No abstract

aimed at high yields. The low yield of upland rice, however, is largely a consequence of its production being Upland rice (Oryza sativa L.), commonly considered to be low limited to low harvest index (HI, ratio of grain to total yielding, can be high yielding if the genotype is improved for harvest index (HI) and the crop is grown relatively free from nutrient and biomass) varieties (George et al., 2001) and to infertile drought stresses. We examined whether high and stable rice yields or drought-prone uplands. could be obtained in aerobic soil. In four experiments of 1-to 3-yr Indications are that upland rice can also be high yieldduration, lime, N, and P were inputs for wet-season upland rice ing if the genotype is improved for yield and the crop 'UPLRi-5' in a favorable rainfed Oxisol. In a 3-yr experiment conis not subject to nutrient and drought stresses. With sisting of two crops per year in an irrigated Ultisol, different lowland improved upland rice, yields approaching 7 Mg ha Ϫ1 can and upland varieties were grown in limed and fertilized aerobic soil. be obtained, even in highly acidic upland soils, given First-season rainfed UPLRi-5 yield varied from 1.5 to 7.4 Mg ha Ϫ1 , adequate amounts of lime, N, and P (George et al., with low yields in fields receiving low early-season rainfall. With 2001). In the highly acidic soil of the 240-million-ha irrigation, the lowland hybrid 'Magat' yielded 7.8 Mg ha Ϫ1 vs. 2.1 Mg Cerrado region in Brazil, upland rice is a commercial ha Ϫ1 for traditional upland rice 'Lubang Red'. Magat's high yield was associated with a HI of 0.43 in contrast to 0.31 of improved upland crop in rotational systems (Guimarã es and Yokoyama, rice variety 'Apo' and 0.17 of Lubang Red. Whether the crop was 1998), producing about 5 Mg ha Ϫ1 with fertilizers and rainfed or irrigated, yield loss was rapid following the first season: irrigation (Stone et al., 1997). In China's Huang-Huai-Grain yields decreased by up to 73% for rainfed UPLRi-5 in the Hai River plains, with no soil acidity constraints, yields second to third season. In the irrigated upland, yield loss in the second approaching 7 Mg ha Ϫ1 have been observed for imto fourth season was reflected in a 16 to 79% decline in 10-wk biomass.proved upland rice lines on irrigated dryland (Huaqi Here, the 13-wk biomass in the fifth crop was only half that of the Wang, personal communication, 2000). simultaneously grown first-season crop. We conclude that while prom-

1998). Traditional upland rice is adapted to soil acidity (Garrity et al., 1990; Wade et al., 1988); moreover, im-Upland rice (Oryza sativa L.) yields on infertile, acid soils in the provement in upland rice adaptation to soil acidity can Asian uplands average only 1 Mg ha Ϫ1 . Phosphorus deficiency is considered a major soil constraint to increased yield, but little quantitative be brought about by breeding (Zeigler et al., 1995; Kirk information is available. We analyzed P responses of traditional rice et al., 1998). Given that upland rice is inherently tolerant on farm in Laos, Thailand, and the Philippines and improved varieties to soil acidity and that improved acid soil adaptation can in researcher-managed trials in the Philippines. Treatments in onbe incorporated into superior germplasm, nutrient defarm trials were a control and 50 kg P ha Ϫ1 Ϯ 100 and 50 kg ha Ϫ1 N ficiency, particularly P, becomes a major soil constraint and K, respectively. Treatments in researcher-managed trials were P to production in upland rice areas (Garrity et al., 1990; rates on an unlimed and limed Ultisol. Mehlich-1 extractable P cor- Kirk et al., 1998). Hedley et al. (1994) found that upland related with on-farm grain yield (r ϭ 0.47). Phosphorus fertilization rice varieties quickly became P deficient in a high P-fixing, increased average grain yield (1 Mg ha Ϫ1 ) by 20%, total biomass acid upland soil in the Philippines. Even in shifting culti-(4 Mg ha Ϫ1 ) by 27%, and P uptake (4.1 kg ha Ϫ1 ) by 53%. Yield vation systems where ash deposits and nutrients from increased 37% with P ϩ N ϩ K, but only 16% of the 2.4 Mg ha Ϫ1 biomass increase was grain. Improved rice in researcher-managed trials mineralized organic matter are considered sufficient for responded to P, with a larger proportion of biomass partitioned to reasonable rice yield in the initial years after clearing grain [i.e., higher harvest index (HI)]. Grain yield of 'UPLRi-5' infallows (Roder et al., 1995a), P deficiency was found to creased from 3.2 to 4.6 Mg ha Ϫ1 in limed soil while that of 'IR55423-01' be a major nutrient limitation to upland rice yield (Van increased from 3.4 to 4.0 Mg ha Ϫ1 in unlimed soil. Phosphorus fertiliza- Reuler and Janssen, 1989). Phosphorus fertilizer applition always increased the frequency of higher yields averaged across cation has been reported to increase upland rice yield in trials, soils, varieties, and growing conditions. Yield gain from onthe Cerrado region of Brazil (Fageria et al., 1982), West farm P fertilization of traditional rice was small because of low HI, Africa (Van Reuler and Janssen, 1996;Sahrawat et al., unlike in improved varieties, which had a HI. We infer that increasing 1995), and Southeast Asia (Schmidt et al., 1990). upland rice yield in Asia would require genotypes with higher HI inThe poor infrastructure and lack of markets in the addition to P fertilization. fragile environments of Asia often do not offer much economic incentive for farmers to use purchased fertilizers. When inputs are purchased, they...

Upland rice (Oryza sativa L.), commonly considered to be low yielding, can be high yielding if the genotype is improved for harvest index (HI) and the crop is grown relatively free from nutrient and drought stresses. We examined whether high and stable rice yields could be obtained in aerobic soil. In four experiments of 1‐ to 3‐yr duration, lime, N, and P were inputs for wet‐season upland rice ‘UPLRi‐5’ in a favorable rainfed Oxisol. In a 3‐yr experiment consisting of two crops per year in an irrigated Ultisol, different lowland and upland varieties were grown in limed and fertilized aerobic soil. First‐season rainfed UPLRi‐5 yield varied from 1.5 to 7.4 Mg ha−1, with low yields in fields receiving low early‐season rainfall. With irrigation, the lowland hybrid ‘Magat’ yielded 7.8 Mg ha−1 vs. 2.1 Mg ha−1 for traditional upland rice ‘Lubang Red’. Magat's high yield was associated with a HI of 0.43 in contrast to 0.31 of improved upland rice variety ‘Apo’ and 0.17 of Lubang Red. Whether the crop was rainfed or irrigated, yield loss was rapid following the first season: Grain yields decreased by up to 73% for rainfed UPLRi‐5 in the second to third season. In the irrigated upland, yield loss in the second to fourth season was reflected in a 16 to 79% decline in 10‐wk biomass. Here, the 13‐wk biomass in the fifth crop was only half that of the simultaneously grown first‐season crop. We conclude that while promise exists for high‐yielding rice in aerobic soil, the rapid yield loss with successive rice cropping must first be overcome.