| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
a International Rice Research Institute (IRRI), DAPO Box 7777, Metro Manila, Philippines
b Univ. of Alberta, Edmonton, AB, Canada T6G 2R3
* Corresponding author (g.atlin{at}cgiar.org).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: Add, Additive effect; Chr, Chromosome • DAS, days after seeding • LOD, logarithm of odds • MAS, marker-assisted selection • PCR, polymerase chain reaction • QTL, quantitative trait loci • SSR, simple sequence repeat.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Progress in breeding for drought resistance has been slow (Fukai and Cooper, 1995). It has been suggested that the development of drought-resistant varieties could be made more efficient by MAS to introgress alleles of QTL conferring improved drought resistance into the genome of widely used cultivars through backcrossing. When the traits that need to be improved are low in heritability, MAS may be more efficient than phenotypic selection (Asíns, 2002). However, for MAS to be worthwhile, the target QTL must have a large and consistent effect on yield (Salekdeh et al., 2002).
Several experiments have reported QTL for yield and yield components under different types of drought stress in rice. Babu et al. (2003) reported five QTL related to grain yield over two different trials conducted in southern India under upland drought stress in the population CT9993/IR62266. In one trial, severe drought stress was applied toward the end of the vegetative stage while in the other, stress was mild, but was applied at the flowering stage. The QTL with the largest effects explained approximately 20 and 28% of the genetic variation for yield in the vegetative and reproductive-stage stress trials, respectively, but none of the yield QTL observed were consistent across trials. Lafitte et al. (2004b), in a population derived from the cross Azucena/Bala, used a drip-irrigation system to precisely withhold water from the beginning of panicle emergence until 50% flowering but failed to identify QTL responsible for increased yield under stress. Lanceras et al. (2004), in the CT9993/IR62266 population, reported on QTL analyses conducted in Thailand in five different water treatments in a single transplanted trial under line-source irrigation. They identified four significant QTL for grain yield under water stress, one of which was consistent in three of the five water treatments and one which was detected twice. The largest-effect QTL identified in this trial explained about 30% of the genetic variance for yield under severe stress. None of those QTL corresponded to those previously identified by Babu et al. (2003) in the same population. Yue et al. (2005) reported a QTL detected in two consecutive years affecting yield, biomass, and harvest index reduction under drought stress in pot experiments. The QTL identified in this experiment (located on Chromosome 9 between markers RM316 and RM219) was consistent and stress-specific but of relatively small effect, explaining only 14 to 25% of the total phenotypic variation (Yue et al., 2005). Of all the QTL identified so far affecting rice yield under drought stress, none are considered to be sufficiently large or consistent to be useful in breeding (Lafitte et al., 2004a).
Despite the large number of QTL reported to affect drought tolerance in upland rice, there have been only a few attempts to introgress drought-tolerance QTL into susceptible genotypes (Courtois et al., 2003; Price, 2002). Large chromosomal regions bearing putative QTL associated with root length in a population derived from a cross between the deep-rooted upland cultivar Azucena and shallow-rooted lowland cultivar IR64 were introgressed into the IR64 background, but the majority of lines carrying the desired introgressions failed to have deeper roots than IR64 (Shen et al., 2001). The reasons for lack of effect of the introgressed segments on root length and yield may be that the target QTL were responsible for a relatively small proportion of the total phenotypic variation in the mapping experiment (5.6–17.7%) or because the introgressed region was large, and therefore desirable genes within it could be lost because of recombination during backcrossing. Azucena root-related QTL have also been introduced into the indica cultivar Kalinga III, but only one of the five target QTL had an effect on root length. Such introgressions have not been reported to consistently improve grain yield under stress (Steele et al., 2006).
Most reported drought-related QTL mapping experiments in rice have employed only five populations (Price et al., 2002). Many of the upland parents used in these populations (e.g., CT9993-5-10-1-M and Azucena) are not considered to be highly drought-tolerant in terms of grain yield under severe drought stress. Many traditional and improved cultivars from drought-prone areas have some tolerance to reproductive-stage drought stress, but they have rarely been used in mapping studies. A more extensive survey of tolerant rice germplasm, currently underway at IRRI, may lead to the identification of lines carrying major genes improving yield under stress. The objective of the present study was to detect QTL with large and repeatable effects on grain yield under reproductive-stage water stress in a cross between an eastern Indian cultivar, Vandana, which produces a relatively high grain yield under severe reproductive-stage stress, and Way Rarem, an Indonesian upland cultivar with high yield potential but poor drought tolerance.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Bulk F3 seeds from this cross were planted at IRRI during the dry season of 2004. F3 plants were harvested individually and their seeds used to constitute the current mapping population. A total of 436 random F3–derived lines were used in the mapping experiment. F3:4 seeds were used in the 2005 experiments and F3:5 seeds harvested from the nonstress treatment of 2005 were used to establish the 2006 trials.
