumption, then the ‘new learning’ produced by R & D, and finally the implications. We use ‘lessons’ in a broad sense to include not just new knowledge, but also the re-affirmation of old truths (Lesson 4) and what might be better described as working hypotheses (Lessons 5 and 6). For reasons of space, this list is not exhaustive and there are some notable omissions, such as policy.

Table 1.

Strategic lessons for R & D for sorghum and millets in SSA.

No.	Initial assumption/ expectation	‘New learning’	Comments	Implications for R & D
1	‘New uses’ will drive demand	'New uses' are not driving demand	Middle-class markets for Smart Foods and clear sorghum beer	Dietary changes required will take time
2	Breeders in SSA should create Green Revolution types of varieties	Plant breeding must adapt to African farmers’ contexts and objectives	Farmers adopt new varieties offering advantages for their objectives and growing conditions	Focus on varietal adaptation and use
2				Farmer- breeder collaboration to guide and implement breeding
3	The seed system is the chief constraint on adoption of new varieties	Availability (supply) is a bigger constraint than access (effective demand)	Access in West Africa determined by cultural factors	Different seed systems for WCA and ESA
4	Adoption of improved varieties will raise yields	High adoption without higher yields	Farmers adopt improved varieties that offer new options	Prioritise varieties with advantage(s) in specific contexts
4	Adoption of improved varieties will raise yields	High adoption without higher yields	Ceiling on adoption of improved varieties	Develop a wide range of varieties with attention to intra-varietal diversity
5	Climate change will reduce yields	Technology can compensate for yield losses	Yield losses are small	Incentives for adoption may be missing
6	Commercialisation will exclude poorer farmers and disempower women	Commercialisation can be gender- and socially - inclusive	Value chain for sorghum beer dominated by big farms, value chain for finger millet in ESA involves small farms	Information and small seed packs boost inclusion
			Commercialisation in ESA benefits women	Groups help women keep control
			Commercialisation in ESA benefits women	More evidence needed on gender impacts

Source: Summarised from text.

Lesson # 1: ‘New uses’ are not driving demand

Aggregate or total demand for sorghum and millet in SAA has risen thanks to population growth. This is reflected in rising production. Yet in SSA, as in India, average or per capita demand has fallen.

There are three reasons. One is urbanisation. In ESA (Ethiopia, Kenya, Tanzania, and Uganda), consumption of sorghum was three times lower in urban areas (7 kg/capita/year) than in rural areas (30/kg/capita/year). By contrast, the rural-urban divide was relatively small for millets: 4 kg/capita/year in urban areas compared to 7 kg/capita/year in rural areas (Orr et al., 2020). Similarly, in WCA (Nigeria and Niger) the average consumption of sorghum and millets was lower in urban areas (Cheng and Larochelle, 2016). Since by mid-century 58% of the population in SSA will be urban (United Nations, 2019a) average consumption will continue to fall.

A second reason is rising income. Except in Ethiopia, demand for sorghum in ESA was lower among higher-income households (Orr et al., 2020). The same was true for sorghum in Nigeria and Niger (Cheng and Larochelle, 2016). In the case of millets, the picture was mixed. Average consumption of millets rose with income in ESA (Orr et al., 2020) and in Niger though not in Nigeria (Cheng and Larochelle, 2016). In general, therefore, urbanisation has reduced the demand for both sorghum and millets while higher income reduced demand only for sorghum.

The third reason is competition from other cereal crops. In Kenya, Tanzania, and Uganda the shares of sorghum and millets in household expenditure on cereals ranged from just 1–5% (Gierend and Orr, 2015), which is too low for R & D to have much impact on poverty and nutrition at the national level. There is greater scope in Ethiopia, where the share of sorghum in food expenditure was 11% (Gierend and Orr, 2015), The greatest scope is in WCA. In Nigeria, although sorghum accounts for only 3% of national household food expenditure (rising to 9% in the North-West). its large population means a big growth in demand (Cheng and Larochelle, 2016). The share of household expenditure on millets is highest in Niger, at 25% (Cheng and Larochelle, 2016).

Yes, there are ‘new uses’. The most prominent is clear sorghum beer – to be distinguished from traditional brews, which are opaque. First produced in Nigeria in 1988, lager-type sorghum beer is now produced in Ghana, Sierra Leone, Kenya, Uganda and Zambia. Other uses of sorghum include starch and livestock feed. In ESA, finger millet porridge has long been used as a weaning food. ICRISAT has promoted sorghum and millets as ‘Smart Food’ for school meals and as health-foods targeted at urban, middle-class consumers who are most at risk from lifestyle diseases such as diabetes (Orr et al., 2020). Similarly, the Sorghum and Millets lnnovation Lab (SMIL) has developed sorghum blends to make wheat-based products in Senegal and Niger and for teff-based injera in Ethiopia (Feed the Future, 2021). These products are still at the experimental stage and production remains small-scale. If widely accepted, they would boost consumer demand for ‘new uses’.

At present, however, ‘new uses’ account for a modest share of total production. Sorghum beer in Nigeria – with the largest brewing industry outside South Africa – uses only 1.4% of the entire sorghum harvest (Akinyoade et al., 2020). For WCA, no reliable estimate of the scale of ‘new uses’ is available.⁵ For ESA (Ethiopia, Kenya, Tanzania, and Uganda), the three major value chains – livestock feed, flour, and beer – utilised just 6% of the total production of sorghum and 11% of millets in 2015. Under favourable assumptions, by 2025 utilisation might rise to 15% for sorghum and 12% for millets (Orr et al., 2017). Admittedly, there is potential demand. The primary driver is consumer demand by Africa's emerging middle class. In 2010 there were 137 million households with per capita incomes of $ 4–20 per day (Lufumpa et al., 2015). Although just 13% of the total population, they account for a disproportionate share of consumer expenditure, and much of their expenditure on food goes on processed products (Tschirley et al., 2015). These ‘aspirational’ consumers are the natural target market for ‘new uses’ of sorghum and millets.

Demand for ‘new uses’ is constrained by four factors, however. First, utilisation as livestock and poultry feed in both regions is limited by the relatively high price of sorghum – 20% higher than maize in ESA (Orr et al., 2017). Similarly, evidence from ESA shows that millets are too expensive for school-feeding programmes (Wangari et al., 2020). Increasing consumer demand for millets would require a significant reduction in price, which would in turn require higher yields. A more realistic alternative is to increase demand (and willingness to pay) by educating consumers about the nutritional benefits of these crops (Revoredo-Giha et al., 2022). Market research in Niger has shown that middle-class urban consumers will pay a quality premium for pearl millet dèguè (Nakelse and Dalton, 2018). Second, many consumers are unaware of the health benefits of these crops. In ESA, one in three consumers already buying sorghum or millets was unaware of these benefits; among non-consumers, the ratio was two in three (Orr et al., 2020). Third, supply may be limited in drought years when (as happened in Kenya in 2011) smallholders refused to sell sorghum to East African Breweries (Orr et al., 2020). This is not a problem in Nigeria, where sorghum supplied to Nigerian Breweries comes mainly from medium and large farms (Akinyoade et al., 2020). Finally, there is government policy. Sorghum beer in Nigeria owes its existence to a ban on imported barley (Akinyoade et al., 2016), while in Kenya and Zambia its survival depends on low excise duty. The biggest threat to the future of this industry is Inconsistent policy (Akinyoade et al., 2016; Orr, 2018b),

Demand will therefore be driven primarily by home consumption. Although demand per head may decline, total demand will grow thanks to population growth. By mid-century, population in SSA will top 2 billion (United Nations, 2019b). Demand for sorghum is forecast to rise from 41 million t in 2025 to 57 million t by 2040, while demand for millets is forecast to rise from 27 to 38 million t (Nedumaran et al., 2017). Most of this increase in demand will come from WCA, which will account for 60% of Africa's demand for sorghum and 84% of its demand for millets (Nedumaran et al., 2017).

