Choice of the effective traits on grain yield as selection indices in progressive F6 populations in rice (Oryza sativa L.)

To investigate the relationships among effective traits on grain yield and determination of and the relationships among them, 137 rice recombinant lines from F6 population along with four control cultivars (Tarom Deilamani, Hashemi, Saleh and Ali Kazmi) were evaluated for nine important agronomic traits and two crop seasons (2015-2016) in a augment design with four blocks. The results of analysis of variance indicated a significant difference among the lines for all evaluated traits. Also, the differences among control varieties were significant for all traits except for leaf length and width. Simple correlation coefficients were calculated for each year separately and showed that 100 seed weight, number of fertile tillers, number of filled grains, panicle length, leaf length and leaf width had a positive correlation with grain yield. Multiple regression analysis by stepwise method was calculated for each year separately and the results showed that for the first year, 100-grain weight and number of filled grain per panicle traits were entered into model, respectively, that explained 62 percent of grain yield variations. For the second year, 100-grain weight, number of filled grain per panicle and number of panicles traits were entered into model, respectively, that explained 73 percent of grain yield variations. Also based on path analysis results, after 100-grain weight (0.926), the number of tiller (0.739) and number of filled grain per panicle (0.682) traits have the most and positive direct effect with grain yield and thus, these traits could be considered as selection criteria for grain yield improvement in rice. The lines include 9, 132, 133, 24, 52, 83, 5, 40, 78 and 19 were selected as elite lines based on the above criteria. Due to the fact that these traits had a great effect on yield, the selected genotypes can effectively be used as high yielding lines for future studies based on these indices.