| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Plant Breeding Dep., Inst. of Agric. and Animal Sci., Rampur, Chitwan, Nepal
b CIMMYT, South Asian Regional Office, Kathmandu, Nepal
* Corresponding author (rsharma{at}ecomail.com.np).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Abbreviations: AUDPC, area under disease progress curve • DHD, days to heading • HLB, Helminthosporium leaf blight • masl, meters above sea level • TKW, thousand-kernel weight • IS, selection index
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
In the past decade, breeders in South Asia have used several exotic sources showing good levels of HLB resistance. However, Dubin et al. (1998) reported that the best foliar blight-resistant wheats in South Asia were late and tall, two less desirable agronomic characters. The use of late-maturing HLB-resistant wheats in crossing programs with early maturing, HLB-susceptible commercial cultivars generated segregating populations where it was difficult to identify early maturing HLB-resistant lines with regular selection schemes (Dubin and Rajaram, 1996; Dubin and Duveiller, 2000).
There is little detailed information on the use of HLB severity rating per se to improve resistance in breeding programs targeting warmer wheat growing areas. There is even less information on the correlated response that such selection might produce on the improvement of other agronomic characters. A previous study reported that it was possible to select for better spot blotch resistance in early generations among progenies of crosses involving resistant and susceptible wheat genotypes (Sharma et al., 1997a). Selection for low disease levels also resulted in increased grain yield. However, the apparent correlation between HLB resistance and late growth stage increased the selection rate of resistant lines that were also late maturing, an undesirable character for wheat grown after rice in South Asia. The short grain-filling period in that region requires the use of short-cycle wheat cultivars to avoid temperature increases that may lead to shriveled grain. To overcome this difficulty, Sharma et al. (1997a) recommended selecting resistant genotypes among early flowering lines in the segregating generations.
Simultaneous selection for early maturity, HLB resistance, and heavier kernels in wheat has proven difficult in the warm, humid region of South Asia because of a complex negative relationship among the three traits. Duveiller and Gilchrist (1994) had suggested that tolerance or the slower rate of foliar blight development in wheat was often associated with late maturity. Dubin et al. (1998) was unable to find foliar blight resistance among early maturing wheat genotypes from South Asia. Duveiller et al. (1998) reported that early maturing wheats showed higher levels of blight compared with late-maturing genotypes. Sharma et al. (1997b) reported that spot blotch severity was negatively correlated with both DHD and kernel weight. Also, a negative correlation is often found between days to maturity and kernel weight in warm wheat growing regions. These early findings, together with the unsuccessful attempts by the wheat breeders in developing early maturing HLB-resistant cultivars, led to a general belief among wheat scientists that there is a genetic linkage between late maturity and resistance. However, two recent studies have shown that maturity and resistance to spot blotch are inherited independently in wheat (Sharma and Bhatta, 1999; Joshi et al., 2002), which suggests that there is no genetic linkage between disease resistance and maturity, and therefore selection for both these traits should be possible. Since simultaneous improvement of HLB resistance, maturity, and grain weight has not been so far successful with conventional selection technique, it is important that a new selection scheme be tested by building on the limited previous findings in the direction (Sharma et al., 1997b; Sharma and Bhatta, 1999; Joshi et al., 2002). Therefore, this study was initiated to simultaneously improve HLB resistance, TKW, and earliness by combining the three traits in an easy-to-use IS. A highly heritable trait, TKW is considered an important indicator of grain quality. Earliness and high TKW are two particularly important characters for the commercial success of a wheat cultivar in South Asia, where farmers prefer bold amber grains. Use of a IS involving these three traits simultaneously would be expected to improve all of them, which is difficult to accomplish through selection for an individual trait at a time. In index selection, the segregates inferior in multiple traits are excluded from the selection cycle, which could not be achieved either through tandem or truncation selection (Simmonds, 1981, p. 179). The use of a IS permits marked superiority in one trait to compensate for moderate inferiority in the other. Despite this advantage of a IS, it is not being used in wheat breeding programs because a complex equation has to be developed with appropriate relative economic or genetic weightings for all traits (Simmonds, 1981, p. 179). Therefore, a relatively simple IS has been used in this study that could be used in the wheat breeding programs with little difficulty. In most wheat breeding programs aiming at improving HLB resistance, maturity, and grain weight, selection in