Early generation selection for amylose content in rice grain: Heritability and response to selection

• Main Authors: Karladee,D., Gavilo,A., Jamjod,S.
• Format: Article
• Published: Chiang Mai University 2015
• Subjects: Multidisciplinary
• Online Access: http://www.scopus.com/inward/record.url?partnerID=HzOxMe3b&scp=84872230919&origin=inward; http://cmuir.cmu.ac.th/handle/6653943832/38039
• Institution: Chiang Mai University

Landraces of glutinous rice contain many desirable traits, but many consumers do not prefer its glutinous grain characteristic. Genetic variability in the glutinous grain type exists, suggesting that it should be feasible to select for non-glutinous grain type. The aim of the present study was to accelerate the selection of non-glutinous rice grain type by examining the amylose content of F3 and F4 lines. The grain amylose content of two hybrids of rice combinations was investigated: Kao Dok Mali 105 (non-glutinous) x Kum Doi Saket (glutinous) and RD15 (non-glutinous) x Kum Doi Saket. The amount of amylose was measured as an extraction from the unmilled ground grains, using 2N NaOH in 95% ethyl alcohol. Fifty F2 plants were selected randomly to build up the F3 families and, after harvesting, seeds of the 50 F3-derived lines were divided into two halves. One half was used for analyzing amylose while the remainder was used to produce 50 F4 families. In the F4 generation, 20 plants were sown from each family and the amylose means and variance values of the 50 F4 families were calculated as a within family variance. Heritability was calculated based on the linear regression coefficient value (b), which was obtained from the regression of the within family mean of F4 families on its relative derived F3 mean (hns 2 = b x 4/7). The response to selection (R) for amylose content was then estimated by using a mathematical model, R = iσh2, with 5% of selected proportion (v =5%). Variation in rice grain amylose content was detected in F3 and F4-derived lines in both combinations of parents, with a population mean intermediate between the two parents' values but closest to the value of the higher parent. This indicated the occurrence of the partial dominant action of gene segregation. Therefore, a selection for the character could be achieved in the F3 generation. In the F4 generation, family variance of the derived lines varied little with standard deviation (σ) = 2.24 and 2.35 in the two combinations. Regression of these variance values on its family mean showed a considerable number of families that regressed closely to the parents, thereby, indicating the homozygosity of genotypes in some F4-derived lines. There were 14 lines (28%) in Kao Dok Mali 105 x Kum Doi Saket combination and 2 lines (4%) in RD15 x Kum Doi Saket combination that showed homozygosity. A significant linear regression of F4 on F3 means was detected only in the derived lines of RD 15 x Kum Doi Saket combination with b = 0.3135. Thus, the low narrow sense heritability for grain amylose content in this combination (hns 2 = 0.179) indicated a large environmental influence was involved. Furthermore, if high performing lines were selected in F4, with v = 5% (i = 2.065), the achieved rate of response to selection for percent grain amylose content was +0.869 per generation, which is too low to change the population mean in advanced generations. Therefore, selection for high grain amylose content in rice should be performed at an early generation.