Phenotyping of the Mapping Population
Field evaluation was conducted under upland conditions at IRRI, Los Baños, Philippines (14°11' N 121°15' E, 21 m above sea level) during the 2005 and 2006 dry seasons. The soil of the IRRI upland farm is a Maahas clay loam (isohyperthermic mixed Typic Tropudalf) (Zhao et al., 2006). In both years, the population was screened in one trial under nonstress conditions and one under stress during reproductive growth and grain filling. For each trial, the design was a two-replicate 
-lattice with 2-m single-row plots. Rows were spaced 0.3 m apart in 2005 and 0.25 m apart in 2006. Six checks with a broad range of drought tolerance (Vandana, Way Rarem, Apo, IR64, IR 55419-04, and IR 74371-54-1-1) were included in the trials and replicated 12 times each. Seeds were dry-direct-seeded in aerobic soil using a seeding density of 2 g per linear meter of row, resulting in a seed rate of approximately 305 seeds m–2 in 2005 and 365 seeds m–2 in 2006. Basal applications equivalent to 40 P and 40 K kg ha–1 were applied in the form of single super phosphate and potassium chloride, and 120 kg ha–1 of N in the form of ammonium sulfate were applied in three even splits around 21, 42, and 61 d after seeding. Planting dates were 22 Dec. 2004 and 21 Jan. 2006 for the nonstress trials and 11 Jan. 2005 and 7 Jan. 2006 for the stress trials.
The nonstress trial in 2005 was sprinkler-irrigated twice weekly, resulting in applications of approximately 36 mm water per irrigation, or 1224 mm over the entire growth cycle. Pan evaporation over this period was 637 mm. In 2006, basin irrigation was used once a week, resulting in the application of 92 mm water per irrigation and a total of 1554 mm for the entire season. Pan evaporation totaled 672 mm for the 2006 season. The nonstress trial of 2006 was affected by a moderate stem borer infestation and a typhoon immediately before the majority of the plots were harvested, which resulted in substantial lodging and seed shattering.
In both years, stress trials were sprinkler-irrigated twice weekly during establishment and early vegetative growth, but irrigation frequency was reduced at 56 and 40 d after sowing in 2005 and 2006, respectively. Plots were re-irrigated periodically when soil water tension fell below –50 kPa at a 30-cm soil depth. At this soil water potential, most lines were wilted and exhibited leaf drying. This type of cyclical stress is considered to be efficient in screening for drought resistance in populations consisting of genotypes with a broad range of growth duration (Lafitte et al., 2004b) and ensures that lines of all durations are stressed during reproductive development.
In the stress trial of 2005, the crop received a total of 129 mm of water from the beginning of stress until the end of flowering (from 57–96 DAS), supplying an average of 3.2 mm of water per day while the pan evaporation averaged 6.4 mm per day. There was a total of five irrigation and rainfall events: 46 mm at 66 DAS, 35 mm at 72 DAS, 6 mm at 75 DAS, 33 mm at 82 DAS, and 7 mm at 95 DAS. During that period, the mean daily minimum and maximum temperatures were 23.1 and 33.3°C, with the temperature reaching a maximum of 37.6°C.
The stress trial of 2006 received 167 mm of water between 57 and 104 DAS, resulting in an average of 3.6 mm per day while the average pan evaporation was 6.2 mm per day. There was a total of six irrigation and rainfall events during this period: 41 mm at 63 DAS, 7 mm at 65 DAS, 28 mm at 76 DAS, 15 mm at 82 DAS, 41 mm at 87 DAS, and 35 mm at 96 DAS. During that period, the mean daily minimum and maximum temperatures were 23.9 and 32.5°C, with the temperature reaching a maximum of 34.8°C.