the early generations is frequently done quite late in the crop season when early maturing genotypes are already under natural senescence. In such cases, natural senescence of leaves is often difficult to separate from diseased leaf area. Consequently, late-maturing genotypes that are still green are selected as resistant. Discarding early maturing lines this way results in their exclusion from further selection. With early maturing wheat out of selection cycle, it is highly likely that genotypes with high kernel weights are also discarded because high temperatures during the grain-filling period often cause shriveled kernels on the late-maturing wheats. So, a selection method that could allow some compromise on all traits of interest could be effective in their simultaneous improvement. The proposed IS in this paper is an attempt in that direction.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
The crosses were made in the 1998-1999 wheat-growing season (November to March) and the F1 plants were grown in an off-season high-altitude field nursery in 1999 (May to October) at Marpha (Mustang), Nepal, to produce the F2 seeds. The F2 populations were grown in the field in Rampur (Chitwan) during the 1999-2000 wheat season to produce F2–derived F3 families. Each F2 population consisted of 
300 single plants, which were individually harvested and threshed to grow F3 plant rows. A number of F2 plants in each population produced <120 seeds and were not advanced. Attempts were made to keep equal number of lines in each F3 family. The F3 populations were grown in the field in Rampur (Chitwan) during the 2000-2001 wheat season. Each F3 family consisted of 234 lines grown as individual 1-m long rows, planted at 120 seeds by 0.25-m row spacing. After emergence, hand thinning was done to maintain 100 seedlings within each row to produce a solid crop stand. Fertilization was applied as recommended for the experimental site, with 100, 60, and 40 kg ha-1 of N, P2O5, and K2O, respectively. Three irrigations were applied, as required in Nepal lowland environments. Plots were kept free from weeds by hand weeding. The trial was seeded at optimum seeding time, the second half of November.
Days to heading was recorded when 50% of plants in a plot had spikes fully emerged from the boot. After anthesis, HLB scores were visually scored for each plot at weekly intervals with the double-digit scale (00–99) developed as a modification of Saari and Prescott's severity scale to assess wheat foliar diseases (Saari and Prescott, 1975; Eyal et al., 1987). The first digit (D1) indicates the disease progress in height, and the second digit (D2) refers to severity measured as the diseased leaf area. Four individual scores were recorded over a 3-wk period. For each score, percentage disease severity was estimated based on the following formula:
 |
The AUDPC was calculated with the percentage severity estimates corresponding to the four ratings as outlined by Das et al. (1992) and shown below:
 |
After maturity, plots were individually harvested and threshed. One thousand kernels were randomly counted from each plot's seed package and weighed to determine TKW. The 234 F3 lines in each family were ranked individually for AUDPC, DHD, and TKW. The entries with the same value for a trait were assigned the same average rank. The following IS was developed by combining AUDPC, DHD, and TKW and was calculated for each F3 progeny:
 |
In practice, the F3 line with the lowest AUDPC value within each family was ranked 1. Similarly, the F3 line with the lowest DHD value was ranked 1. In contrast, the F3 line with the highest TKW value was ranked 1. The IS of a given line is the sum of the three ranks. This IS was considered among many possible because of its simplicity with which it can be used in a breeding program. Equal weights were given to all three traits, again to keep the index simple for its use in applied wheat breeding programs.
After the IS values were calculated, the 20 lowest and 20 highest scoring lines in each population were selected. Each set of 40 genotypes was tested at two locations in the lowlands of Nepal (Manara and Rampur) during the 2001-2002 growing season, with a randomized complete block design with three replicates. For the past several years, HLB severity on wheat has been high at both Manara and Rampur. Environmental conditions are also different at the two sites, which are 260 km apart. The Rampur site (27°40' N and 84°19' E) is located at 228 m above sea level (masl) and 20 km away from the Himalayan foothills. The Manara site (26°43' N and 85°58' E) is situated at 84 masl and a further distance (50 km) from the foothills. Thus, when wheat is grown in the field during the winter or rabi season, temperatures are a little cooler at Rampur (average of growing season is 21°C) than Manara (23°C). Average humidity for the wheat crop-growing season at Rampur and Manara sites are 70% (range 68 to 73%) and 63% (range 62 to 65%), respectively. Average rainfall for the wheat-growing season at Rampur and Manara are 74 and 47 mm, respectively. Annual precipitation at Rampur is 2500 mm, compared with 2000 mm at Manara. The soil type at Rampur is a medium-textured loam compared with heavy clay soil at the Manara site. Rampur is a part of a large valley, whereas Manara is a part of the vast eastern Gangetic plains.