Lesson # 2: Plant breeding must adapt to African farmers’ contexts and objectives

Developing improved varieties of sorghum and millets for Africa required re-orienting the breeding programme based on learnings of farmers’ contexts and objectives. ‘Adapt’ is used as shorthand to describe this process.

West African breeders spent decades developing new sorghum varieties based on breeding materials that they introduced from India and elsewhere (Yapi et al., 2000). The assumption was that the Green Revolution ideotype – short stature, high harvest index, fertiliser-responsive, insensitive to photoperiod, with wide adaptation – would provide high yield potential. This ideotype and the corresponding introduced Caudatum-race breeding materials were extensively used for developing ‘improved varieties’ for SSA. However, Malian farmers’ adoption of these new varieties was low in comparison to their (four-fold higher) adoption of varieties developed by selecting and purifying the farmers’ local varieties (Yapi et al., 2000).

Several learnings help to understand farmers’ varietal adoption decisions. First, the sensitivity to photoperiod of farmers’ varieties helps them adjust their maturity to better match the end of the rainy season despite the highly variable and unpredictable onset of rains in WCA. (Clerget et al., 2004, 2008). The tall height and loose open panicles of farmers’ varieties reduce risks of damage by birds, insects and cattle, and their hard grain favours both good cooking quality and reduced losses in storage and processing (Vom Brocke et al., 2010; Diallo et al., 2018). Furthermore, understanding grew over time of the importance of varietal diversity in farmers’ strategies for optimising output and minimising risks under the diverse and unpredictable growing conditions and the differing needs of individuals and communities over landscapes and years. Even in a single well-defined agroecology there was no ‘one best variety’ of sorghum or pearl millet (Smale et al., 2016; Christinck et al., 2014).

These learnings prompted West African sorghum breeders to adapt their programmes to emphasise use of the predominant race cultivated by farmers, the Guinea race, which possesses an array of adaptation and quality traits (Weltzien et al., 2018), New varieties developed by a Malian breeder from local varieties through mutagenesis were among the first improved Guinea race sorghum varieties to attract widespread interest from farmers (Breteaudau, pers. comm.). This success in creating genetic improvement within the Guinea race, without crossing to potentially high yielding but un-adapted introduced materials, encouraged the breeding programme in WCA to systematically explore options to breed Guinea race varieties with improved yield and good adaptation. Diverse breeding methods, including the development of Guinea race hybrid parents (Rattunde et al., 2013), are now used in WCA to breed new varieties based on a local germplasm base.

A second adaptation of the programmes was the increased emphasis on Participatory Plant Breeding (PPB). Farming Systems Research (FSR) had moved the testing of improved varieties onto farmers’ fields. PPB engaged farmers in the process of plant breeding. The breeding programmes in Mali and Burkina Faso developed testing procedures that rely on farmers’ collaboration and expertise for conducting trials, assessing plants, plots or varieties. This collaboration also oriented the selection programmes to better target farmers’ priorities (Rattunde et al., 2018; Weltzien et al., 2018). Resulting changes in the research directions included focusing on improving grain yield under poor soil fertility, especially low plant-available phosphorous (Leiser et al., 2018), improving the digestibility of stover for livestock feed, and biofortified sorghum to make nutrient-dense porridge to improve children's nutrition (Bauchspies et al., 2017). Farmers joining researchers in the breeding process thus helped to guide those programmes as well as overcome the limited testing capacity and number of researchers – for every scientist working on sorghum in WCA there were six in India (Ndjeunga et al., 2015). Equally important were joint farmer-research discussions, both in the field and during presentations of trial results in feedback and planning meetings, which were instrumental for adapting the selection strategy towards jointly understood and agreed goals (Christinck et al., 2020; Rattunde et al., 2021).

An outgrowth of these activities was the development of R & D ‘networks’ in which farmer organisations and cooperatives, national breeding programmes and international programmes collaborated (Haussmann et al., 2020). Such networks created many diverse new varieties, including high yielding photoperiod-sensitive Guinea race hybrids (Rattunde et al., 2013; Kante et al., 2017). In addition, the farmer seed-producer cooperatives and unions of cooperatives in Mali and Burkina Faso established or strengthened their capacities for large-scale production of certified seed of newly bred sorghum OPV and hybrid varieties as well as early generation seed, while continuing to conduct variety trials for continued variety improvement (Rattunde et al., 2021; Vom Brocke et al., 2020).

Lesson # 3: The seed system is the chief constraint on adoption of new varieties

The analysis focuses on two key elements of the seed system: availability and access of seed of preferred quality. Availability represents the supply side, or whether enough seed of the desired varieties is produced and available at the right time. Access denotes effective demand, or whether smallholders can afford to buy or barter seed, and whether seed is physically accessible or can be found within an acceptable distance.

In terms of availability, the challenge is that the public-sector cannot produce enough seed (Christinck et al., 2018), while private seed companies find OPVs unprofitable because growers can save and recycle seed. The result is that most seed is obtained through informal channels, where certified seed may be unavailable or, if available, of poor quality and unknown provenance.

Several business models have been developed to overcome this market failure for certified seed. Breweries use a vertically – integrated model to supply the white, low-tannin varieties needed for clear sorghum beer (Akinyoade et al., 2020; Orr et al., 2013). Another model in ESA (Tanzania, Uganda) is the production of Quality Declared Seed (QDS) by individual farmers, farmer groups, or cooperatives (Waithaka et al., 2021). A third model, found in WCA (eg. Mali), is the production of certified seed by farmer cooperatives and associations, who are contributing to the variety development (see above) (Weltzien et al., 2018). These cooperatives produce seed in close collaboration with local researchers, and may be contracted by seed trade companies, or agro-dealer associations. Lastly, there are networks of agro-dealers, linked to private seed companies, which are widespread in ESA but not in WCA, where farmers buy fewer inputs and infrastructure is less developed and (AFAP, 2020).

There is no question that these models have increased the availability of certified seed. In WCA (Mali, Niger, Burkina Faso), seed producer cooperatives have increased both the volume and diversity of certified seed. Villages where seed is produced have seen a sharp increase in adoption (Christinck et al., 2014). Farmer cooperatives now supply most of the certified sorghum seed in Mali (Weltzien et al., 2018). In Tanzania, the HOPE project used agro-dealers to market small packs of certified seed. Sixty percent of the seed was sold (Audi et al., 2015). This suggests that supply creates its own demand (Say's Law) and that availability is a bigger constraint than access.