Days to flowering was recorded as the number of days from sowing until 50% of the plants in a plot had flowering tillers. Final plant height was the length in centimeters from the soil surface to the tip of the panicle on the main tiller at maturity. Grain yield under stress and nonstress conditions is reported on an oven-dried basis (0% moisture). Biological yield was sampled by selecting a uniform section 50 cm in length in each plot and harvesting the plants in this section at ground level. Biomass samples were then oven-dried, weighed, and threshed. In 2006, the number of panicles within the biomass sample was recorded in both trials. Harvest index was estimated as the ratio of grain weight to whole plant weight. Flowering delay was calculated as the difference between the 2-yr line mean days to flowering under stress and days to flowering under nonstress. Drought-response index was calculated as the difference between predicted yield under stress based on the number of days to flowering and grain yield under nonstress conditions and the observed grain yield under stress divided by the standard error of the line mean (Bidinger et al., 1987).
Plant height at 4 wk after seeding was measured on four plants per plot in the 2005 nonstress trial, the 2006 stress and nonstress trials, and in another trial using the same population and design sown on 7 Feb. 2006. The last trial could not be reliably drought-stressed because of an early onset of the wet season in 2006, so yield data from this trial have not been used in the current analysis. Data from the four trials where plant height 4 wk after seeding was measured have been combined because the trials were all treated in a similar way at the beginning of the season, before the imposition of stress. Only the data combined over trials were analyzed for this trait.
DNA Extraction and Amplification
Eight fresh leaves from each line were collected in bulk 21 DAS. The leaves were then freeze-dried and ground in liquid nitrogen. The powdered leaves were then mixed with CTAB (cetyltrimethylammonium bromide) extraction buffer and incubated at 65°C for 30 to 60 min. Chloroform: isoamyl alcohol (24:1) was then added to the mixture. The mixture was then put on a shaker at room temperature for 20 min. Samples were then centrifuged and the upper phase was then transferred into a 2-mL Eppendorf tube. DNA was then precipitated with isopropanol precooled to a temperature of –20°C. The samples were then place in a –20°C freezer for at least 1 h and then centrifuged in order for DNA to pelletize at the bottom of the tubes. The DNA pellet was then rinsed with 70% (v/v) ethanol and recentrifuged. The ethanol was then drained off and then DNA pellet was left to dry. The DNA was then dissolved in 200 µL of TE (Tris-EDTA) buffer and treated with RNase. DNA was then quantified with a spectrophotometer and subsequently diluted to a final concentration of 10 ng µL–1 of double-distilled water.
Polymerase chain reaction (PCR) was performed using a volume of 20 µL per reaction. Amplifications were performed using 40 ng of DNA, 1x PCR buffer, 100 µM of dNTPs, 250 µM of oligonucleotide primers, and 1 unit of Taq polymerase. This mix was prepared on 96-well polycarbonate plates and the thermocycling used follows the method described by Panaud et al. (1996). After the PCR reaction was completed, 4 µL of 6x loading dye was added to each well. Four microliters of the resulting solution mix was then loaded into an 8% (w/v) polyacrylamide gel for size separation of the amplified DNA fragments using a minivertical electrophoresis system (CBS scientific, model MGV-202–33). DNA fragments were then stained with ethidium bromide and visualized with a UV transilluminator.
Genetic Analysis of the Mapping Population
A total of 474 SSR markers were screened for polymorphism between the parents. A total of 169 markers showed polymorphism on our gel system out of which we initially selected 121 to be used in the experiment. The markers were taken from previously published rice genetic and sequence maps (IRGSP, 2005; McCouch et al., 2002; Temnykh et al., 2001).
Quantitative trait loci with large effects were initially detected by selectively genotyping (Darvasi and Soller, 1992; Lander and Botstein, 1989) the highest-yielding and lowest-yielding 12% of lines from 2005 stress and nonstress trials. To compensate for the fact that grain yield under stress is negatively correlated with the number of days to flowering under nonstress conditions, selection of lines for inclusion in the tails was done on the basis of stratification into three categories according to the number of days to flowering in the 2005 trial: 60 to 73, 74 to 83, and 84 to 96 d to flowering. In each of those categories, the highest- and lowest-yielding 12% of the lines were selected for genotyping. A total of 39 and 38 lines from the highest- and lowest-yielding tails of the stress treatment and 35 and 38 lines from the highest- and lowest-yielding tails of the nonstress treatment were genotyped as well as 92 random lines. This resulted in a total of 242 lines being genotyped.
