Evaluation of grain yield stability of pure lines of rice in Guilan province

Introduction of new varieties of rice with a higher performance than local cultivars is one of the important goals of rice breeding programs. Grain yield stability analysis of the varieties and estimation of the genotype by environment interactions should be performed in a conventional breeding program. Nine pure lines of rice selected from the International Network for Genetic Evaluation of Rice (INGER) and cv. Khazar, as check variety, were compared in an experiment using randomized complete block design with three replications for three years (2002-2004) in Roudsar, Rasht and Talesh in Guilan province, Iran. Grain yield was adjusted by 14% moisture content. Analysis of variance (ANOVA) showed significant differences for grain yield among the lines. Results of combined ANOVA indicated that there were significant differences for grain yield of the genotypes averaged over years and locations. Due to significant genotype × location × year interaction, grain yield stability analysis was performed using Lin and Binns and AMMI methods. Lines No. 6 (CT9807-3-5 1-1-2P-M-1) and 9 (IR2101-4-159-1-3-3) were characterized as stable lines because of small coefficient of variation and interlocation mean squares as well as smallest distance from the offset of coordinates in AMMI vectors. The mean grain yield of these lines were 4970 and 4770 kg.ha-1, respectively. Line No. 5 (CNA 4364-4-4-4-1-3-1) with grain yield of 4790 kg.ha-1 showed specific adaptation to Talesh and can be recommended for this region.