The reality is more complicated. Rising adoption suggests that farmers are interested in the new varieties and can sometimes afford to buy certified seed. But even when certified seed is available, smallholders continue to recycle saved seed because sorghum is largely self-pollinated, and farmers have expertise in the production and storage of quality seed. Among farmers supplying Nigerian Breweries, 78% planted saved seed, although certified seed was available from the brewery and most growers were large farmers who could easily afford to buy (Akinyoade et al., 2020). Unlimited availability of certified seed does not guarantee a dramatic increase in adoption. In northern Nigeria, where making more farmers aware of improved varieties of sorghum had a minimal impact on adoption, because the varieties available lacked the traits that farmers wanted (Ndjeunga et al., 2011). Access is constrained by price. Although price did not limit access to hybrid seed in Mali (Christinck et al., 2018), the price of sorghum hybrids in ESA (Tanzania) can be double that of OPVs (Hambloch et al., 2021). Distance and time required to acquire seed limit access in WCA (Mali), where rainfall is so unpredictable that many farmers decide which variety to grow only at the time of sowing (Christinck et al., 2014). Thus, new seed must physically accessible. Finally, access is determined by cultural factors. In Mali, seed is a communal resource which is not sold but shared (Jones, 2017). In principle, therefore, everyone had access to seed. This was true even of hybrid seed. Once within the village, it was shared like any other seed, making it accessible to those without the cash to buy (Jones, 2014). In sum, the critical entry point for seed system development in SSA is making diverse seeds more widely available, including useful, reliable information about them. Involving farmers and their organisations in all the critical functions of seed system development facilitates this process and generates benefits not only through the use of the seed, but also from its production, processing and marketing.

Lesson # 4: High adoption without higher yields

Already by 2000 in semi-arid SSA there were pockets of high adoption. A review of seven ‘success stories’ in 2000 found a median adoption rate of 35% (Ahmed et al., 2000). Adoption has increased over time. A recent estimate (for a different sample of countries) suggests that the area planted to improved varieties of sorghum grew from 19% in 1998 to 32% in 2010, while that for pearl millet grew from 22% to 32% (Walker et al., 2015). For SSA overall, the rate of adoption was lower – 23% for sorghum and 14% for pearl millet (Fuglie and Marder, 2015).

Yet the impact on average yields has been minimal. Time-series data for the period 2000–2020 shows a negative trend in yields for sorghum, while the trend for millets is weakly positive (FAO, 2022). Cross-section data presents a mixed picture, with variation between countries and crops. In northern Nigeria, for example, improved pearl millet gave a small increase in yield while improved sorghum gave the same yield as local varieties (Ndjeunga et al., 2011). Where yields are low, an increase from improved varieties of, say, even 15% may scarcely be detectable. Meanwhile, in neighbouring Mali, Guinea-race sorghum hybrids increased yields by 30% (Smale et al., 2016).

The low yields of these crops are explained primarily by the hostile environments in which they grow. Indeed, farmers in SSA cultivate these crops, in part, because they can more reliably produce a harvest despite unfavourable moisture and soil-fertility conditions. Early - maturing improved varieties are sometimes adopted as a risk-reduction strategy for years when the rainy season is short. However, in average years these varieties would likely yield less than varieties with a longer crop duration. The result is that they are planted only on small areas or when the rains fail (Shapiro and Sanders, 2002). In Mali, farmers plant improved early-maturing varieties only when the rains set in late (Christinck et al., 2014) or when last year's harvest was poor, so that households must harvest new grain as early as possible (Siart, 2008). Moreover, farmers often plant sorghum and millets on less fertile land, which usually receives little or no fertiliser. In WCA, following the expansion in the area planted to maize, sorghum and especially pearl millet are increasingly farmed in more marginal fields and with riskier sowing dates – before or after cash crops are sown. In average years, therefore, improved early - maturing varieties may even depress yields (Ahmed et al., 2000; Lacy et al., 2006).

Varieties that give higher grain yields do not necessarily mean higher ‘food yields’ for farmers (Isaacs et al., 2018). Sorghum varieties created by plant breeding programmes, especially those with a high proportion of introduced caudatum-race germplasm, can have higher losses during storage and processing. Farmers in Mali assess new varieties not just for their potential grain yield but also for whether they are susceptible to attack during storage, their losses during food processing (most sorghum is decorticated), and the swelling capacity of the starch (which determines the volume of cooked food from a given quantity of grain). Varieties with low ‘food yields’ have not been readily adopted.

Higher productivity is a goal most R & D programmes in SSA have set for themselves. They often assume that farmers share the same goal. This assumption has some justification. After all, Malian farmers adopted high - yielding Guinea-race sorghum hybrids soon after their development and commercialisation (Sissoko et al., 2019). Yet farmers in SSA adopt new varieties of sorghum and pearl millet for diverse reasons. One broad motivation is for resilience and food security, with farmers adopting new varieties expected to provide adaptive advantages for changing conditions (Vom Brocke et al., 2014). Varieties that enable new cropping patterns, such as extra early varieties that make relay cropping possible, is another. Still another motivation is the possibility to increase the value of crop products. Novel dual-purpose sorghum varieties that produce both good quality grain for food and fodder for livestock in Mali (Weltzien et al., 2018) and varieties for industrial malting in Nigeria and Kenya are such examples. Thus, the understanding of farmers’ motives for adoption and its expected impacts should be widened. What determines adoption and impact is offering farmers a particular advantage, whether that be yield, value of crop products, or new cropping options.

Higher yields come not just from improved varieties but from improved crop management. Commercialisation is expected to give farmers the incentive to invest in inorganic fertiliser. However, the evidence suggests otherwise. In Kenya smallholders supplying the value chain for sorghum beer did not apply inorganic fertiliser (Orr et al., 2013) while in Nigeria growers producing sorghum for Nigeria Breweries increased supply not by using inorganic fertiliser to increase yields, but by expanding the area planted (Akinyoade et al., 2020). In ESA, only finger millet in Kenya and Ethiopia received any inorganic fertiliser – and finger millet in Ethiopia was rarely sold (Orr and Muange, 2022). Microdosing – the spot application of small amounts of inorganic fertiliser – is popular for pearl millet in Niger and sorghum in Nigeria (Yahaya et al., 2022; Winter-Nelson and Mazvimavi, 2014), where they are staple food crops, but not for sorghum or millets in Zimbabwe, where the staple is maize (Orr, 2018a). Smallholders are market opportunists rather than business investors seeking to maximise cash returns.

Lesson # 5: Technology can compensate for climate change

Climate models predict that the yields of sorghum and millets in SSA could go either up or down (Lipson et al., 2010). Falling yields are the result of higher temperatures, which shorten the time required for crops to reach maturity. By 2050, climate change in ESA may reduce yield of millets by 15% while the impact on sorghum will be minimal (<5%) (Adhikari et al., 2015). Climate models for WCA give similar results – yield declines of < 10% for both crops (MacCarthy et al., 2021).

Farmers can more than compensate for these losses by adopting existing technology. Water conservation, micro-dosing with inorganic fertiliser, or denser planting would cancel out the effects of climate change (Adam et al., 2020; Cooper et al., 2009). This is because temperature is not the main factor limiting yields. Yet while these results are encouraging, they are based on crop models and not on observation. If farmers are hesitant to adopt new technology now, why would they do so in future? If the yield loss is so small and gradual as to be imperceptible, the incentive for adoption is lacking. Moreover, the same constraints on adoption (Lesson 4) would still apply. ICRISAT's ‘hypothesis of hope’ (Cooper et al., 2009) may be just wishful thinking.