Initial QTL analysis of the random lines and the tails of the phenotypic distribution (242 lines) identified a major QTL on Chromosome 12. To increase precision of the position and effect estimate, five markers were added to Chromosome 12 and the whole population (436 lines) was genotyped at a total of nine marker loci on Chromosome 12 (RM3472 to RM3739).
This genetic information was subsequently used to create a linkage map using MapManager QTX (Manly et al., 2001). The marker orders used to create the linkage map corresponded to the published rice genome SSR marker orders (IRGSP, 2005). Map distances were calculated using the 92 random lines exclusively, except in the 10-marker interval (RM3472-RM3739) on Chromosome 12 where the data from all 436 lines was used.
Statistical Analysis
The model for the combined analysis over years and within irrigation treatments was:
 | [1] |
For estimating line means within years, replicates, and blocks within replicates were considered random factors, with lines considered fixed. For estimating line means over years, year effects were also considered random. The results for the yield trial were analyzed by the REML algorithm of PROC MIXED of SAS V.9.1 (SAS Institute, Inc., 2002–2003). Degrees of freedom were computed using the Kenward-Rogers version of the Satterthwaite correction for estimating degrees of freedom (Piepho et al., 2003; SAS Institute, Inc., 2004). The REML algorithm of PROC VARCOMP was used to obtain the variance components required to calculate genetic correlations (rG) among traits, broad-sense heritability (H), and proportion of the genetic variance explained by QTL. For variance component estimation, the model described in [1] was used, with all effects considered random. Differences are only discussed when p < 0.05.
The following formula was used to compute H across years:
 | [2] |
G2 is the genotypic variance, GY2 is the genotype x year variance, E2 is the plot residual variance, and r and y are the number of replicates and years, respectively. For traits measured in 1 yr only, heritability was estimated as:
 | [3] |
G'2 is the genotypic variance from the single-year analysis. This estimator of broad-sense H is biased upward by the genotype x year interaction variance but was used to approximately compare the relative repeatability of panicle number per m–2 across stress levels as this trait was measured in only a single season (Table 1a and b).
|
 | [4] |
For genetic correlations between two traits measured in the same environment:
 | [5] |
G12 and G22 are the genetic correlation coefficient between traits 1 and 2 within the same trial, genetic covariance of traits 1 and 2, and the genotypic variances of traits 1 and 2, respectively.
QTL Analysis
Composite interval mapping (CIM) was performed by QTL Cartographer (Wang et al., 2005). The minimal LOD value required to declare a QTL was obtained empirically from 1000 permutation tests (Churchill and Doerge, 1994). The LOD thresholds obtained correspond to an experiment-wise type I error rate of 0.05. The thresholds values obtained for different traits varied from 3.5 to 37.7 with an average of 8.0. The linkage map covered 1591 cM using the Kosambi mapping function, resulting in an average of 1 marker every 12.7 cM (Fig. 1
). CIM was conducted separately for means within years and means over the 2 yr.
|
 | [6] |
Variance components were estimated using the REML option of SAS PROC VARCOMP.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
There was a higher average biomass accumulation under stress in the 2006 drought trial than in 2005, but a lower harvest index in that same year resulted in lower grain yield on average. The earlier onset of drought stress in 2006 might be responsible for the lower harvest index. In the nonstress trials, a higher biomass accumulation combined with a lower harvest index also contributed to slightly lower grain yield in 2006. The lower harvest index in the 2006 nonstress trial can partially be explained by the stem borer (unidentified species) infestation and the typhoon that hit this trial when a lot of material was ready to be harvested.
In all trials, there was a large growth duration difference between Vandana (64 d to flower) and Way Rarem (91 d to flower). The two parents differed for grain yield and harvest index under stress but not under nonstress conditions (Table 1a and b). Those differences do not necessarily imply that Vandana is much more drought resistant than Way Rarem because of the large difference in growth duration. In this experiment, the drought-response index of the parents did not differ significantly, indicating that the difference in grain yield under stress between Vandana and Way Rarem may have been due mainly to their very different phenologies (Table 1a). The parents also differed significantly in the number of panicles produced per m2, with Vandana producing many more.
Under nonstress conditions, population means for grain yield, biomass yield, harvest index, final plant height, and plant height at 4 wk were all higher than in either parent. However, the mean number of days to 50% flowering for the population was intermediate to the parental means (Table 1b).
