Go back to JSRRweb homepage:ホームページへ戻る
Genetic and Environmental Factors Affecting
Root Properties; Putative Traits (for Upland)
and Rainfed Lowland Rice Improvement
L.J. Wade* IRRI, P.O. Box 933, Manila, Philippines.
Introduction
Rainfed rice is grown on 69 million ha, with yields averaging 1.9 t/ha providing the staple food for more than one billion of the world's poorest peoples. Soil conditions range from aerobic in uplands to mostly anaerobic in flood-prone. In rainfed lowland, soils are generally anaerobic early, but change to aerobic as drought intensifies. This paper examines genotypic variation for key environmental factors, and the implications for rice improvement in rainfed ecosystems.
Environmental Factors and Key Traits
Upland rice (UR) encounters consistently aerobic soils due to the absence of ponded water, so its growth condition is comparable to wheat or maize. As drought is a major feature, deeper and more extensive root systems have been shown to be advantageous for extraction of soil water. Hardpans and acidic subsoils may provide additional barriers to development of an extensive root system. In contrast, irrigated rice (IR) is grown in ponded water, so soils are anaerobic and roots are shallow, and root/shoot ratios have been reduced in modern cultivars.
Rainfed lowland (RL) and flood-prone (FP) encounter features of both IR and UR, since shallow roots are commonly developed in anaerobic soils, but water is to be extracted from deeper soil layers as drought intensifies later in the season. Submergence may occur during the peak of the wet season, with depth and duration of flooding being greater in FP. As direct-seeding becomes more common, RL and FP are exposured to early-season drought as well. These contrasting soil conditions provide substantial challenges to root systems of rice in RL and FP. Root penetration to depth may be hindered by hardpans or low pH subsoils, and perhaps nutrient distribution. But unlike UR, oxygen supply, root signals and rate of stress onset may further complicate development of an effective root system for water extraction and nutrient uptake.
Approaches to Crop Improvement
Because of the importance of drought, crop improvement programs have focussed on root traits for both UR and RL (maximum rooting depth, root thickness, total root dry weight, root dry weight per tiller and deep root dry weight per tiller). RL has added hardpan penetration capacity and osmotic adjustment to this list of traits. Since all of these traits are difficult to screen directly, both UR and RL have sought to develop molecular markers for these complex traits.
Genotypic variation has been reported for all of the root traits listed. Lines differing in these traits were crossed to develop doubled haploid and recombinant inbred populations for identification of markers. Molecular markers have now been reported for root traits for UR, and have been identified for RL. Marked-assisted selection has commenced, to support the direct selections already practiced in drought-screening nurseries. Recently, root pull resistance has been added as a simple field-screening procedure for lines with good roots, due to a strong correlation with root dry weight. Of concern, however, is the relevance of screening procedures currently used in selection, or in phenotyping the breeding populations for marker development, relative to the conditions which actually define plant response in the taarget ecosystems. The concern is greater for RL than UR, because of the contrasting soil conditions encountered within the season in RL.
Genotypic Variation and the Implications for Selection
To address these concerns directly, the breeding populations are screened in anaerobic conditions in the greenhouse, to identify constitutive variation in root traits prior to the onset of stress. This provides a benchmark against which to evaluate variation in field screening nurseries conducted in the dry season. Field screening for root traits for drought avoidance is complicated by the low pH soils currently used. Responses in other locations are now being evaluated. Capacity of roots to penetrate wax-petrolatum layers in the greenhouse is compared with measurements of root pressures on compressed soils, and ability to penetrate a haardpan in the field. The critical assumption, that these traits confer an advantage under water stress, is being tested in greenhouse and field, by examining the consequences of their inclusion for water extraction and nutrient uptake. Finally, expression of these traits is examined relative to actual field conditions, where development of an effective root system under drought may be dependent upon successful root growth during an earlier anaerobic phase. Effects of low pH subsoils, hardpans, nutrient distribution, rate of stress onset, oxygen and root signals on root growth and function are examined at Ubon Ratchathani, Rajshahi, Raipur and IRRI in the next cycle of research. This research will confirm whether the current suite of traits would confer an adaptive advantage under drought in UR, FP, and especially RL. Related studies of genotype by environment interactions will assist crop improvement programs in deploying these trait combinations for different target subecosystems, such as RL late season drought.
References