Yield Potential Trends of Tropical Rice since the Release of IR8 and the Challenge of Increasing Rice Yield Potential

In 1966, IRRI released IR8, the first high-yielding modern rice cultivar for the irrigated tropical lowlands. Since the release of IR8 in 1966, 42 additional indica rice (Oryza Since then, the primary emphasis of rice improvement sativa L.) cultivars developed by the International Rice Research Institute (IRRI) for the irrigated and favorable rainfed lowlands have has been directed towards incorporation of disease and been released in the Philippines. The maximum yield of IR8 has been insect resistance, shortening of growth duration, and reduced by about 2 Mg ha21 during the past 30 yr. Empirical breeding improving grain quality (Khush, 1990). As a result, 42 for population improvement within the indica germplasm has resulted indica inbred cultivars have been released for the irriin the maintenance of rice yield potential in the tropics of about 10 gated and favorable rainfed lowland rice systems in the Mg ha21. To break the yield barrier, several approaches are being Philippines. In the late 1960s and early 1970s, yield of explored. These include development of a new plant type (NPT) 9 to 10 Mg ha21 was often reported for IR8 and other with low tillering capacity and large panicles from tropical japonica early IRRI cultivars under favorable irrigated condigermplasm and exploitation of heterosis through intervarietal and tions in the Philippines (De Datta et al., 1968; Chandler, intersubspecific hybrids. Hybrid rice between indicas increased yield 1969; Yoshida and Parao, 1972). The yield potential of potential by about 9% under the tropical conditions. The higher yield potential of indica/indica hybrids compared with indica inbred culticurrent high-yielding inbred rice cultivars grown with vars was attributed to the greater biomass production rather than irrigation in the tropical lowlands is 10 Mg ha21. Thereharvest index. New plant type breeding has not yet improved yield fore, rice yield potential has remained almost constant potential due to poor grain filling and low biomass production. Factors in the tropical environments. The theoretical potential that cause poor grain filling and low biomass production of the NPT yield has been estimated at 15.9 Mg ha21 in these envilines have been identified. Selecting parents with good grain filling ronments based on the total amount of incident solar traits, introduction of indica genes into NPT’s tropical japonica backradiation during the growing season (Yoshida, 1981). ground, and a refinement of the original NPT design are expected to On the basis of this estimate, there appears to be a large improve the performance of the NPT lines. Further enhancement in gap between the yield potential of the best available yield potential may be possible from use of intersubspecific heterosis rice cultivars and the maximum theoretical yield. At between indica and NPT lines. issue is whether this gap can be narrowed by genetic crop improvement. Despite the apparent yield barrier, the quest for G rice production must reach 800 Tg of paddy higher rice yield potential continued. Ku et al. (1999) rice to meet projected demand in 2025, which is explored the possibility of converting rice’s C3 photosyn266 Tg more than rice production in 1995 (Table 1). thetic pathway into the C4 pathway by introducing the This additional rice must come mainly from irrigated maize (Zea mays L.) phosphenolpyruvate carboxylase land in Asia because improving rice yields in most (PEPC) gene into rice. Zhang et al. (1996) introduced rainfed regions is constrained by drought, flooding, and a gene from Escherichia coli encoding for ADP-glucose poor soil quality (Cassman, 1999). Moreover, the scope pyrophosphorylase (AGP) enzyme into rice with the for expansion of irrigated rice area is limited. In fact, goal of increasing starch biosynthesis during seed develcultivated rice area is decreasing in many Asian counopment. There is preliminary evidence that quantitative tries as a result of urbanization and industrialization. trait loci (QTL) alleles from a wild rice Oryza rufipogon Therefore, the average yield of Asia’s irrigated rice land Griffith when introduced to cultivated rice, may increase must increase from 5.0 to 8.5 Mg ha21 in the 30-yr period the yield potential by as much as 20% (Xiao et al., 1998). from 1995 to 2025. A mean yield of 8.5 Mg ha21 is very Setter et al. (1995) proposed to reduce panicle height in close to the estimated climate-adjusted yield potential the canopy to increase canopy light interception, canopy of present rice cultivars in the major rice-growing areas photosynthesis and grain yield. In the past 10 yr, IRRI (Matthews et al., 1995). Rice cultivars with higher yield scientists initiated two new approaches to increase yield potential must be developed to maintain an exploitable potential: (i) the development of a new plant type (NPT) gap between yield potential and average farm yields. from tropical japonica germplasm, and (ii) development of indica hybrid rice for the tropical environments. The purpose of this paper is to evaluate the yield potential trends of indica inbred cultivars since the release of IR8, S. Peng and J. Sheehy, Agronomy, Plant Physiology and Agroecology and to report recent progress and challenges related to Division, International Rice Research Institute (IRRI), P.O. Box 3127, the yield potential of the NPT and hybrid rice. 1271 Makati City, Philippines; K.G. Cassman, Dep. of Agronomy, Univ. of Nebraska, Lincoln, NE 68583; S.S. Virmani and G.S. Khush, Plant Breeding, Genetics, and Biochemistry Division, IRRI. ContribuAbbreviations: DAT, days after transplanting; DS, dry season; IRRI, tion of IRRI. Received 28 Dec. 1998. *Corresponding author (s.peng International Rice Research Institute; LAI, leaf area index; NPT, @cgiar.org). new plant type; PhilRice, Philippine Rice Research Institute; WS, wet season. Published in Crop Sci. 39:1552–1559 (1999).