Under severe drought stress conditions, large variation in performance was observed among the lines. Averaged over the 2 yr, a total of 44 lines yielded less than 100 kg ha–1, while 10 yielded more than 1000 kg ha–1. Population mean biomass yield and plant height under stress were higher than for either parent in both years. Population mean flowering delay (1.9 d) was also higher than for either parent, but the average grain yield, harvest index, days to flowering, and panicle number of the population was intermediate to that of the parents (Table 1a). Despite the relatively early onset of stress in 2006, panicle number was not greatly affected by stress, with a reduction of only 5% compared with the nonstress trials. Many lines flowered earlier under stress than under nonstress conditions, some by up to 14 d. Acceleration of flowering under drought stress has been observed previously in some short-duration rice cultivars (Lafitte et al., 2006).
Heritability estimates for the number of days to flowering and plant height were high in all trials. The estimate obtained from the combined analysis over years of grain yield under stress (0.70) was much higher than that obtained under nonstress conditions (0.23) (Table 1a and b). The genotype x year interaction variance for grain yield under stress was low (Table 2), and H estimates were relatively high within individual years (Table 1). Biomass yield and harvest index also exhibited a higher level of heritability under drought stress than under nonstress conditions. This demonstrates that the drought screening protocol, which ensured that water-stress was imposed for the entire period of reproductive growth, was highly repeatable. Several factors such as germination problems in both years due to water-logging immediately after sowing, stem borer infestations, and the 2006 preharvest typhoon contributed to the low H of the nonstress trials. Although rice commonly is grown in flooded fields, the crop does not always germinate well when direct seeded under anaerobic conditions (Dennis et al., 2000), and stem borers are a major insect pest of rice that cause formation of "dead hearts" and sterile panicles. Effective chemical control of stem borers is difficult since the insect lives within the rice stems (Ho et al., 2006). Stress trials were planted on different dates compared with nonstress trials and escaped the worst of these problems.
|
|
|
|
A total of five genomic regions had significant effects on grain yield under nonstress conditions (qtl2.1, qtl3.1, qtl7.1, qtl8.1, and qtl10.1). The only region that had significant effects in both years was qtl2.1, located near RM3212 on Chromosome 2. This region was also associated with panicle number under stress. The yield increasing allele at this locus was contributed by Vandana and explains about 11% of the phenotypic variation in the combined analysis over years (Table 4b).
Days to Flower
A large-effect QTL for days to flower under both stress and nonstress conditions was identified on Chromosome 3 near marker RM523 (qtl3.1) (Table 4a and b). This QTL had a LOD of 34.4 (stress) and 52.3 (nonstress) for means over 2 yr, resulting in an additive effect of 11.2 and 8.6 d, respectively, for the stress and nonstress trials, with the duration-increasing allele contributed by Way Rarem, the long-duration parent. This region had a significant effect on days to flower in all four trials as well as on biomass yield under stress and nonstress conditions and plant height under nonstress conditions. The effect measured at this locus is probably caused by the gene Hd9, which has previously been fine-mapped within a few centimorgans of RM523 (Lin et al., 2002).
Another QTL (qtl5.1) affecting days to flower over more than one trial was significant in the nonstress trial of 2006 as well as in the stress trial of 2005 and for average of the two stress trials. The duration-increasing allele in this case came from Vandana. Two other QTL were detected for days to flower under stress in 2005 and over the 2-yr average (qtl1.1 and qtl7.2). In both cases, the duration-increasing allele was contributed by Way Rarem.
The effect of qtl12.1 on flowering occurred only under stress conditions. The effect of the Way Rarem allele was to reduce days to flowering under stress in both years by approximately 3 d, or, in other words, to reduce flowering delay due to stress.
Biomass Yield
As noted above, qtl12.1 had a large effect on biomass yield under stress. Qtl3.1 and qtl6.1 also significantly affected this trait in 2005, but not in 2006. The effect of qtl3.1 on biomass probably results from the fact that the Way Rarem allele at this locus increases crop duration. In the case of qtl6.1, the favorable allele was contributed by Vandana and the increase in biomass is probably a consequence of an increased panicle number, as a QTL for panicle number under stress is located in the same marker interval.
