Genetic analysis of grain quality traits and marker assisted selection for fragrance trait in selected Malaysian rice (Oryza sativa L.) varieties

Rice has reached a yield plateau and is susceptible to a major disease outbreak, so genetic information of the breeding material is important for developing a high quality variety. As a result of breeding that has focused on yields, grain quality has declined. This study is an evaluation of the combining ability, heritability, and correlation of grain quality traits for selected Malaysian rice varieties and the actions of the genes involved in the inheritance of these traits. It also introduces marker-assisted selection for the fragrance trait in a selected F2 population. The mean squares values for general combining ability (GCA) were significant for grain length (GL), grain width (GW), milled grain length (MGL), milled grain width (MGW), length to width ratio (LW), milled rice recovery (MRR), head rice recovery (HRR), amylose (AMYL) and gel consistency (GC), which indicated the importance of additive gene effects to the inheritance of these traits. Rice varieties MR 84 and MR 267 were the best combiners for most of the traits. Q 85 was a good combiner for GC, GL and HRR. The results of the specific combining ability (SCA) effect promoted 7 out of 21 total combinations to the next generation cycle (F2) for further phenotypic selection: MR 84 Χ MRQ 74, MR 84 Χ MRQ 76, MR 263 Χ Q84, MR 263 Χ MRQ 74, MR 267 Χ MRQ 74 and MRQ 76 Χ Q 84. Seven combinations showed large and significant REC effects: MRQ 76 Χ MR 84, MRQ 76 Χ MR 263, and Q 84 Χ MR 263 MRQ 76 Χ MRQ 74, Q 84 Χ MRQ 74, Q 85 Χ MRQ 74 and Q 84 Χ MRQ 76. The genetic parameters of the grain quality traits showed higher additive variance compared to the dominance variance. The broad sense heritability for GL was moderate, while it was comparatively higher in GW,MGL, MGW, LW AND HRR. The narrow sense heritability of the grain quality traits was high for GW, MGL, MGW and LW and moderate for AMYL, GC and GL. Positive correlations were observed between 10 pairs of grain quality traits: AMYL and HRR, GL and MGL, GL and LW, GW and MGW, MGL and LW, MGL and MRR, MGL and HRR, LW and MRR, LW and HRR and MRR and HRR. Generation mean analysis revealed the importance of additive gene action in GL, MGL and MRR for a population of high and low amylose parents. However, the populations of intermediate and high amylose parents and intermediate and low amylose parents shared similar dominant gene actions for most of the physical grain quality traits. Large differences between the parents in the traits resulted in simple heritability. Heritability in the F2 ranged from low to high in the population of high and low amylose parents despite the non-heritability of some traits in the populations of intermediate and high amylose parents and intermediate and low amylose parents. Marker-assisted selection (MAS) was introduced for fragrance detection, and allelic specific amplification (ASA) successfully differentiated the fragrant plants from the non-fragrant. Chi-squared analysis revealed that phenotyping and genotyping of the fragrance trait was not significant for a segregation ratio of 3:1 and significant for the segregation ratio of 1:2:1. Out of a total of 35 rice microsatellite markers (RMs) used, 10 were identified as polymorphic between the evaluated parents. The markers were RM 25, RM 44,RM 72, RM 80, RM 152, RM 210, RM 281, RM 330, RM 342 and RM 342A. However, none of these markers fit the 1:2:1 segregation ratio. The best breeding approach for MAS is at the seedling stage before transplant as opposed to conventional methods in which it is normally carried out at the mature stage. Marker assisted selection (MAS) is introduced for fragrance detection. Allelic specific amplification (ASA) had successfully differentiated the fragrance plant from the non-fragrant. Chi-square analysis revealed that phenotyping and genotyping of fragrance trait was non-significant for segregation ratio of 3:1 and significant for the segregation ratio of 1:2:1 respectively. Out of a total of 35 rice microsatellite marker (RM) used, 10 out of 35 markers were identified as polymorphic between the evaluated parents. The markers were RM 25, RM 44,RM 72, RM 80, RM 152, RM 210, RM 281, RM 330, RM 342 and RM 342A. However, none of these markers fit to 1:2:1 segregation ratio. The best breeding approach for MAS is at the seedling stage before transplanting as compared to the conventional methods which was normally carried out at maturity stage