1 INTRODUCTION

The world is at a critical juncture. COVID-19 has wiped out years of progress on poverty, hunger, undernourishment, food security, etc. For the first time in over two decades, global poverty reduction has not only slowed down but also reversed due to the effect of COVID-19, conflict, climate change, etc. (Sankof, Zheng, & Lenert, 2020). For a quarter of a century (between 1990 and 2015), the rate of decline in global extreme poverty was an average of about 1 percentage point per year. However, the rate of decline between 2013 and 2015 dropped to 0.6 percentage point per year (Sankof et al., 2020). World Bank earlier projected that in 2020 more than 100 million additional people became the COVID-induced extreme poor (World Bank, 2020). Most of the COVID poor are from countries of Asia and Africa. Similarly, the prevalence of undernourishment significantly increased from 8.4% to around 9.9% between 2019 and 2020 (FAO, IFAD, UNICEF, WFP, 2021). Of the total 768 million undernourished people in 2020, more than half (418 million) live in Asia and more than one-third (282 million) in Africa. Likewise, of the 2.37 billion people facing moderate or severe food insecurity in 2020, more than half (1.2 billion) live in Asia and one-third (799 million) in Africa (FAO, IFAD, UNICEF, WFP, 2021). The Sustainable Development Goals (SDGs), the United Nations’ flagship plan to end poverty, hunger, etc., are not going to be achieved by 2030. Instead, it is projected that around 660 million people may still face hunger in 2030 (FAO, IFAD, UNICEF, WFP, 2021).

Rice is the single most important crop of the world as half of the world population eats rice every day. Rice supplies 20% of the world's dietary energy supply, while wheat and maize supply 19 and 5% respectively. In some Asian countries, rice provides over 70% of calorie supply (GRiSP, 2013). Moreover, rice is the key staple food for the world's poorest and undernourished people living in Asia and Africa as they cannot afford – or do not have access to – nutritious foods. Therefore, rice is considered one of the most strategic commodities for the world; not only linked with global food security but also closely connected with economic growth, employment, social stability and regional peace (Yadev & Kumar, 2018).

Rice production and research have met unprecedented challenges in recent years. Yield and total production have plateaued for many years in some major rice production countries (Shetty, Hegde, & Mahadevappa, 2013) while the demand from populations in poverty is ever increasing. Rice research can contribute to poverty alleviation, both directly (improved variety resulting in higher yield and higher profits for farmers) or indirectly (higher yield resulting in lower consumer prices). Therefore, in the review article, we have analysed the trends in rice research in the past three decades, particularly on the mega-projects that attempted to revolutionize rice yield, sustainability and quality of both Asian (Oryza sativa) and African (O. glaberrima) rice together with their impact on rice cultivation. We have also analysed the trends in population growth, rice cultivation, production, price and consumption along with their projections for 2030 and beyond. Furthermore, we have analysed recent trends in variety release using Bangladesh as an example. Finally, we have identified future challenges and priorities of rice research.

2 RISING TRENDS IN POPULATION GROWTH: KEY CHALLENGE FOR SUSTAINABLE DEVELOPMENT

Rapid population growth is the key challenge in the effort to combat hunger and malnutrition for sustainable development. Therefore, we have first examined the trends in population growth. The world population has grown from 2.5 billion in 1950 to 7.7 billion in 2019 (Figure 1a). The latest projections of UN population division predict the continuation of a similar rising trend for most parts of the century (United Nations Department of Economic and Social Affairs, 2019). The world's population is projected (based on medium fertility) to grow from 7.7 billion in 2019 to 8.5 billion in 2030 (10% increase), and further to 9.7 billion in 2050 (26%) and to 10.9 billion in 2100 (42%). The world's population could reach more than 17 billion in 2100, if current fertility and mortality rates remain unchanged throughout the rest of the century (no change variant; Figure 1a).

Global population growth rates vary greatly across regions. For example, the population of sub-Saharan Africa is projected to double by 2050 (99% increase), whereas the population of Europe and Northern America will remain almost the same (only a 2% increase). Moreover, both Europe and Latin America are expected to have declining populations by 2100. The population of Asia is expected to reach its peak (5.2 billion) in 2066, then start to decline. However, the population growth of Africa will remain strong for the rest of the century. Between 2020 and 2100, Africa's population is expected to increase from 1.3 billion to 4.3 billion (Figure 1b).

For the first time in history, the world's population is also expected to stop growing by the end of this century. By 2070, the global fertility rate is expected to fall below the replacement fertility rate (2.1 births per woman; Figure 1c). The total fertility rate of Africa is also expected to sharply decline, but still will remain slightly over the replacement rate till 2100 (Figure 1d). The projected growth in the world's population is expected to be concentrated in Africa and South Asia. Between 2020 and 2050, the global population is projected to increase by nearly 2 billion, with more than half in Africa, especially sub-Saharan Africa.

3 TRENDS IN GLOBAL RICE PRODUCTION

Rice is an Asian crop, as most of the rice is produced and consumed in Asia. In 1961, globally rice was cultivated on 115 million hectares of land. Between 1961 and 2019, global rice cultivation areas increased slightly over 40% (Figure 2a). Among different regions, the highest increase in rice cultivation areas was observed in Africa (516%), whereas in South America, the cultivation areas remained almost same (only 3.5% increase), despite a sharp rise (40–50%) in the late 70s (Figure 2b).

Global rice (paddy) production increased more than threefold between 1961 and 2019, from 215 million tonnes to 755 million tonnes; most of the production was from Asia (Figure 2c). Seven Asian countries, namely China, India, Indonesia, Bangladesh, Vietnam, Myanmar and Thailand together account for more than 80% of the global rice production. Among other regions, rice production has rapidly increased in Africa (almost ninefold increase) (Figure 2d) where leading producers are Egypt, Nigeria and Madagascar. In South America, rice production has increased from 7.1 million tonnes to 23.9 million tonnes in the past six decades. Brazil, Peru, Colombia and Ecuador are the top rice producers in South America.