A bigger challenge for adaptation may be the growing frequency of ‘extreme events’ (eg. heavy rainfall during drier periods, or longer dry-spells) which will make weather even more variable and unpredictable. In WCA (southern Mali), farmers’ main strategy for coping with variable weather is varietal diversity (Lacy et al., 2006). In Mali, sorghum farmers generally grow several varieties. Farmers change varieties frequently, particularly after a drought, when farmers test varieties they have not previously grown (Siart, 2008). In the Sahelian Zone, where pearl millet is grown, farmers use the same strategy of varietal diversity, but in a different way. Farmers grow only one or two varieties. But the high out-crossing with pearl millet means that there is great diversity within the same variety. This includes traits like time of flowering, which can help avoid climatic stresses (Haussmann et al., 2007, 2012). Farmers exploit this intra-varietal diversity by using a diversity of panicles for use as seed of the varieties that will cope best with the whole range of weather conditions in their own area. Varietal diversity – meaning different types of variety as well as intra-varietal diversity - thus offers breeding programmes a proven strategy for addressing the challenge of climate change.

Lesson # 6: Commercialisation can be gender- and socially – inclusive

Commercialisation is said to disadvantage poorer smallholders and women. Is this also true for sorghum and millets? Certainly, this is true for clear sorghum beer in Nigeria, where sorghum is supplied by bigger farmers, and where demand ‘is encouraging the development of larger scale farming’. (Akinyoade et al., 2020: 9). Likewise, supply in ESA (Kenya, Uganda, and Tanzania) is concentrated among medium- to large-scale producers (Hambloch et al., 2021). Where poorer smallholders participate, as in Kenya, this has been driven by social enterprise rather than by breweries (Orr et al., 2013).

Nevertheless, the commercialisation of sorghum and millets is not confined to bigger farms. In ESA even the smallest farms sell at least part of their crop. Sorghum in Kenya, finger millet in Tanzania and Uganda are cash crops (Gierend and Orr, 2015). ‘Commercial farmers’ are not a homogeneous group. In central Tanzania, there were not one but three groups of ‘commercial’ finger millet growers: one a group of large farms, the second average-sized farms farming intensively using purchased inputs, and the third small farms farming intensively using family labour (Appiah-Kubi, 2017). Likewise, in Western Kenya, selling finger millet was ‘more common among female producers and poorer households’ (Handschuch and Wollni, 2016a: 356). Income from crop sales was important for small farmers. Women selling sorghum to East African Breweries earned less than $ 100 but needed the cash for school fees (Orr et al., 2013).

Adoption of improved varieties has not been inclusive. Improved varieties of sorghum have been adopted more frequently by bigger farms. This is because bigger farmers are better connected. Once there is equal access to information, the difference can disappear, as happens with sorghum in Tanzania (Kaliba et al., 2018; Muange and Schwarze, 2014) or finger millet in Kenya (Handschuch and Wollni, 2016b), though not in Nigeria (Ndjeunga et al., 2011). Improving access to information would make adoption more inclusive. Small seed packs – packs of seed (100 g) priced at just USD 10 to 20 cents – have proved effective in reaching smaller farmers in WCA (Mali, Niger, Burkina Faso) (Christinck et al., 2014). These not only provide information but allow farmers to test varieties for themselves. In Tanzania, 74% of farmers buying small seed packs from agro-dealers were trying these varieties for the first time (Audi et al., 2015).

The conventional wisdom is that commercialisation disempowers women because men take control of ‘women's crops’. The evidence for ESA suggests otherwise. In western Kenya, where finger millet is a cash crop, ‘we do not find an obvious gender bias in the marketing activity of finger millet’ (Handschuch and Wollni, 2016a: 350). Women benefitted more from commercialisation if they belonged to a collective marketing group. Membership helped women in Kenya access the value chain for clear sorghum beer (Orr et al., 2013) and get higher prices for finger millet (Handschuch and Wollni, 2016a). In WCA, ‘women's crops’ have a different meaning. As food crops, sorghum and millet are grown on family fields controlled by men but may also be grown on fields controlled by women, where they are used to buy household necessities (Cagley et al., 2009; Cook et al., 2009). In Burkina Faso, new technology for sorghum and millet increased the time women worked on family fields but reduced the time available for their own crops. Nevertheless, farm modelling shows that this increased income for women because they can earn more by working on family fields than on their own, which are less productive (Sanders et al., 1996). Whether women themselves feel this compensates them for their loss of autonomy is quite another matter.

Discussion

These lessons reveal that R & D for sorghum and millets in SSA is based on a paradox: there is a limited market for grain but there is a thriving market for seed. This paradox has strategic implications for the R & D programme.

The limited market demand for grain challenges the universality of the commercialisation model. Market-led development, where the adoption of new technology is spurred by ‘new uses’, has limited relevance for most sorghum and millet growers in SSA. This calls for a different Theory of Change in SSA (Orr and Muange, 2022). The dominance of demand for home consumption also determines the strategy for crop improvement. The primary objective is to protect household food security. Here the lessons have important implications for crop improvement. Lesson 3 (‘Higher adoption without higher yields’) suggests that rather than focus narrowly on realising yield potential, breeding should focus on realising the potential of variety improvement. Higher yields are not the only potential benefit from improved varieties and may be less important for farmers than early – maturity or stover digestibility Lesson 5 (‘Technology can compensate for climate change’) highlights the value of varietal diversity. Breeding programmes, especially commercial programmes, focus on the uniformity of their new varieties. This limits their adaptability in highly diverse environments. Targeting one specific high priority ‘product profile’ may limit their expected impacts to very specific situations. High genetic uniformity (e.g., single-cross pearl millet hybrids) can also increase production risks. Climate change makes it even more urgent for breeding programmes to give a higher priority to varietal diversity.

Priorities for plant breeding have implications for target – setting. The first concerns adoption rates. If smallholders plant a mix of varieties for specific reasons and will not replace them all with just one new variety, this imposes a ceiling on the adoption of improved varieties.⁶ In some countries, this adoption ceiling may already have been reached. The second concerns productivity. Without better crop management, improved varieties do not always mean higher yields. Put another way, ‘productivity gains from ‘naked’ varietal diffusion – adoption without changing input use or management practices – are likely to be statistically significant but small in rainfed agriculture in SSA’ (Walker et al., 2015: 402). The exception is hybrid varieties – like the Guinea-race sorghum hybrids in Mali – which have greater genetic gains than OPVs. If the target is to raise productivity, crop improvement should focus on hybrids.

The limited market for grain also determines the strategy for seed supply. Take the example of sorghum. There are three archetypes (BMGF, 2015). Where sorghum is grown for subsistence, using OPVs, seed supply is dominated by the public sector. In ESA (Kenya, Tanzania), seed supply is unprofitable for the private sector except as an ‘add-on’ to their core business (Hambloch et al., 2021). The gap has been filled by informal seed-producer groups or state – run seed farms. Where there are niche markets like clear sorghum beer, but where the market is too small to attract private seed companies, there will be public-private partnerships. This is the case with clear sorghum beer, where the market is a tiny fraction of total demand, and low-tannin hybrids were developed in partnership with Nigerian Breweries (Akinyoade et al., 2020). Finally, where there is demand for hybrids, there is scope for a commercial system with both breeding and seed supply controlled by the private sector. Hybrid sorghum is widely grown in Mali. However, private seed companies play no role either in plant breeding or in seed supply. Farmer organisations have a competitive advantage because they are decentralisation and can respond quickly to demand that is both localised and unpredictable (Christinck et al., 2014).