Three QTL for biomass yield under nonstress conditions were identified on the basis of the 2-yr mean. One of those loci corresponds to Hd9 (qtl3.1) and is contributed by Way Rarem. A larger-effect QTL is also contributed by Way Rarem (qtl9.1), while the one with the smallest effect (qtl4.1) was contributed by Vandana.
Plant Height
A QTL for plant height under both environments contributed by Vandana is located near marker RM315 on Chromosome 1 (qtl1.1). This QTL was identified in both nonstress trials as well as the 2006 stress trial and also affected plant height 4 wk after seeding. Given that RM315 has previously been mapped to the same location as RG220 (Robin et al., 2003) and that RG220 is known to be only 0.3 cM away from Sd1 (Lin et al., 2002), we assume that the differences in plant height observed here result from polymorphism between the two parents at the Sd1 locus.
Under nonstress conditions, a further four QTL affecting plant height were identified (qtl3.1, qtl4.2, qtl5.2, and qtl6.2). Qtl3.1 was identified in 2005 only and qtl4.2 in 2006 only. Qtl5.2 and qtl6.2 were significant based on the combined analysis and in both cases; the height-enhancing allele was contributed by Vandana. Qtl5.2 also showed a significant effect on plant height 4 wk after seeding, but the three other QTL did not. Qtl12.1 is the only locus that had an effect on plant height under stress conditions only.
Analysis of plant height 4 wk after seeding, on the average of four trials, revealed five significant QTL (qtl1.1, qtl5.2, qtl8.2, qtl9.1, and qtl12.1). Three of those QTL had been identified to have a significant effect on other traits (qtl1.1, qtl5.2, and qtl12.1), while the last two were detected exclusively for plant height 4 wk after seeding.
Harvest Index
Only qtl12.1 had a significant effect on harvest index under stress. Two loci (qtl2.2 and qtl8.2) had significant effects on harvest index under nonstress conditions in 2005, but none did in 2006 and no loci had significant effects in the combined analysis over the 2 yr.
Characterizing the Effects of qtl12.1 through Single-Marker Analysis
To better understand the effects of the major QTL for grain yield under stress identified at qtl12.1 in the CIM analysis, we performed a single-marker analysis for all traits under both stress and nonstress conditions for RM511, the marker within the qtl12.1 region with the largest effect on grain yield under stress. Under severe stress, RM511 had a significant effect on grain yield, biomass yield, plant height, days to flower, drought-response index, and flowering delay (Table 2). RM511 had no significant effect on any trait measured in the nonstress environments, except for a very small effect on plant height at 4 wk after seeding, with this locus explaining only 8% of the total genetic variance.
Variance components presented in Table 2 indicate that qtl12.1 explained a large proportion of the genetic variance for the following stress traits: grain yield (51%), biomass yield (40%), harvest index (35%), and plant height (36%). It contributed to the genetic variance of days to flower under stress (10%) and plant height 4 wk after seeding (8%) to a lesser extent. Genotype x year interaction affected all traits except harvest index under nonstress conditions (Table 2). However, in the traits for which qtl12.1 had a significant effect, the proportion of this interaction explained by qtl12.1 was very low, with most of the genotype x year variance resulting from variation among lines within marker genotype classes. Qtl12.1 accounted for only 12% of the total genotype x year interaction for grain yield under stress and 7% for biomass yield under stress, indicating that this QTL is fairly stable across years and would be suitable for MAS.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The effect size of this locus in the current population seems to be large enough to support fine-mapping, and the very low genotype x year interaction gives this locus a potential for use in MAS to improve the drought tolerance of the eastern Indian upland parent, Vandana. Because this parent is already considered to have good drought tolerance, and is currently grown in the severely drought-prone upland environment of the eastern Indian plateau, an improvement in its drought tolerance could be of significant benefit to farmers in the region who depend on upland rice for food security. Fine-mapping of the locus is currently under way.
|
ACKNOWLEDGMENTS |
|---|
Received for publication July 27, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. K. Jena and D. J. Mackill Molecular Markers and Their Use in Marker-Assisted Selection in Rice Crop Sci., July 1, 2008; 48(4): 1266 - 1276. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. C. Collins, F. Tardieu, and R. Tuberosa Quantitative Trait Loci and Crop Performance under Abiotic Stress: Where Do We Stand? Plant Physiology, June 1, 2008; 147(2): 469 - 486. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