Before the first green revolution, yields of both global and Asian rice (paddy) production were below 2 tonnes/hectare (Figure 2e). Between 1961 and 2019, global rice yield increased from 1.86 to 4.66 tonnes/hectare. Among different regions, rice yield always remained the lowest in Africa; still, it remains around 2 tonnes/hectare. Remarkably, rice yield in the developed regions like North America, Europe and Oceania is significantly higher than the global average (Figure 2f).

According to the latest projections by the OECD-FAO Agricultural Outlook, global rice production (milled equivalent) is expected to increase by 11.4%, reaching 567 million tonnes (Mt) by 2030 (OECD-FAO, 2021). The majority of the projected production increase (52 Mt, i.e. 89.7%) is expected to occur in Asia where India (+20 Mt), China (+6 Mt), Vietnam (+4.5 Mt) and Thailand (+2.5Mt) will account for the lion's share of the production increase. The production of rice in developed regions will remain stagnant or fall slightly. Rice cultivation area is also expected to decline in different countries, for example, China, Vietnam, etc., despite their production increase through yield improvements (OECD-FAO, 2021).

4 TRENDS IN RICE PRICE

The price of rice in the world market shows a rising trend. In 2020, the reference export price of rice (milled, 100% Grade B, fob Bangkok) was USD 512/tonnes, the highest price since 2013 (Figure 3). Rice price in past decades, particularly during 2005–2010, was volatile. In 2007–2008, rice price spiked in few months. Remarkably, the price hike at that time was not due to crop failure. Instead, the rise in rice price was explained by tight global rice supply, more specifically, trade restrictions by several major suppliers, panic buying by several large importers, a weak dollar, record oil prices, etc. (Childs & Kiawu, 2009). The prices of other major agricultural commodities such as wheat, corn and soybeans also spiked during the time. Global rice prices also spiked during the oil crisis in the early 70s. Oil/energy prices could directly affect prices of fertilizers and agrochemicals. The trends in rice price in the world market are clearly associated with the trends in fertilizer prices (Pearson's r = 0.87; Figure 3).

According to the latest projections, the trend in rice price will continue to increase until 2023, then the price is expected to decline to USD 476/tonnes (OECD-FAO, 2021). For rest of the years of the decade, the price in the world market will remain lower than the current price. Growing demand in Asia, Africa and the Middle East will increase the price, however, larger supply is expected to limit the price at USD 492/tonnes by 2030 (OECD-FAO, 2021).

5 TRENDS IN RICE CONSUMPTION

Change in demand of rice depends on the change in population, economic growth, per capita income, relative price of rice and other foods (Maclean, Dawe, Hardy, & Hettel, 2002). At low levels of income (below US$ 200 per capita) when meeting energy needs is a serious concern, rice is considered a luxury commodity. With increasing income, demand shifts from coarse grains and root crops (cassava, sweet potato, etc.) to rice. But at high levels of income, consumers substitute rice for high-cost quality food like fish, meat, fruits, vegetables, etc. (Maclean et al., 2002).

Per capita rice consumption in Asia is far higher than in all other regions. On average, more than 77 kg of rice is consumed annually by every person in Asia (Figure 4a). In 2018–2020, per capita rice consumption in other regions, i.e. Latin America and Caribbean, Africa, Europe, Oceania and North America, were 28.0, 27.4, 20.7, 13.5 and 6.3 kg respectively. By 2030, all these regions, except Africa will experience a smaller gain or loss of consumption per capita rice (Figure 4a). The demand for rice consumption in Africa will significantly increase, mainly in countries in West Asia and sub-Saharan Africa. The reasons for rapid increase in demand in Africa are fast population growth, rising urbanization, rural-to-urban migration, rise of income, change in food habits, particularly from root-based crops to rice, etc. The growth rate of rice consumption in West Asia and sub-Saharan Africa by 2030 will be 3% to 4% per year, (OECD-FAO, 2021).

In Asia, the consumption of rice in the middle- and high-income countries such as Japan, South Korea and Thailand shows a decreasing trend (Figure 4b). For instance, per capita rice consumption in Korea has decreased from 106 kg in 1975–1977 to 60 kg in 2020 and it is expected to decrease even further to 53 kg by 2030 (OECD-FAO, 2021). Oppositely, per capita rice consumption is increasing in some low-income Asian countries, such as Indonesia, the Philippines, Bangladesh and Vietnam. Per capita rice consumption in the Philippines has increased from 85 kg in 1975–1977 to 122 kg in 2020 and it is expected to increase even further to 129 kg by 2030 (Figure 4b).

6 TRENDS IN RICE RESEARCH

In the past three decades, different rice research institutes/universities around the globe took a large number of research initiatives to meet the future challenges to cope with the increasing demand. Most of the research projects/initiatives aimed to tackle one of the two fundamental challenges: (1) to break the barrier of yield ceiling or (2) to improve the yield sustainability. However, only few projects or initiatives were launched with the aim to release improved varieties. In this section, therefore, we have only analysed few unprecedented long-term (more than a decade long) research initiatives/projects of the past 30 years that were intended to revolutionize rice yield in both Asian and African rice, with their impact on rice cultivation. As climate change will be a big driver of the future of rice production in many rice-growing areas, we have also analysed a mega-project on the development of climate-smart rice varieties that attempted to stabilize rice production in rain-fed ecosystems in South Asia and Africa through the use of modern technology. Despite focusing on rice yield and yield sustainability, we have not ignored the mega-project that was launched to improve the nutritional value of rice to combat vitamin A deficiency. In the following sub-sections, we have presented these mega-projects in order of their inception (Table 1).

7 NEW PLANT TYPE (NPT) INITIATIVE

In 1989, IRRI initiated the super rice or new plant type (NPT) initiatives. The NPT initiative was based on ideotype breeding. The concept of ideotype breeding for yield increase in cereals was proposed by Donald (1968). Crop ideotypes are idealized plant type with a combination of characteristics favourable for photosynthesis, growth and grain production. Simulation models predicted a considerable increase in yield potential by modification in specific traits of the conventional plant type (Dingkuhn, Penning de Vries, De Datta, & van Laar, 1991).

