History and progress in genetic improvement for enhancing rice yield in sub-Saharan Africa

Highlights

• Efforts of genetic improvement of rice in sub-Saharan Africa (SSA) are reviewed.

• Yield of selected varieties was assessed for 4 so-called breeding target domains.

• Improving yield performance of existing upland NERICA varieties is possible.

• Lowland interspecifics out-yielded improved O. sativa varieties released in SSA.

• Old varieties from Asia are still common in SSA, especially in irrigated lowland.

Abstract

Genetic improvement has been given high priority in rice research for development in sub-Saharan Africa (SSA). This paper provides an overview of historical efforts of genetic improvement in SSA focused on improving rice productivity. It further describes yield gain of new rice varieties evaluated in recent field experiments in four selected breeding target domains, and summarizes adoption studies on rice varieties. Efforts to develop rice varieties adapted to diverse rice production systems in SSA have been made by several research organizations, including national agricultural research institutes, bilateral organizations, and three CGIAR centers. The efforts have resulted in the release of around 570 rice varieties in 10 major rice-producing SSA countries by 2020. Among these varieties, the most well-known are interspecfic upland rice varieties “New Rice for Africa (NERICA)” that were developed from crosses between improved Oryza sativa tropical japonica and Oryza glaberrima varieties. However, recent field assessements demonstrated that upland rice yield can be further improved through introduction of upland indica materials from Asia. In contrast, lowland interspecific rice varieties out-yielded the previously released improved O. sativa varieties in selected SSA countries. Field assessments in Senegal and Madagascar did not demonstrate the yield advantage of the recently-developed varieties including hybrids over best performing ones released before. The adoption level of improved varieties in terms of share of area occupied by improved varieties in 18 SSA countries was 40 % in 2008 with old varieties that were released before 2000 being dominant and 7% of the total rice area occupied by NERICA varieties. Especially in irrigated and rainfed lowland rice systems, the majority of dominant varieties were still those introduced from Asia. Panel data between 2002 and 2019 from the Senegal River Valley also confirmed this, and Sahel 108, which was introduced from Asia and released in Senegal in 1994, accounted for >70 % of the total rice area in both the wet and dry seasons in 2019. The implications of this review for future rice genetic improvement and varietal replacement in SSA are discussed.

Introduction

Rice is a staple food for more than half of the world’s population. Worldwide, more than 3.5 billion people depend on rice for at least 20 % of their daily calorie intake (GRiSP (Global Rice Science Partnership), 2013). Asia is the dominant rice consumer and accounts for 90 % of its global consumption, followed by Africa, where rice is the fastest-growing food staple. Although rice self-sufficiency has not been achieved in Africa (Saito et al., 2015), global rice demand has been met until today through increased rice productivity and expansion of rice area. Genetic improvement has played an important role in enhancing rice productivity (Pingali, 2012). Improved rice varieties having superior grain yield potential and input-responsiveness combined with improved management practices and inputs such as fertilizer and irrigation have contributed to large grain yield increases in lowland rice production systems especially during the ‘Green Revolution’ period in Asia (Evans, 1993; Cassman, 1999).

It has been frequently indicated that the Green Revolution bypassed sub-Saharan Africa (SSA) (Pingali, 2012), where rice is grown in diverse rice production systems, including irrigated lowland, rainfed lowland, rainfed upland, deep water, and mangrove swamps. The last two production systems are of relatively minor importance in terms of surface area (Saito et al., 2013). Except for irrigated lowland rice, rice in SSA is grown extensively with limited inputs, resulting in low yield (Saito and Futakuchi, 2009; van Oort et al., 2017). However, recent field studies in the 2010s showed that improved varieties combined with improved field management have contributed to improved grain yield in Africa (Arouna et al., 2017; Niang et al., 2017; Senthilkumar et al., 2018). Rice grain yield levels in three major production systems in Africa, namely irrigated lowland, rainfed lowland, and rainfed upland rice, were in similar ranges as those recently observed in Southeast and South Asia (Tanaka et al., 2017; Senthilkumar et al., 2020). Irrigated rice production systems had a higher adoption rate of improved varieties and fertilizer inputs than the other production systems and have benefitted from the introduction of genetic materials developed in Asia (Dalton and Guei, 2003; Niang et al., 2017), resulting in its highest yield. Low national average rice grain yield levels in Africa (on average around 2 t/ha) are mainly attributed to a relatively large share of rice-grown area in rainfed lowland (32 %) and rainfed upland (28 %) systems in contrast to Asia where irrigated lowland is the dominant production system (Diagne et al., 2013).