These lessons from R & D raise three more general questions. The first concerns commercialisation. Stage-theory sees the present seed system in SSA as a temporary phase, to be replaced as speedily as possible by a fully-fledged commercial system. The seed system in SSA is a far cry from that in India, with its consortium of public and private breeding and its myriad private seed companies. But the seed system in India is also a far cry from that in the United States, with its oligopoly of giant seed companies. The alternative is to see the seed system in SSA as a different form of system, as an integrated system which combines old and new elements (Louwaars et al., 2013). The market for hybrid seed in WCA provides an unusual example. The taboo on private sale of seed has resulted in two types of seed market (Jones, 2017). Outside the village, the market for hybrid seed is commercial. There are buyers/sellers, the cash nexus, and profit/loss. Inside the village, the market for hybrid seed remains traditional. There are givers/receivers, barter transactions, and reciprocity. These are not parallel markets but interconnected. The traditional market has been adapted to take advantage of the commercial market, but in ways that do not violate the taboo on the sale of seed to fellow farmers.⁷ Hybrid seed has a dual identify, with both economic and social values. Thus, the commercialisation of seed production has not meant the complete commercialisation of seed markets.

The two other general questions concern R & D programmes. One is about metrics. ‘Adoption’ – of certified seed, improved crop management, and inputs – is a key metric for these programmes. Recent critiques have highlighted its shortcomings. By measuring ‘adoption’ as a binary choice, it ignores how the technology was developed, or its potentially harmful effects (Glover et al., 2019).⁸ Adoption also depends on institutions and policies, which are outside the programme's control (Orr, 2018). The R & D programme in WCA addresses these problems in an interesting way. First, PPB changes the process of technology development. Farmers are not mere adopters but co-creators of new technology. Farmer agency is baked – in. This reduces the risk of unintended negative effects. Second, R & D integrates crop improvement with seed supply, with certified seed being produced and disseminated through farmer co-operatives. Thus, the design of the R & D programme can make ‘adoption’ a more meaningful metric.

The final question is about the form of R & D programmes. The R & D programme for sorghum and millets in SSA is not merely a less-developed version of that in India or elsewhere. It should not be judged against some ideal model but on its own merits. Again, the example of sorghum in Mali is instructive (Rattunde et al., 2021). The conventional model of R & D – breeding by the public sector, seed supply by private companies – gave disappointing results. This sparked the search for a more effective model. Breeders involved farmers in setting breeding priorities, while farmer cooperatives produced and marketed seed, with de-centralised seed committees deciding which improved varieties to produce based on local demand. The evolution of this ‘network’ model took 20 years, but the upshot is that 33% of the area planted to sorghum in Mali is now covered by improved varieties (Ndjeunga et al., 2015; Smale et al., 2016). The wider lesson is that R & D programmes evolve in response to a specific context. Their success or failure reflects how well they understand and adapt to that context.

How likely is it that these lessons will be learned? Here the record is mixed. The lesson on ‘higher adoption without higher yields’ originated with FSR long ago in the 1980s (Matlon, 1990). However, the continued primacy given to productivity suggests that this lesson has not yet been learned. This reflects both external and internal factors. Externally, this lesson conflicted with the narratives of the Green Revolution, on which ICRISAT's vision for the drylands was based,⁹ or with market-led development, which saw productivity as driven by commercialisation. Internally, FSR lacked an institutional base. Once donors lost interest, FSR personnel were not replaced. Consequently, FSR's central credo, that R & D should be ‘organised around an understanding of farmers’ conditions and practices’ (Tripp, 1991: 4) was never fully internalised. The features of FSR that have survived, such as on-farm trials or systems analysis, were those that could be incorporated in ways that did not challenge the original assumptions underlying the R & D strategy.

By contrast, the lesson that ‘plant breeding must adapt to the African context’ has been taken to heart, at least in Mali and Burkina Faso. Why the difference? Once again, both external and internal factors were at play. Learning took 20 years. Internally, therefore, continuity of personnel within a core programme ensured an institutional base. This allowed innovative approaches, such as PPB, to be internalised. Externally, breeding objectives can now be fitted into the narrative of the Green Revolution, thanks to the genetic gains from hybrids. Externally, too, the ‘network’ approach created a coalition of interest groups which gave R & D an added legitimacy and a greater claim on resources. A decentralised model also elevates the importance of NARES. Breeding for heterogeneous environments shifts the emphasis from international public goods, the mandate of ICRISAT, towards goods that are national or even local in scope, the mandate of NARES. Thus, the network approach has far-reaching implications for how future R & D should be resourced, managed, and implemented. This re-structuring will be challenging. But if Mali and Burkina Faso can initiate such a network approach there is no reason, in principle, why it should not be equally successful in neighbouring, better-resourced NARES, and in a shorter time.

Conclusion

ICRISAT's R & D for sorghum and millets is based on market-led development. This makes sense for India, where they are fast becoming commercial crops. Is it also appropriate for Africa?

We explored lessons learned from R & D over the past two decades. The main lesson is that sorghum and millets are grown for home consumption and that commercialisation is limited. There are pockets of commercialisation, but they are islands in a sea of subsistence agriculture. In consequence, plant breeding remains in the public sector and private firms play a limited role in the supply of certified seed. Improved varieties are designed to protect household food security in drought-prone environments. They are often early – maturing and grown on marginal land. This means that improved varieties do not always increase yields, which helps put a ceiling on their adoption. Adapting R & D to the African context has required several innovations. A striking example of this adaptation is the ‘network’ model of R & D in WCA, where breeders work with farmers to develop improved varieties and certified seed is supplied by farmer cooperatives. The success of this approach in Mali shows what can be achieved when R & D is re-designed to make it more appropriate for Africa.

These lessons have implications for R & D strategy. Plant breeding should prioritise food security rather than markets. The focus should not be just higher yields but on a wider set of traits that meet specific needs and growing conditions. Varietal diversity, which reduces the risk from a variable and changing climate, is more important than uniformity and wide adaptation of individual varieties. For seed supply, the lessons point to the need for a decentralised structure that can respond quickly to local demand. In WCA, this is provided by farmer organisations. In ESA, a pluralistic system with different types of suppliers offers the best option. The lessons also have wider implications. Chief among these is the need to judge the R & D programme in SSA on its own merits, rather than by its evolution towards some ideal commercial model. A pragmatic approach is to view the programme as a form of R & D adapting to a unique environment and ask what further adaptations would make it more effective.

There is still a lot we do not know – the known unknowns – let alone the unknown unknowns. The lessons have identified some knowledge gaps, including the impact of commercialisation on women, or whether technology really will compensate for climate change. We leave these questions to future researchers who, in reviewing these lessons, may find that we have invented some myths of our own.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Bill and Melinda Gates Foundation, CGIAR Research Program on Dryland Cereals

ORCID iD

Alastair Orr

Notes

References

Adam

MacCarthy

Traoré

PCS

, et al. (2020) Which is more important to sorghum production systems in the Sudano-Sahelian zone of West Africa: climate change or improved management practices? Agricultural Systems 185: 102920.