Based on studies, breeders' experiences and the results of simulation modelling, the idealized new plant type was chosen to be moderately tall (90–100 cm) and strong (thick and sturdy stems) plants with a short growth duration (100–130 days), having few (3–4 tillers when direct seeded) but effective tillers (few unproductive tillers), desired leaf (thick, dark green and erect) and root (profuse and deeper) traits, with large panicle (200–250 grains per panicle; Khush, 1995). Most of these morphological traits were easy to include in the breeding programme. The goal was to develop an NPT with a yield potential 20–25% higher than the top performing inbred variety (Khush, 1995).

Bulu/javanica germplasms (tropical japonica) were identified as the potential source/donor of these traits. For instance, Daringan, Djawa Serang and Ketan Gubat were selected for donors of large panicle trait, whereas Sengkeu, Sipapak and Sirah Bareh for thick stems, after field screening of nearly 2000 germplasms (Peng, Khush, & Cassman, 1994). To achieve the desired plant type, 1000 of crosses were made, 100 of 1000 pedigree lines were analysed and 100 of NPT-TJ (tropical japonica) lines were evaluated in observational yield trials (Peng, Khush, Virk, Tang, & Zou, 2008).

Expectedly, the NPT-TJ lines had large panicles, few unproductive tillers and lodging resistance. However, the grain yield of these lines was not satisfactory, mainly due to poor biomass and poor grain filling (Peng et al., 2008). The probable causes of poor grain filling of NPT lines were panicle morphology (Peng et al., 2008), source limitation because of early leaf senescence (Ladha et al., 1998), the lack of apical dominance within a panicle (Yamagishi, Peng, Cassman, & Ishii, 1996), etc. Moreover, NPT-TJ lines were also susceptible to diseases and insects. Furthermore, the grain quality of NPT-TJ lines was not preferred in tropical and sub-tropical rice-growing countries.

In 1995, the NPT initiative was revised to NPT-2 with reduced number of grains per panicle (200–250 to 150) to increase the yield potential. Moreover, the revised NPT was also designed for the preferred grain quality along with multiple diseases and insect resistance to withstand disease and insect pressure in tropical environments. Therefore, NPT-TJ lines were crossed with elite indica varieties/breeding lines (Peng & Khushg, 2003). Over 400 NPT-IJ (Indica-Japonica) lines were evaluated in yield trials, and reduced panicle size and increased tillering capacity were achieved. Unfortunately, reports came out with a poor performance of NPT-IJ lines compared to the best inbred varieties (Peng et al., 2008).

The NPT initiatives were not successful in terms of number of varieties released or their coverage of areas of cultivation. However, two first-generation NPT lines (IR64446-7-10-5 and IR69097-AC2-1) were released as varieties (Diancho 1, 2 and 3) in Yunnan province of China and their yield was around 13 t/ha (1.0–1.5 tonnes more than local cultivars; Virk, Khush, & Peng, 2004). Two second-generation NPT lines (IR77186-122–2-2-3 and IR78581-123-3-2-2) were released as varieties in the Philippines in 2007 (NSICRc 158/Tubigan 13) and 2009 (NSICRc 222/Tubigan 18; Tonini & Cabrera, 2011). The maximum yield of Tubigan 13 and Tubigan 18 were 8.1 and 10 t/ha respectively. Remarkably, Tubigan 18 is still one of the most preferred rice varieties of Filipino farmers in irrigated lowland fields nationwide, both for dry and wet seasons (Bautista, 2018). Two NPT-IJ lines were also released as varieties in Indonesia.

The most important contribution of NPT initiatives was the introduction of japonica germplasm into otherwise indica breeding materials (Virk et al., 2004). The NPT lines were distributed to the breeding programmes throughout the world through the International Network for the Genetic Evaluation of Rice (INGER) nurseries. Before the NPT project, japonica gene pool was essentially excluded from the breeding programmes in tropics. The NPT lines are used in numerous hybridization programmes that successfully released varieties. For instance, one NPT-TJ line (IR66154-521-2-2) was one of the parents of Ciapus (cross-Memberamo//IR66154–521-2-2/Memberamo), a released variety in Indonesia in 2003 (Virk et al., 2004). Some other NPT-TJ lines including the aforementioned one were widely used in hybridization programmes in China and Vietnam. Finally, the launch of super hybrid rice project of China (described below) was also inspired by the NPT project.

8 GOLDEN RICE PROJECT

Deficiency of dietary micronutrients is a major source of morbidity and mortality worldwide, especially among infants, children and pregnant women. Vitamin A deficiency (VAD) has been recognized as a public health problem in developing countries. VAD causes numerous health complications including childhood blindness (Black et al., 2008). Unfortunately, VAD is prevalent among the poor who mostly depend on vitamin A-poor carbohydrate-rich diets. Therefore, food-based intervention or biofortification is considered a realistic and sustainable solution to prevent VAD. However, screening of rice germplasms did not find any rice cultivar capable of accumulating β-carotene (provitamin A) in the grain, although rice plants produce β-carotene in green tissues. Therefore, the introduction of the provitamin A pathway into the rice endosperm through genetic engineering is an excellent solution to prevent VAD.

In 1992, Ingo Potrykus (ETH Zurich, Switzerland) and Peter Beyer (University of Freiburg, Germany) proposed a plan to the Rockefeller Foundation to genetically engineer the provitamin A pathway into the rice endosperm. Within 7 years of their proposal, they proved that β-carotene can be produced in the rice grain (Ye et al., 2000). The colour of the transgenic rice grain turned yellow–orange or golden, hence the project was called Golden Rice project. Besides the Rockefeller Foundation, numerous other organizations also funded the project. It included the Bill & Melinda Gates Foundation, USAID, the Philippine Department of Agriculture, HarvestPlus, the European Commission, Swiss Federal Funding and the Syngenta Foundation (www.goldenrice.org).

