Grouping of rice (Oryza sativa L.) lines based on multivariate analysis under drought stress condition

For evaluation and grouping of 150 inbred lines of rice based on some morphological and physiological traits, an experiment was conducted using Augment design based on randomized complete block design (RCBD) with four replications in Agricultural College of Guilan University, in 2013. Correlation coefficients between traits showed a significant correlation between grain yield with almost all of traits. Factor analysis showed that four independent and main factors explained 94.23 percent of total variance in all lines. The first factor with 51.99 percent of variance was named as the yield components and fertility rate. The second factor with 25.41 percent of variance was nominated as fertility rate, the yield and the yield components. The third factor with 11.50 percent of variance was nominated as physiologic traits. The fourth factor with 5.33 percent of variance was nominated as plant morphologic. Cluster analysis by ward method with a 92.7 percent of original grouped cases correctly classified four distinct groups for studied lines. Finally, to select high-yielding lines and tolerate to drought stress in rice simultaneously selection can be carried out for traits such as number of total spikelet per panicle, number of filled grain per panicle; spikelet fertility, proline content, chlorophyll a, chlorophyll b, total chlorophyll and relative water content.