Adhikari

Nejadhashemi

Woznicki

(2015) Climate change and Eastern Africa: a review of impact on major crops. Food and Energy Security 4(2): 110–132.

AFAP (2020) African Fertilizer and Agribusiness Partnership, Assessment of Sustainable Agro-dealer Networks. Report to Bill and Melinda Gates Foundation. 64 pp.

Ahmed

Sanders

Nell

(2000) New sorghum and millet cultivar introduction in Sub-Saharan Africa: impacts and research agenda. Agricultural Systems 64: 55–65.

Akinyoade

Ekumankama

Uche

(2016) The use of local raw materials in beer brewing: Heineken in Nigeria. Journal of the Institute of Brewing 122: 682–692.

Akinyoade

Kazimierczuk

Ekumankama

, et al. (2020) Sorghum Value Chain in Nigeria. Explorative study. Research Report. African Studies Centre, University of Leiden.

Alumira

Heinrich

(2003) Market-led development: R & D strategies for agriculture in Southern Africa: Summary proceedings of a regional research strategy workshop 30–31 July 2002, Bulawayo, Zimbabwe. International Crops Research Institute for the Semi-Arid Tropics, Bulawayo.

Appiah-Kubi

(2017) Commercialization Dynamics and Welfare Implications for Dryland Cereal Farmers in Central Tanzania: A Typology Study. M.Sc. Thesis, University of Bonn, 113 pp.

Audi

Sakwera

Ziwa

, et al. (2015) The effectiveness and complementarity of field days and small seed packs (SSPs) in delivering Dryland Cereal technologies: A survey of field day participants and agro-vets in Singida and Iramba districts of central Tanzania. Working Paper Series No 61. ICRISAT Research Program Markets, Institutions and Policies. Patancheru, Telangana, India: International Crops Research Institute for the Semi-Arid Tropics. 40 pp.

10.

Bauchspies

Diarra

Rattunde

, et al. (2017) “An Be Jigi”: collective cooking, whole grains, and technology transfer in Mali. FACETS 2: 955–968.

11.

BMGF (2015) Bill & Melinda Gates Foundation, Early Generation Seed Study. April. Mimeo, 122 pp.

12.

Cagley

Anderson

Klawitter

(2009) Gender and Cropping: Millet in Sub-Saharan Africa. Evans School Policy Analysis and Research (EPAR). Evans School of Public Affairs, University of Washington. Mimeo, 6 pp.

13.

Chambers

(2017) Can We Know Better? Reflections for Development. Rugby: Practical Action Publishing Ltd

14.

Cheng

Larochelle

(2016) Estimating household demand for millet and sorghum in Niger and Nigeria. ICRISAT Socio-Economic Discussion Paper Series. Series Paper Number 39. Mimeo, 39 pp.

15.

Christiaensen

Demery

(2018) Agriculture in Africa: Telling Myths from Facts. Directions in Development. Washington, DC: World Bank.

16.

Christinck

Diarra

Haberler

(2014) Innovations in Seed Systems Lessons from the CCRP-Funded Project “Sustaining Farmer-Managed Seed Initiatives in Mali, Niger, and Burkina Faso”. McKnight Foundation Collaborative Research Program. Minnesota: McKnight Foundation. Mimeo, 75 pp.

17.

Christinck

Rattunde

Kergna

, et al. (2018) Identifying options for the development of sustainable seed systems: Insights from Kenya and Mali, ZEF Working Paper Series, No. 165, University of Bonn, Centre for Development Research (ZEF), Bonn.

18.

Christinck

Rattunde

Weltzien

(2020) Building collaborative advantages through long-term farmer-breeder collaboration. In: Westengen

Winge

(eds) Farmers and Plant Breeding. Current Approaches and Perspectives. London: Routledge, pp.79–94.

19.

Clerget

Dingkuhn

Chantereau

, et al. (2004) Does panicle initiation in tropical sorghum depend on day-to-day change in photoperiod? Field Crops Research 88: 11–27.

20.

Clerget

Dingkuhn

Gozé

, et al. (2008) Variability of phyllochron, plastochron and rate of increase in height in photoperiod-sensitive sorghum varieties. Annals of Botany 101: 579–594.

21.

Cook

Curran

Anderson

(2009) Gender and Cropping: Sorghum in Sub-Saharan Africa. Evans School Policy Analysis and Research (EPAR). Evans School of Public Affairs, University of Washington. Mimeo, 6 pp.

22.

Cooper

Rao

KPC

Singh

, et al. (2009) Farming with current and future climate risk: advancing a ‘hypothesis of hope’ for rainfed agriculture in the semiarid tropics. SAT e-Journal 7: 1–19.

23.

Dalton

Hodjo

, (2020) Trends in global production, consumption, and utilization of Sorghum. In: Tonapi

Talwar

Are

(eds) Sorghum in the 21st Century: Food – Fodder – Feed – Fuel for a Rapidly Changing World. Singapore: Springer, pp.3–16.

24.

Diallo

Isaacs

Gracen

, et al. (2018) Learning from farmers to improve sorghum breeding objectives and adoption in Mali. Journal of Crop Improvement 32(6): 829–846.

25.

Doss

Meinzen-Dick

Quisumbing

, et al. (2018) Women in agriculture: four myths. Global Food Security 18: 69–74.

26.

FAO (2022) FAOSTAT Statistical Database. Rome, Italy: Food and Agriculture Organization of the United Nations (FAO). Available at: https://www.fao.org/faostat/en/#home (accessed 30 June 2022).

27.

Feed the Future Innovation Lab for Collaborative Research on Sorghum and Millet (2021) Annual Performance Report FY 2021. Sorghum and Millet Innovation Lab, Kansas State University, Manhattan, KS, pp. 57.

28.

Fuglie

Marder

(2015) The diffusion and impact of improved crop varieties in Africa. In: Walker

Alwang

(eds) Crop Improvement, Adoption and Impact of Improved Varieties in Food Crops in Sub-Saharan Africa. Wallingford: CAB International, pp.338–369.

29.

Gierend

Orr

(2015) Consumer Demand for Sorghum and Millets in Eastern and Southern Africa: Priorities for the CGIAR Research Programme for Dryland Cereals SocioEconomics Discussion Paper Series No. 42. International Crops Research Institute for the Semi-Arid Tropics: Nairobi.

30.

Glover

Sumberg

Ton

, et al. (2019) Rethinking technological change in smallholder agriculture. Outlook on Agriculture 48(3): 169–180.

31.

Hambloch

Kahwai

Mugonya

(2021) Contextualizing private sector-based seed system development: the case of sorghum in Eastern Africa. Outlook on Agriculture 50(4): 1–14.

32.

Handschuch

Wollni

(2016a) Improved production systems for traditional food crops: the case of finger millet in western Kenya. Food Security 8: 783–797.

33.

Handschuch

Wollni

(2016b) Traditional food crop marketing in Sub-Saharan Africa: Does gender matter? The Journal of Development Studies 52(3): 343–359.

34.

Haussmann

BIG

Aminou

Descheemeaker

, et al. (2020) Tackling key issues for smallholder farmers: The farmer research network (FRN) approach. In: Tonapi

Talwar

Are

et al. (eds) Sorghum in the 21st Century: Food – Fodder – Feed – Fuel for a Rapidly Changing World, Singapore: Springer, 315–329. https://doi.org/10.1007/978-981-15-8249-3_13

35.