For the development of Golden Rice, Ye et al. (2000) used Agrobacterium-mediated transformation to introduce transgenes into Taipei309, a japonica rice variety. The genetic construct consisted of a plant phytoene synthase (psy) originating from daffodil (Narcissus pseudonarcissus), a bacterial phytoene desaturase (crtI) originating from Pantoea ananatis and lycopene ≈-cyclase (lcy) from daffodil (Ye et al., 2000). Transgenic lcy was not necessary for β-carotene accumulation in the grain as native lcy was also expressed in the endosperm. Low carotenoid content (1.6 μg/g) was a big concern of the first-generation golden rice (GR1) as it may not be able to combat VAD. For the development of the second-generation golden rice (GR2) with higher β-carotene levels, GR2 was developed using the bacterial phytoene desaturase (crtI) originating from P. ananatis, phytoene synthase of maize (instead of daffodil) and Asanohikari variety (instead of Taipei309). GR2 accumulated up to 37 μg/g carotenoids, of which 31 μg/g was β-carotene (Paine et al., 2005).

GR2 was donated to International Rice Research Institute (IRRI) in 2006 to help combat VAD in developing countries including the Philippines, Bangladesh and Indonesia. Breeders at the Philippine Department of Agriculture – Philippine Rice Research Institute (DA-PhilRice), the Bangladesh Rice Research Institute (BRRI) and the Indonesian Center for Rice Research (ICRR) are developing Golden Rice versions of existing farmer-preferred mega-varieties like IR64, PSBRc82, BR29, etc. For instance, the transgenic GR2 BRRI dhan29 yielded an average of 7.0 t/ha under confined field conditions in Bangladesh. There was no significant difference between the transgenic and non-transgenic BRRI dhan29 (Biswas et al., 2021).

A safety assessment of Golden Rice revealed that the introduction of DNA occurred without modifications (Oliva et al., 2020). Neither phytoene synthase nor carotene desaturase proteins display sequence homology with allergens or toxins. Moreover, both proteins were rapidly digested in simulated gastric fluid and their enzymatic activity was inhibited upon heat treatment (Oliva et al., 2020). In 2018, Golden Rice received food safety approvals from Food Standards Australia New Zealand, Health Canada and U.S. Food and Drug Administration (Owens, 2018). In July 2021, the Philippine Department of Agriculture-Bureau of Plant Industry (DA-BPI) approved the commercial cultivation of Golden Rice in the Philippines. Therefore, the Golden Rice is going to be the world’s first commercialized GM biofortified crop.

9 NEW RICE FOR AFRICA (NERICA) PROJECT

The cultivated rice of Africa (Oryza glaberrima) is adapted to the African environment, but prone to lodging and grain shattering. Asian rice (O. sativa), on the other hand, has high yield potential but is poorly adapted to African conditions. In 1994, the West Africa Rice Development Association (WARDA) (currently the AfricaRice Centre) successfully hybridize African rice with Asian rice to develop interspecific hybrid suitable for African conditions. The interspecific cultivars were called New Rice for Africa, simply NERICA. NERICA project was funded by the Japanese government, the African Development Bank and the United Nations Development Programme (WARDA, 2001).

The idea of interspecific rice hybrid was not new but previous attempts of hybridization between the two species had failed as the resulting offspring were all sterile. In early 1990s, WARDA breeders successfully overcome hybrid sterility. To overcome the limitations, breeders employed embryo rescue technique and subsequent back-crossing with the sativa parent for several cycles (WARDA, 2001). After the fertility restoration, anther culture technique was used for the production of double haploids. NERICA rice was first field tested in 1994 and the first NERICA variety was released in Côte d'Ivoire in 2000 (Mohapatra, 2019). The development of NERICA rice was a major scientific breakthrough in African rice development, gaining several international awards including World Food Prize in 2004 to Dr. Monty Jones (a senior breeder at AfricaRice) (WARDA, 2001).

Being interspecific hybrid, NERICA varieties display heterosis, i.e. better yield or stress tolerance than either parent. There are 82 NERICA varieties – 18 upland, 60 rain-fed lowland and 4 irrigated varieties (Mohapatra, 2019). Beside good yields, upland NERICA varieties are early maturing (in 75–100 days) and are tolerant to major local stresses. NERICA-4, tolerant to drought and phosphorus deficiency, is the most widely adopted upland variety, grown in more than 10 countries in sub-Saharan Africa. Among lowland varieties, NERICA-L-19 is the most widely adopted lowland NERICA in sub-Saharan Africa (Mohapatra, 2019). An impact study in 2017 showed that the adoption of NERICA varieties significantly reduced poverty (lifted about 8 million people out of poverty) and ensured food security to 7.2 million people in 16 African countries (Arouna, Lokossou, Wopereis, Bruce-Oliver, & Roy-Macauley, 2017). The success of NERICA varieties has now expanded beyond the African continent. NERICA varieties are also being used by breeders in varietal improvement programmes in Bangladesh, China, India and several other countries around the world (https://www.africarice.org/nerica). For example, Bangladesh Rice Research Institute and Bangladesh Institute of Nuclear Agriculture developed and released three drought-tolerant rice varieties from NERICA-10 for aus season. These are BRRI dhan 82, BINA dhan 19 and BINA dhan 21 (BRRI, 2020).

10 SUPER HYBRID RICE PROJECT

In 1996, the Ministry of Agriculture, China, launched a national key project on ‘China Super Rice Research’ to ensure food security of growing population in China. In 1998, Prof. Yuan Longping proposed a strategy for developing super hybrid rice. The strategy combined two ways of yield increase, i.e. morphological improvement through the ideotype approach along with the utilization of intersubspecific heterosis (Cheng et al., 2007).

The morphological improvement was primarily aimed at three aspects, (1) tall erect-leaf canopy for higher photosynthetic efficiency, (2) lower panicle position for lodging resistance and (3) larger panicle size for higher yield potential (Cheng et al., 2007). The utilization of heterosis or hybrid vigour for cross-pollinated crops such as maize and sorghum has been effectively exploited since 1950s. Super hybrid rice project successfully utilized interspecific heterosis in rice. Indica/japonica hybrids have a very large sink and rich source, and the yield potential of indica/japonica hybrids is 30% higher than commercially used indica/indica hybrids.