Individual plots (1.5 m2) were seeded at the standard seeding rate of 120 kg ha-1. Each plot consisted of two rows 0.25 m apart. Fertilizers were applied at the same rate in each plot (i.e., 100, 60, and 40 kg ha-1 of N, P2O5, and K2O, respectively). Three irrigations were applied according to soil conditions. The plots were kept free of weeds by hand weeding. The trial was planted on 25 Nov. and 5 Dec. 2001 at the Rampur and Manara sites, respectively. The optimum seeding time for the two sites is between 25 November and 10 December. Days to heading, four HLB scores at weekly intervals, and TKW were determined, and AUDPC calculated as explained for the F3 lines. At maturity, plant height from ground level to the tip of the spikes was measured in each plot. Plants were harvested by hand at ground level to collect all aboveground biomass. After threshing, grain weight was recorded and harvest index (%) was calculated as the ratio of grain yield to biomass yield.
Trials were conducted under natural disease infection in both seasons. Disease severity in the susceptible parent (Sonalika) exceeded 90% in both years. Weather conditions were optimal for wheat cultivation in 2001, but a little more humid than normal in 2002. Grain yield of commercially grown varieties was normal in both years.
Response to selection was estimated in the F4 generation for each population based on the mean difference of the 20 lines with low IS and 20 lines with high IS. Analysis of variance was conducted separately for each population and location to determine differences between the low and high IS groups for all parameters: AUDPC, DHD, TKW, plant height, biomass yield, grain yield, and harvest index. Also, a combined ANOVA across locations was conducted for each population to determine the effect of genotype x location interaction. Rank correlation coefficients were determined between the F3 and F4 lines to examine the change in relative differences within each IS group in the two generations.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
|
|
In this study, the use of a IS made it possible to detect desirable correlated responses in biomass yield, grain yield, and harvest index. This suggests that the proposed IS might prove a useful indirect selection criterion to improve grain yield in the warmer regions of South Asia where HLB is a special concern. The associated positive responses in biomass yield and harvest index indicate that agronomically superior lines were identified, an added advantage resulting from applying the IS. A previous study has reported that selection for foliar blight alone could bring improvement in grain yield (Sharma et al., 1997a); however, HLB severity was negatively correlated with DHD and TKW. Hence, the use of IS in this study appeared to be more efficient than using HLB alone, as done previously (Sharma et al., 1997a). Though this study did not aim at comparing selection for HLB alone with the IS in bringing out improvement in grain yield and other traits, it will be interesting to accomplish that in the same study.
The rank correlation coefficients between the two generations were mostly significant (Table 3) for all three traits considered in the IS. The values were generally higher for the low IS group and at Rampur. This means that the relative ranking of genotypes in the low IS category was more consistent than that of genotypes in the high IS group. The differences in correlation coefficients between the two locations could in part be attributed to the certain environmental conditions that differed between the two sites. These results suggest that irrespective of environmental effects, and since less genotype x environment interaction occurred among lines in the low IS group than in the high IS group, selection for low index value could be done very efficiently.
|
The results of this study suggest that selection for HLB-resistant wheat lines with high grain yield and high kernel weight is possible with a IS. Many early maturing HLB-resistant lines that could be directly used in wheat improvement programs were identified with the tested screening criterion. The method did not significantly penalize the crop for maturity and plant height, two important agronomic traits in wheat.
This is the first report on the use of a simple IS to improve HLB resistance and maturity in wheat. It is likely to prove particularly useful to wheat breeders who usually base genotype selection on means and ranks. Since the three characters that are part of the IS in this study are recorded regularly in most wheat breeding programs, the use of a IS does not increase workload. One could argue that using truncated selection could be as good as or even simpler than using a IS; however, breeders using truncated selection would be tempted to select first for early heading, then for disease resistance, and finally for kernel weight. Truncated selection would thus consider maturity more important than disease resistance and TKW. A breeder might consider selecting a line showing high levels of HLB resistance and high TKW with even intermediate maturity. This flexibility is inherent to the proposed IS and cannot be done with truncated selection. The advantage of using a IS is that it is flexible enough to allow balancing moderate defects in one trait with obvious gain in others. This approach of using multiple traits in the IS in this study converges with the concept of intuitive index selection described by Simmonds (1981)(p. 180).
|
ACKNOWLEDGMENTS |
|---|
Received for publication December 26, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Related articles in Crop Science:
This article has been cited by other articles:
|
|
 |
|
  W. Yan and J. Fregeau-Reid Breeding Line Selection Based on Multiple Traits Crop Sci., March 19, 2008; 48(2): 417 - 423. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Long, J. B. Holland, G. P. Munkvold, and J.-L. Jannink Responses to Selection for Partial Resistance to Crown Rust in Oat Crop Sci., April 25, 2006; 46(3): 1260 - 1265. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Plant Registrations | Soil Science Society of America Journal | ||||
| Journal of Natural Resources and Life Sciences Education |
Journal of Environmental Quality |
||||