Haussmann

BIG

Boureima

Kassari

, et al. (2007) Two mechanisms of adaptation to climate variability in West African pearl millet landraces – a preliminary assessment. E-Journal of SAT Research 3(1): 1–3.

36.

Haussmann

BIG

Rattunde

Weltzien-Rattunde

, et al. (2012) Breeding strategies for adaptation of pearl millet and Sorghum to climate variability and change in West Africa. Journal of Agronomy and Crop Science 198: 327–339.

37.

ICRISAT (2010) ICRISAT Strategic Plan to 2020: Inclusive Market-oriented Development for Smallholder Farmers in the Tropical Drylands. Patancheru: International Crops Research Institute for the Semi-Arid Tropics.

38.

ICRISAT and partners (2017) CGIAR research program: Grain legumes and dryland cereals agri-food systems. Full Proposal. Mimeo, 100 pp.

39.

Isaacs

Weltzien

Diallo

, et al. (2018) Farmer engagement in culinary testing and grain quality evaluations provides crucial information for sorghum breeding strategies in Mali. In: Tufan (ed) et al. (ed.) State of the Knowledge for Gender-Responsive Breeding: Case Studies for Practitioners. CGIAR Gender and Breeding Initiative Working Paper No. 3. Lima: International Potato Center (CIP), pp.74–85.

40.

Jayne

Snapp

Place

, et al. (2019) Sustainable agricultural intensification in an era of rural transformation in Africa. Global Food Security 20: 105–113.

41.

Jones

(2014) Emerging seed markets, substantive seed economies, and integrated seed systems in West Africa: A mixed methods analysis. PhD Thesis, University of Pennsylvania. 413 pp.

42.

Jones

(2017) Social institutions mediating seed access in West African seed systems. FACETS 2: 998–1014.

43.

Kaliba

Mazvimavi

Gregory

, et al. (2018) Factors affecting adoption of improved sorghum varieties in Tanzania under information and capital constraints. Agricultural and Food Economics 6: 18.

44.

Kante

Rattunde

Leiser

, et al. (2017) Can tall Guinea-race sorghum hybrids deliver yield advantages to smallholder farmers in West and Central Africa? Crop Science 57: 1–10.

45.

Kumara Charyulu

Moses

Bantilan

, et al. (2016a) Rainy Season Sorghum Technology Adoption and Impact Study in Maharashtra. Research Report 70. Patancheru 502 324. Telangana, India: International Crops Research Institute for the Semi-Arid Tropics. 80 pp.

46.

Kumara Charyulu

Moses

Bantilan

, et al. (2016b) Pearl Millet Technology Adoption and Impact Study in Maharashtra. Research Report 71. Patancheru 502 324. Telangana, India: International Crops Research Institute for the Semi-Arid Tropics. 76 pp.

47.

Lacy

Cleveland

Soleri

(2006) Farmer choice of Sorghum varieties in Southern Mali. Human Ecology 34(3): 331–353.

48.

Leach

Mearns

(eds) (1996) The Lie of the Land: Challenging Received Wisdom on the African Environment. Oxford: James Currey Ltd

49.

Lipson

Po-Chedley

Smoliak

, et al. (2010) Crops & Climate Change: Sorghum and Millet. EPAR Brief No. 115, Evans School Policy Analysis and Research (EPAR), University of Washington.

50.

Louwaars

de Boef

Edeme

(2013) Integrated seed sector development in Africa: a basis for seed policy and law. Journal of Crop Improvement 27(2): 186–214.

51.

Lufumpa

Mubila

Ben Aissa

(2015) The dynamics of the middle class in Africa. In: Ncube

Lufumpa

(eds) The Emerging Middle Class in Africa. London: Routledge, pp.9–33.

52.

MacCarthy

Adam

Freduah

, et al. (2021) Climate change impact and variability on cereal productivity among smallholder farmers under future production systems in West Africa. Sustainability 13: 5191.

53.

Matlon

(1985) A critical review of objectives, methods, and progress to date in sorghum and millet improvement: a case study of ICRISAT/Burkina Faso. In: Ohm

Nagy

(eds) Appropriate Technologies for Farmers in Semi-Arid West Africa. West Lafayette, Indiana: Purdue University International Program in Agriculture, pp.154–178.

54.

Matlon

(1990) Improving productivity in Sorghum and pearl millet in Semi-Arid Africa. Food Research Institute Studies 22(1): 1–44.

55.

Meena

Joshi

Bisht

, et al. (2021) Global scenario of millets cultivation. In: Kumar

Tripathi

Joshi

, et al. (eds) Millets and Millet Technology. Singapore: Springer, pp.33–50.

56.

Muange

Schwarze

(2014) Social Networks and the Adoption of Agricultural Innovations: The Case of Improved Cereal Cultivars in Central Tanzania. SocioEconomics Discussion Paper Series No. 18. Nairobi: International Crops Research Institute for the Semi-Arid Tropics.

57.

Nakelse

Dalton

(2018) Are Urban Consumers in Niger Willing to Pay for Safe and Nutritious Food? Unpublished paper. AgEcon Search http://ageconsearch.umn.eduaesearch@umn.edu.

58.

Ndjeunga

Mausch

Simtowe

(2015) Assessing the effectiveness of agricultural R & D for groundnut, pearl millet, pigeonpea and Sorghum in West and Central Africa and East and Southern Africa. In: Walker

Alwang

(eds) Crop Improvement, Adoption and Impact of Improved Varieties in Food Crops in sub-Saharan Africa. Wallingford: CAB International, pp.123–147.

59.

Ndjeunga

Umar

Ahmed

, et al. (2011) Adoption and impacts of modern sorghum and pearl millet varieties in Northern Nigeria. Working Paper. Markets, Institutions and Policies Programme. Hyderabad: International Crops Research Institute for the Semi-Arid Tropics. Mimeo, 81 pp.

60.

Nedumaran

Gbegbelegbe

Alene

, et al. (2017) Foresight Analysis for Grain Legumes and Dryland Cereals. Hyderabad: International Crops Research Institute for the Semi-Arid Tropics. Mimeo, 12 pp.

61.

Orr

(2018b) Killing the goose? The value chain for sorghum beer in Kenya. Journal of Agribusiness in Developing and Emerging Economies 8(1): 1–22.

62.

Orr

(2018a) Markets, institutions and policies: a perspective on the adoption of agricultural innovations. Outlook on Agriculture 47(2): 81–86.

63.

Orr

Gierend

Choudhary

(2017) Value chains for sorghum and millets in Eastern and Southern Africa: Priorities for the CGIAR research program for dryland cereals. ICRISAT Socio-Economics Discussion Paper Series No. 42. Mimeo, p. 81.

64.

Orr

Muange

(2022) Hedgehog or fox? Theories of change for dryland cereals in Eastern Africa. Journal of International Development 34(3): 532–549.

65.

Orr

Mwema

Mulinge

(2013) The value chain for sorghum beer in Kenya, Socioeconomics Discussion Paper Series No. 16. ICRISAT, Nairobi. Mimeo, 54 pp.

66.

Orr

Schipmann-Schwarze

Gierend

, et al. (2020) Why invest in research & development for sorghum and millets? The business case for East and Southern Africa. Global Food Security 26: 100458.