The super hybrid rice project was inspired by IRRI's NPT project, although the NPT project was not very successful in terms of yield increment and number of varieties released. Unlike the NPT project, source sink relations were well balanced in the super hybrid rice project. In general, the yield advantage among super hybrid rice resulted from a high rate of flag leaf photosynthesis, slow leaf senescence, efficient remobilization of carbohydrates, higher leaf area density, greater specific leaf weight and high root activity. The most notable point of improvement of plant architecture was achieved by emphasizing more on the top three leaves and panicle position within the canopy, most importantly, the emphasis on the upper three leaves (long, erect, narrow, V-shaped and thick) (Peng et al., 2008).

The super hybrid rice project of China was a very successful project in terms of achievement of project goals, number of varieties developed/released and cultivation areas of these varieties. The yield targets of all four initial phases were achieved. Moreover, the yield target of phase V (16 tonnes/ha) was also achieved in a 6.6-hectare demonstration field in Gejiu country, Yunnan province, in 2015 (Yuan, 2017). As of 2017, the Ministry of Agriculture of China has certified 110 indica and 56 japonica super hybrid rice cultivars. The annual planting area of super rice cultivars in recent years is maintained at about 9 million hectares in China (Chen, Xu, & Tang, 2017).

11 STRESS-TOLERANT RICE FOR AFRICA AND SOUTH ASIA (STRASA) PROJECT

Similar to other crops, rice productivity is also dependent on weather and climate conditions. Climate models predict that climate change is likely to change numerous parameters that might lead to unfavourable growing conditions in crop fields. For instance, climate change is likely to change precipitation patterns, resulting in more droughts or flooding (Trenberth, 2011). The world is experiencing rising global temperatures (GISTEMP Team, 2022). Rising temperatures and sea levels will result in more heat stress and saline intrusion/flooding respectively. The occurrence of more frequent and severe weather extremes (droughts/flooding/wildfires, etc.) is well evident across the globe (Cai et al., 2014). To cope with the challenges of climate change, therefore, it is essential to develop climate-smart rice varieties with concerted research effort (Bin Rahman & Zhang, 2022).

In 2007, IRRI in collaboration with former West Africa Rice Development Association (WARDA; current AfricaRice) launched a mega-project, funded by the Bill & Melinda Gates Foundation (BMGF), to develop stress-tolerant rice for poor farmers in Africa and South Asia or STRASA (AfricaRice, 2022; IRRI, 2022). The general goal of the STRASA project was to reduce poverty for at least 18 million farmers and stabilize rice production in rain-fed ecosystems in South Asia and Africa through the use of modern technology. The STRASA project was focused to develop tolerant varieties of five major abiotic stresses: drought, submergence, salinity, iron toxicity and low temperature. Stress-tolerant varieties should have at least 1 t/ha yield advantage over existing varieties under stress conditions and no yield penalty if no stress occurs (IRRI, 2022). The project covered three countries in South Asia (i.e. Bangladesh, India and Nepal) and 18 countries in Africa (i.e. Benin, Burkina Faso, Burundi, Côte d'Ivoire, Ethiopia, Gambia, Ghana, Guinea, Kenya, Madagascar, Mali, Mozambique, Nigeria, Rwanda, Senegal, Sierra Leone, Tanzania and Uganda; AfricaRice, 2022; IRRI, 2022).

The STRASA project utilized a marker-assisted back-crossing (MAB) approach to transfer stress-tolerant QTLs/genes into improved cultivars. Major QTLs such as SUB1 (Xu & Mackill, 1996), Saltol (Bonilla, Mackell, Deal, & Gregorio, 2002), qDTY_1.1, qDTY_2.2, qDTY_3.1, qDTY_3.2, qDTY_6.1, and qDTY_12.1 (Kumar et al., 2014) conferring submergence, salinity and drought tolerance, respectively, were introgressed into farmer-preferred, widely accepted varieties for faster adoption of the improved cultivars. CR Dhan 801 is an example of a recent release of climate-smart rice variety under the STRASA project (Pradhan et al., 2019). This variety is tolerant to both submergence and drought. It was released in 2018 for cultivation in the Indian states of Odisha, West Bengal, Uttar Pradesh, Andhra Pradesh and Telangana. For the CR Dhan 801 development, SUB1 and qDTY1.1, qDTY2.1 and qDTY3.1 were introgressed into the mega-variety, Swarna. Under normal conditions, CR Dhan 801 produces about 6.3 t/ha yield, whereas under submergence and drought conditions, it yields 4 and 2.9 t ha-1 respectively (Pradhan et al., 2019). Beside farmer-preferred varieties, the project also utilized some innovative approaches like participatory variety selection, head-to-head trials, crop cafeteria, cluster demonstrations, etc. for faster varietal adoption. The project was completed in three phases: Phase I (March 2008 to February 2011), Phase II (March 2011 to February 2014) and Phase III (March 2014 to March 2019).

Thirty stress-tolerant rice varieties (STRV) were released in the STRASA Phase II. Among the STRVs, 16 varieties were released in South Asia and 14 in sub-Saharan Africa. In the Phase III, 71 STRVs were released in India (58), Bangladesh (6) and Nepal (7). These STRVs were tolerant to drought (27), submergence (23), salinity (12) and nine tolerant to multiple stresses like drought and heat (2), drought and submergence (6) and submergence and salinity (1) (IRRI, 2022). In Africa, 46 STRV were released in the phase III. These included drought-tolerant varieties (11), flooding-tolerant varieties (6), salt-tolerant varieties (10), iron toxicity-tolerant varieties (12) and cold-tolerant varieties (9) (AfricaRice, 2022).

The STRASA was a remarkably successful project. A total of 182 stress-tolerant rice varieties were released in various countries in South Asia and in sub-Saharan Africa (for the list of the varieties, see at https://strasa.irri.org/varietal-releases). Beside the release of STRVs, the STRASA project also trained 74,000 farmers and scientists – including 19,400 women farmers – in producing good-quality seeds through its capacity-building component. Moreover, STRASA has successfully reached 18 million farmers, produced over 565,000 tonnes of climate-smart seeds and covered 5 million hectares (IRRI, 2022). Recent impact analysis showed that the adoption of STRVs, more specifically, submergence-tolerant rice varieties, significantly increased rice yield and profit of poor farmers in Bangladesh and India (Bairagi, Bhandari, Kumar Das, & Mohanty, 2021; Raghu, Veettil, & Das, 2022).

