Canonical Correlation Analysis to Determine the Best Traits for Indirect Improvement of Wheat Grain Yield under Terminal Drought Stress

Selection based on the traits affecting grain yield is used for its indirect improvement. The objective of the present study was to determine the quantity and mechanism of effects of agro-morphological and physiological traits on grain yield of bread wheat under terminal drought conditions. Thirty six advanced lines of wheat were evaluated in a Randomized Complete Block Design with three replications during three sequential growing seasons 2010-2013 under rainfed conditions. Stepwise regression, path analysis, and canonical correlation analysis were conducted. All three components of yield had a significant positive effect on grain yield. The first canonical variable of predictive traits (U1) and yield components (V1) were studied. In general, selection for shorter vegetative stage and longer grain filling period is recommended to improve wheat grain yield per plant under rainfed conditions. Also, the desirable lines were those that had high amounts of biomass, average plant height and spike length, and low amount of canopy temperature. These types of lines are expected to produce higher numbers of grain per plant (not per spike) and 1000-grain weight, simultaneously, and thus higher grain yield per plant.