67.

Palacios-Lopez

Christiaensen

Kilic

(2017) How much of the labor in African agriculture is provided by women? Food Policy 67: 52–63.

68.

Parthasarathy Rao

Basavaraj

(2015) Status and prospects of millet utilization in India and global scenario. In: Millets: Promotion for Food, Feed, Fodder, Nutritional and Environment Security, Proceedings of Global Consultation on Millets Promotion for Health & Nutritional Security. Society for Millets Research, ICAR Indian Institute of Millets Research, Hyderabad, pp. 197–209.

69.

Pray

Nagarajan

(2009) Improving crops for arid lands: Pearl millet and sorghum in India. In: Spielman

Pandya-Lorch

(eds) Millions Fed: Proven Successes in Agricultural Development. Washington, D.C.: International Food Policy Research Institute, pp.83–88.

70.

Rattunde

Sidibé

Diallo

, et al. (2018) Involving women farmers in variety evaluations of a ‘Men’s Crop’: consequences for the Sorghum breeding strategy and farmer empowerment in Mali. In: Tufan

Grando

Meola

(eds) State of the Knowledge for Gender in Breeding: Case Studies for Practitioners, Lima: International Potato Center (CIP), pp.95–107. https://cgspace.cgiar.org/handle/10568/92819.

71.

Rattunde

Weltzien

Sidibé

, et al. (2021) Transforming a traditional commons-based seed system through collaborative networks of farmer seed-cooperatives and public breeding programs: The case of sorghum in Mali. Agriculture and Human Values 38: 561–578.

72.

Rattunde

HFW

Weltzien

Diallo

, et al. (2013) Yield of photoperiod-sensitive sorghum hybrids based on Guinea-race Germplasm under farmers’ field conditions in Mali. Crop Science 53: 2454–2461.

73.

Revoredo-Giha

Toma

Akaichi

, et al. (2022) Exploring the effects of increasing underutilized crops on consumers’ diets: the case of millet in Uganda. Agricultural and Food Economics 10(1): 1–21.

74.

Sanders

Shapiro

Ramaswamy

(1996) The Economics of Agricultural Technology in Semiarid Sub-Saharan Africa. Baltimore: Johns Hopkins University Press.

75.

Shapiro

Sanders

(2002) Natural resource technologies for semi-arid regions of sub-Saharan Africa. In: Barrett

Place

Aboud

(eds) Natural Resources Management in African Agriculture: Understanding and Improving Current Practices. Wallingford: CAB International, pp.261–274.

76.

Siart

(2008) Strengthening Local Seed Systems: Options for Enhancing Diffusion of Varietal Diversity of Sorghum in Southern Mali. Weikersheim, Germany: Magraf Publishers GmbH.

77.

Sissoko

Smale

Castiaux

, et al. (2019) Adoption of new Sorghum varieties in Mali through a participatory approach. Sustainability 11(17): 4780.

78.

Smale

Kergna

Diakité

(2016) An Economic Assessment of Sorghum Improvement in Mali. Impact Assessment Report No. 2. Patancheru 502 324, Telangana, India: International Crops Research Institute for the Semi-Arid Tropics. 52 pp.

79.

Spielman

Kennedy

(2016) Towards better metrics and policymaking for seed system development: Insights from Asia’s seed industry. Agricultural Systems 147: 111–122.

80.

Leiser

Weltzien-Rattunde

(2018) Sorghum tolerance to low-phosphorus soil conditions. In: Sustainable Cultivation of Sorghum , Rooney

(ed) Achieving Sustainable Cultivation Of Sorghum Volume 1: Genetics, Breeding and Production Techniques. Cambridge: Burleigh Dodds Science Publishing Limited, pp.247–272.

81.

Tripp

(ed.) (1991) Planned Change in Farming Systems: Progress in On-Farm Research. Chichester: John Wiley and Sons.

82.

Tschirley

Reardon

Dolislager

, et al. (2015) The rise of a middle class in East and Southern Africa: Implications for food system transformation. Journal of International Development 27: 628–646.

83.

UN (2019a) United Nations, Department of Economic and Social Affairs, Population Division World Urbanization Prospects 2018: Highlights (ST/ESA/SER.A/421).

84.

UN (2019b) United Nations, Department of Economic and Social Affairs, Population Division World Population Prospects 2019: Highlights (ST/ESA/SER.A/423).

85.

Vom Brocke

Kondombo

Guillet

, et al. (2020) Impact of participatory sorghum breeding in Burkina Faso. Agricultural Systems 180: 102775. DOI: 10.1016/j.agsy.2019.102775.

86.

Vom Brocke

Trouche

Weltzien

, et al. (2010) Participatory variety development for sorghum in Burkina Faso: farmers’ selection and farmers’ criteria. Field Crops Research 119: 183–194.

87.

Vom Brocke

Trouche

Weltzien

, et al. (2014) Helping farmers adapt to climate and cropping system change through increased access to sorghum genetic resources adapted to prevalent sorghum cropping systems in Burkina Faso. Experimental Agriculture 50(02): 284–305. DOI: 10.1017/S0014479713000616.

88.

Waithaka

Mugoya

Mabaya

, et al. (2021) Decentralized Seed Services in Africa: An Assessment of Tanzania and Uganda. ZEF Working Paper Series, No. 206, Centre for Development Research, University of Bonn.

89.

Walker

(2010) Challenges and opportunities for agricultural R & D in the semi-arid tropics. Internal Document for Strategic Planning. ICRISAT. March. Mimeo, 76 pp.

90.

Walker

Alwang

Alene

, et al. (2015) Varietal adoption, outcomes and impact. In: Walker

Alwang

(eds) Crop Improvement, Adoption and Impact of Improved Varieties in Food Crops in sub-Saharan Africa. Wallingford: CAB International, pp.388–405.

91.

Wangari

Mwema

Siambi

, et al. (2020) Changing perception through a participatory approach by involving adolescent school children in evaluating smart food dishes in school feeding programs – real-time experience from Central and Northern Tanzania. Ecology of Food and Nutrition 59(5): 472–485.

92.

Weltzien

Rattunde

HFW

van Mourik

, et al. (2018) Sorghum cultivation and improvement in West and Central Africa. In: Rooney

(ed.) Achieving Sustainable Cultivation of Sorghum Volume 2: Sorghum Utilization Around the World. Cambridge: Burleigh Dodds Science Publishing Limited, pp.1–24.

93.

Winter-Nelson

Mazvimavi

(2014) Economic Impact Evaluation of the ICRISAT Jewels. Patancheru 502 324, Telangana, India: International Crops Research Institute for the Semi-Arid Tropics. 40 pp.

94.

Yahaya

Shimelis

Nebie

, et al. (2022) Sorghum production in Nigeria: opportunities, constraints, and recommendation. Acta Agriculturae Scandinavica, Section B — Soil & Plant Science 72(1): 660–672.

95.

Yapi

Kergna

Debrah

, et al. (2000) Analysis of the economic impact of sorghum and millet research in Mali. Impact Series no. 8. Patancheru, Andhra Pradesh, India: International Crops Research Institute for the Semi-Arid Tropics. 60 pp.

96.

Zereyesus

Dalton

(2017) Rates of return to sorghum and millet research investments: a meta-analysis. PLoS ONE 12(7): e0180414.