12 GREEN SUPER RICE PROJECT

Despite significant yield increase, hybrid rice comes with several constraints such as higher seed cost, higher incidence of pests and diseases and poor grain quality. These constraints limited hybrid rice adoption among poor and marginal rice farmers in less developed countries. Poor and marginal rice farmers have limited land and financial capacity. Unfortunately, climate change poses a big threat to these smallholder farmers due to the unpredictable weather patterns, i.e. increase the risk of crop losses. Numerous other general constraints such as high input-dependent performance of super inbred/hybrid rice, excessive use of fertilizers and pesticides, growing water scarcity in agriculture and poor soil fertility of marginal lands make rice production neither environment friendly nor sustainable.

To address these concerns, the Green Super Rice (GSR) concept was proposed in 2005 to promote resource-saving and environment-friendly rice production (Yu, Ali, Zhang, Li, & Zhang, 2020; Zhang, 2007). The aims of the GSR concept were to develop new rice varieties with various green traits, keeping high and stable yield. Simply put, green stands for an environment-friendly rice production system with fewer inputs, whereas super represents high and stable yield. The GSR project was launched in December 2008 to develop rice cultivars that can produce high and stable yields with less inputs (water, fertilizers and pesticides) or under unfavourable environmental conditions. The GSR project was led by the Chinese Academy of Agricultural Sciences (CAAS) in partnership with IRRI and AfricaRice. The Bill and Melinda Gates Foundation and the Ministry of Science and Technology of China funded the project (Yu et al., 2020).

For the development of GSR varieties, molecular breeding strategy was used. The strategy was a fusion of back-cross breeding, marker-assisted selection (MAS) and gene/QTL pyramiding. The procedure starts with hybridization of elite germplasms of the desired traits (donor) with widely adaptable recurrent parents (recipient), followed by successive back cross (BC) with the recurrent parents, then strong phenotype selection of segregating BC populations for multiple stresses in different target countries/environments, genetic characterization of selected introgression lines using high-throughput genotyping and finally, development of superior GSR varieties by designed QTL pyramiding and/or molecular recurrent selection (Li & Ali, 2017). This molecular breeding strategy significantly reduced the duration of breeding cycle from 8–10 years to 4–6 years.

Up to 2018, 66 GSR varieties had been registered in China. The cumulative planting area of these cultivars exceeded over 10.87 million hectares across five main rice-growing regions of China from 2014 to 2018 (Yu et al., 2021). Currently, 59 GSR varieties tailored to individual conditions have been introduced in 16 countries in South Asia (Bangladesh, India, Pakistan and Sri Lanka), Southeast Asia (Indonesia, Laos, Cambodia, the Philippines and Vietnam) and Africa (Mozambique, Uganda, Rwanda, Nigeria, Senegal and Mali). The cumulative cultivation areas including demonstration plots in these countries have reached 6.1 million hectares (Yu et al., 2021). In addition to these released varieties, more than 100 cultivars are still in the pipeline (Yu et al., 2020).

The adoption of GSR varieties significantly increased the yield and net income of farmers in the Philippines, Bangladesh and Mozambique (Kodama et al., 2022; Yu et al., 2020). An impact analysis showed that, in the Philippines, the extent of increase in yield and economic output were 0.89–1.83 tonnes and Philippine peso 9400.26 (current equivalent of USD 179.16) per hectare respectively (Yorobe et al., 2016). A very significant yield increase was observed in a recent survey in three stress-prone, sub-Saharan regions of Mozambique after the adoption of GSR varieties. The yield advantage of GSR adopting smallholder farmers was about 10 times higher than GSR non-adopting farmers (Kodama et al., 2022).

13 C₄ RICE PROJECT

C₄ rice project is a collaborative research initiative launched in 2008 after successive workshops on the possibility of yield increase in rice manipulating photosynthetic mechanisms (Sheehy, Mitchell, & Hardy, 2000; Sheehy, Mitchell, & Hardy, 2007). To assemble necessary technologies and knowledge, International Rice Research Institute (IRRI) formed an international consortium of scientists to chart and conduct the research (Sheehy et al., 2007). The objective of the project was to develop a new type of rice with improved photosynthesis capacity by incorporating C₄ photosynthetic pathway. Rice has a C₃ photosynthetic pathway. C₄ plants such as maize are more efficient at carbon assimilation than C₃ plants. Successful introduction of a C₄ photosynthetic pathway in rice has been predicted to increase rice yields by up to 50% (Hibberd, Sheehy, & Langdale, 2008; Sheehy et al., 2007). After the first two phases of the project, the project leadership was handed over to the University of Oxford, UK. The Bill & Melinda Gates Foundation awarded a $15 million grant to the University of Oxford for the ongoing Phase IV to develop a prototype for C₄ metabolism (https://c4rice.com/the-project-2/our-history/).

Redesigning rice photosynthesis requires the identification of the genes of C₄ photosynthesis and installation in rice. The goal appears simple and straightforward. However, it is not easy as it sounds. It includes numerous steps such as identification of cassette of C₄ genes, development of molecular toolbox to transform dozens of genes into rice, optimization of their expression and C₄ function and finally, integration into local varieties (Hibberd et al., 2008; Sheehy et al., 2007; Von Caemmerer, Quick, & Furbank, 2012). Besides biochemical traits, introduction of the C₄ pathway into rice also requires the manipulation of anatomical traits, commonly known as Kranz anatomy. To introduce Kranz anatomy into rice, significant change in vein spacing pattern and cell-specific differentiation are required for increased photosynthetic capacity (Sedelnikova, Hughes, & Langdale, 2018). Similarly, the activation of chloroplast development in the bundle sheath cells is prerequisite for the efficient C₄-type photosynthesis. Considering altogether, the C₄ rice project is an extremely challenging project on crop plants.