Besides enhancements in yield potential and input-responsiveness, genetic improvement in Asia has resulted in the development of stress-resilient varieties that are tolerant to abiotic stresses such as submergence, drought, and salinity (Yamano et al., 2016). Large-scale dissemination of such varieties could provide a great incentive to farmers for improving their management practices as these varieties reduce climate risks especially in rainfed production systems (Haefele et al., 2016). In the case of SSA, improvement of abiotic and biotic stress tolerance has mostly been attempted since the rainfed production systems possess high share in area. For instance, upland and lowland New Rice for Africa (NERICA) varieties tolerant to a range of abiotic and biotic stresses were developed from crosses between improved Oryza sativa tropical japonica and Oryza glaberrima varieties and between improved Oryza sativa indica and Oryza glaberrima varieties, respectively. These NERICA varieties were disseminated to a wide range of countries (Saito et al., 2012, 2018).

Despite these achievements, great challenges remain that require further efforts on genetic improvement for rice and other research domains for enhancing rice grain yield and production. Global rice consumption remains high and will continue to increase. For example, it was estimated that global demand will increase from 479 million tons milled rice in 2014/15 to 536–551 million tons in 2029/30 (GRiSP (Global Rice Science Partnership), 2016). In Africa, annual rice consumption per capita is projected to increase by about 5 kg over the period of 2019–2028 (OECD/FAO, 2019). Current rice grain yield trends are insufficient to meet the global demand by 2050 (Ray et al., 2013). In addition, climate variability including debilitating heat-waves, drought, torrential rains and other weather extremes and change will exacerbate food insecurity in areas currently vulnerable to hunger and undernutrition. Further genetic improvement, combined with better improved management practices, could be indispensable to addressing those challenges and be a key factor for the adaptation of cropping systems to climate change for greater and more consistent crop production (Atlin et al., 2017; Bailey-Serres et al., 2019).

Recently, private-sector breeding approaches for accelerating genetic gain have been introduced to public research organizations in Asia such as the International Rice Research Institute (IRRI) (Cobb et al., 2019; Collard et al., 2019). These approaches include rapid generation advance, earlier multi-location trials, increased selection pressure for grain yield, an increased use of molecular breeding, and using variety product profiles. These breeding approaches are also gradually being introduced to SSA. It is therefore essential to take stock of past achievements in genetic gain to provide insights into what has been achieved and which challenges remain. Such a review could help to set the direction for future use of these new approaches to accelerate genetic gain in rice breeding programs in SSA.

This paper will review efforts and achievements of genetic improvement in rice with a focus on SSA, where rice demand is rapidly increasing. The objectives of this review paper are to (i) provide an overview of historical efforts of genetic improvement in SSA with a focus on improvement of rice yield, (ii) review field assessments of yield advantage of new rice varieties recently introduced, released or developed, compared to varieties released before, and (iii) summarize the results of adoption studies on these improved varieties. Lastly, the implications of this review for future efforts on genetic improvement as well as agronomy research in SSA is discussed.

In 2013, the book named ‘Realizing Africa's Rice Promise’ was published (Wopereis et al., 2013), which provides a comprehensive overview of state-of-the-art research on rice including genetic improvement (see Section 2 for genetic improvement). Recently, Saito et al. (2018) reviewed progress in genetic improvement of upland rice productivity in the tropics. However, these publications provided general overviews, and did not provide data on actual gains in yield observed in field experiments and impact studies, which are the primary focus in this paper. Before discussing such gains, we briefly describe the efforts of genetic improvement in rice in SSA and their key outputs – released varieties.