Initially, it was estimated that constructing C₄ rice might take at least 15 years to complete (Sheehy et al., 2007), whereas release to farmers' fields might be ready by 2030. Meanwhile, the C₄ rice prototype that expresses all known genes required to support the C₄ biochemical pathway would be ready for fine-tuning of biochemistry and anatomy (Von Caemmerer et al., 2012). Despite significant progress, the development of a fully functional C₄ rice prototype still requires a significant amount of time. It would not be unlikely to exceed 2030 just to develop a fully functional prototype. Recently, for the first time in using a single construct, five genes from maize that code for five enzymes in the C₄ photosynthetic pathway were installed into rice plant (Ermakova et al., 2021), although the transgenic plants were not yet working efficiently like C₄ plants. Moreover, dozens more genes need to be installed in rice for a functional C₄ prototype. The development of fully functional bundle sheath remains another big challenge. Constitutive expression of 60 candidate genes found individually was not sufficient to induce C₄-like vein patterning or cell-type differentiation in rice (P. Wang et al., 2017). Moreover, the complete regulatory network of Kranz anatomy is yet to be revealed. Simply, the more we understand the molecular insight of C₄ pathway, the more we realize that this approach is an extremely challenging biotechnology, demanding substantial fine-tuning of both biochemistry and anatomy as well (Furbank, 2016).

It appears that necessary knowledge and technologies required for C₄ rice development are not mature enough for the successful installation of dozens of genes simultaneously in the correct cell type, at the right time, to the right level. There are also growing concerns on whether successful installation of C₄ photosynthetic pathway can lead to a drastic yield increase (Araus, Sanchez-Bragado, & Vicente, 2021). After successful installation and desired yield performance (even if everything turns successful), the engineered rice must have to go through rigorous testing for national and international safety and nutritional standard. More than two decades have passed since Golden rice (vitamin A enriched genetically modified rice) was developed; still golden rice is not in farmers' fields. Therefore, it may take a long time, perhaps decades, before the C₄ variety is released to farmers.

14 RECENT TRENDS IN VARIETY RELEASE USING BANGLADESH AS AN EXAMPLE

Rice is currently grown in over a hundred countries. Earlier, most of the rice-growing countries had a tendency to release IRRI-developed varieties directly. Later, IRRI stopped releasing varieties directly, instead, national institutes utilized IRRI-developed breeding lines as parents in the local hybridization programmes. We deliberately selected Bangladesh to identify the recent trend in variety release because in Bangladesh, rice provides very high caloric demand (nearly 70%). Poor people in Bangladesh practically live on rice. Moreover, Bangladesh is not only one of the largest rice-producing countries but also one of the most vulnerable countries of climate change (Bin Rahman & Zhang, 2018).

Bangladesh is considered the centre of rice diversity (Bin Rahman & Zhang, 2013, 2016, 2018, 2022). The distribution/cultivation of two deep-water rice types (Bhadoia/Ashwina and Rayada) is confined to Bangladesh (and some adjacent Indian states) (Bin Rahman & Zhang, 2013). Previous studies identified these two special types of deep-water rice as two distinct varietal groups (groups III and IV) (Glaszmann, 1987). Deep-water rice-specific variant of SEMIDWARF1 (SD1-DW) is predominantly available in deep-water rice varieties of Bangladesh, whereas deep-water rice in other countries only harbour SNORKEL1/2 (Kuroha et al., 2018). Therefore, a narrow geographic region centred around Bangladesh is also considered as the centre of origin of Asian rice (Bin Rahman & Zhang, 2018).

In Bangladesh, most of the rice varieties are released by Bangladesh Rice Research Institute (BRRI) and Bangladesh Nuclear Research Institute (BINA). Since 2010, BRRI and BINA released 54 and 17 rice varieties respectively. The release of high-yielding varieties with numerous add-on features is the common trend in Bangladesh. The common add-on features are stress tolerance, biofortification (high zinc, iron and protein), good eating quality, aromatic and slender white grain, etc.

Flood is the most prevalent and recurring abiotic stress in Bangladesh and millions of hectares of land are flood prone (3.02 million hectares – river flood prone, 1.47 million – flash flood prone and 2.18 million – tidal surge) with varied intensities and depth (Bin Rahman & Zhang, 2018). BRRI and BINA released six flash flood-tolerant (submergence-tolerant) rice varieties (BINA 11, BINA 12, BINA 23, BRRI dhan 51, BRRI dhan 52 and BRRI dhan 79) that possess Sub1A gene (BRRI, 2020). These varieties can withstand 14–21 days of complete submergence. A recent impact study showed that approximately 42% of the farmers adopted Sub1 rice varieties (in the sampled areas) in northwest Bangladesh and the adoption of Sub1 rice varieties increased rice yield (6.0% higher), profit (55.0% higher) and household consumption (15.0% higher) (Bairagi et al., 2021). Remarkably, BRRI also released a high-yielding deep-water rice variety (BRRI dhan 91) for aman season in lowland areas. For the areas of tidal surge, two rice varieties (BRRI dhan 77 and BRRI dhan 78) were released where BRRI dhan 78 can also withstand salinity (6–9 dS/m) during seedling and flowering stages (BRRI, 2020).

Over half a million hectares of cultivable land, mostly from the coastal belt of Bangladesh, suffer from mild to severe soil salinity (Bin Rahman & Zhang, 2018). BRRI and BINA released 11 salt-tolerant rice varieties with different extents of salinity tolerance for the salt-affected areas. For instance, BRRI dhan 99, a recent release (in 2020) for boro season, can tolerate soil salinity up to 14 dS/m at the seedling stage and 8–10 dS/m at the reproductive stage. Several million hectares of cultivated lands of Bangladesh are drought prone, and for these areas seven drought-tolerant rice varieties were released since 2010. For example, BRRI 83, a short duration (100–105 days) drought-tolerant variety, released in 2017 for aus season, particularly for drought-prone areas in Kushtia, Jhinedha, Magura, etc., whereas BRRI 71, released in 2015 for aman season, can withstand 21–28 rainless days during the reproductive stage. Some recently released varieties can withstand multiple abiotic stresses (BRRI, 2020). For example, BINA 23 has the capacity to grow in saline (8.0 dS/m) and flash flooded area for up to 15 days in submerged conditions, whereas BRRI 55 (for both aus and boro seasons) can tolerate salinity (8–10 dS/m for 3 weeks), mild drought and cold.

The release of biofortified rice varieties is another recent trend in Bangladesh (Calayugan et al., 2021). BRRI has released seven high zinc rice varieties under the HarvestPlus initiative (BRRI dhan 62, BRRI dhan 64, BRRI dhan 72, BRRI dhan 74, BRRI dhan 84, BRRI dhan 100 and BRRI dhan 102; BRRI, 2020; Calayugan et al., 2021; http://www.brri.gov.bd/). BRRI dhan100 contains nearly 26 mg of zinc where popular rice varieties usually contain 15–16 mg/kg zinc and it has a high yield potential of nearly 7.7 tonnes per hectare. Similar to BRRI, BINA also released three high zinc rice varieties. Besides zinc, BRRI also released several varieties with high Fe (e.g. BRRI dhan 84-Fe 10.1 mg/kg). A recent release, BRRI dhan 96, has significantly higher protein (10.8%) content than popular varieties (7.0–8.0%). BRRI dhan 69, a released variety in boro season, has a marginally low glycaemic index (GI) (54.9). Finally, vitamin A-enriched rice (Golden rice) is now awaiting final approval in Bangladesh.

Similar to Bangladesh, similar trends of breeding goal and varietal release are observed in other major rice-growing countries. India also trended to release high yielding rice varieties with additional features like drought tolerance (Sahbhagi Dhan, DRR dhan 42, 43 DRR dhan, DRR dhan 44, etc.), submergence tolerance (Swarna-Sub1, Samba Mahsuri-Sub1, IR64-Sub1 and CR-1009-Sub1) and salt tolerance (CSR 43, CSR 46, CSR49, CSR52, etc.), with high protein (CR Dhan 310) and high zinc (DRR Dhan45 and Chhattisgarh Zinc Rice-1; Calayugan et al., 2021). Several other countries also released Sub1 rice varieties, such as IR64-Sub1 in the Philippines, Ciherang-Sub1 in Indonesia and Nepal and FARO 66 and 67 in Nigeria. The release of zinc biofortified rice varieties is also a recent trend in some countries like Indonesia (Inapari Nutri Zn), the Philippines (NSIC Rc 460), Colombia (Fedearroz BioZN-035), Bolivia, El Salvador, etc. (Calayugan et al., 2021).

15 FUTURE CHALLENGES AND PRIORITIES

Climate change is projected to undermine rice productivity (IFPRI, 2009). A field study with IR8, the revolutionary variety of first green revolution, has demonstrated that environmental changes can significantly decline rice productivity of the same variety over time (Peng et al., 2010). To cope with climate change, rice varieties need to be developed with increased yield potential and resistance to multiple environmental stresses, Therefore, sustainable rice production using less land, less water, less labour, fewer chemical inputs and the reduction in greenhouse gas emissions in rice cultivation are the two major challenges.

Another major challenge in rice production is to achieve the dual goal of increasing grain yield and saving water. Fresh water is becoming increasingly scarce for agriculture where rice consumes about 80% of the total irrigated freshwater resources in Asia. Several agricultural practices, such as post-anthesis controlled soil drying, alternate wetting and drying irrigation and non-flooded straw mulching cultivation, could substantially enhance water use efficiency and maintain or even increase the grain yield of rice (Yang & Zhang, 2010). Interestingly, alternate wetting and drying could be a key strategy to reduce both water consumption and greenhouse gas emissions in rice cultivation (Livsey et al., 2019).

Low nitrogen use efficiency (NUE) is a serious problem in some rice production areas, such as in China. The use of excess amount of N fertilizer causes low NUE. Super hybrid rice is well adapted to high N fertilizer conditions. Therefore, farmers tend to apply excess amount of N fertilizer to harvest a higher grain yield, although super rice per se is not responsible for the low NUE (Wang & Peng, 2017). Average NUE between the GSR candidate varieties and the super hybrid rice showed no significant difference (Huang, Sun, Yuan, Peng, & Wang, 2018). Previous studies showed that it is possible to achieve a high grain yield and high NUE by applying precision agricultural practices, more specifically, improvements in N fertilizer management practices such as site-specific N management (SSNM) and real-time N management (RTNM) (Dobermann et al., 2002; Peng et al., 2006; Wang & Peng, 2017).

Rice grain quality improvement has always been one of the top priorities in rice breeding programmes. Similar trends will also continue in the future as rice quality is a primary determinant of its market price and consumer acceptance. Biofortified rice with enriched micronutrient content will be a common trend of upcoming varietal release. Understanding the genetic control of eating and cooking quality of some elite varieties like Koshihikari, Basmati, IR64, Khao Dawk Mali, etc. and transfer the properties into popular varieties is going to be one of the key priorities of rice research in near future.

Similarly, the development of improved inbred and hybrid rice varieties having both high yielding and superior quality would be a key area of research in the future. Simultaneous improvement of grain yield and quality is possible through precise ideotype breeding (rational design breeding). However, pyramiding multiple desirable traits require precise genetic dissection of complex agronomic traits and high-resolution chromosome haplotyping (Bao, 2019; Qian, Guo, Smith, & Li, 2016; Zeng et al., 2017). Although results of a recent attempt seem promising (Zeng et al., 2017), pyramiding complex traits like yield and quality require much more research.

The development of rice varieties with slow digestible starch is gaining significant attention due to the prevalence of diabetes in Asia. However, genetic manipulations of glycaemic index (GI) trait come with yield penalty and undesirable cooking and textural properties (Jukanti, Pautong, Liu, & Sreenivasulu, 2020). Moreover, the lack of high-throughput phenotyping technique of GI is the key limitation for the development of low GI rice varieties.

Rice is grown on primarily 250 million Asian farms, most of them are smaller than 1 hectare (Maclean et al., 2002). To make rice cultivation more cost-effective and to enhance the agricultural productivity of smallholder rice farmers, mechanization and improved postharvest management are necessary as mechanized dry direct seeding can significantly reduce production costs. Therefore, farm mechanization is expected to be a future trend in rice cultivation.

CONFLICT OF INTEREST

Both authors declare that they have no competing interests.