پژوهشهای زراعی ایران (Sep 2022)
Detection Yield Related Traits of Wheat under Cyclic Drought Stress Condition Using Discriminant Analysis
Abstract
IntroductionWheat is one of the most important cereals in the human diet and widely used in many processed nutrition products. Water deficit stress is a main limiting factor of wheat growth and productivity in the world. Major objective of plant breeding is improving grain yield under drought stress condition. In the breeding programs, selection based on multi-traits is an important approach to improve grain yield. This research was conducted out to evaluate the effect of phenological and agronomic traits of 10 bread wheat near isogenic lines (in three genetic backgrounds) and cultivars on grain yield under cyclic drought stress condition and detection a function to use all effective secondary traits simultaneously.Materials and MethodsSix Near-Isogenic Lines (NILs) as well as their parents were evaluated at the research field of Shahid-Bahonar University of Kerman, during growing seasons of 2018-2019 and 2019-2020 under cyclic drought stress condition based on randomized complete block design (RCBD) with four replications. The field was irrigated every 28 days in autumn and winter and every other week in the spring. In the present research grain yield, phenological and agronomic traits were measured. Analysis of variance was performed using SAS v9.1. Broad sense heritability (h2bs) was calculated following the method of Fehr (1987) as follows:h2bs= σ2g / σ2g + σ2e (1) Phenotypic coefficient of variability (PCV) and genotypic coefficient of variability (GCV) were calculated as the following formula proposed by Singh and Chaudhary (1985):PCV= (σp/ µ) × 100 (2)GCV= (σg/ µ) × 100 (3)Where µ, σp and σg are mean, phenotypic standard deviation and genotypic standard deviation, respectively. Expected response (R) to selection in breeding programs was calculated following the methods of Falconer and Mackay (1996) as follows:R= ih2bσp (4)Where i is selection intensity, which is equal to1.694 if 10% of genotypes are selected (p = 10%) in breeding program.The studied genotypes were designated as group one and two based on grain yield under drought stress condition. Those traits that could significantly separate two groups based on t-test entered the discriminant analysis. These traits were standardized before discriminant analysis, as follows:Zij = (Xij - µ) / Si (5)where Zij is standard score for jth genotype in ith traits, Xij is raw data of for jth genotype in ith traits and Si is standard deviation of ith traits. Discriminant analysis was performed using MINITAB.Results and DiscussionAmong several secondary traits only awn length, flag leaf length and grain number per spike (grains/spike) could significantly distinguish high and low yield genotypes under water stress condition. These results showed the importance of the mentioned traits in the breeding programs for drought prone environments. Discriminant function of these traits was used as a comprehensive index for selection of high yield genotypes (Eq. (6)).DS= -1.32 + 2.07 FLL + 1.63 AL – 0.04 GNS (6) Where DS, FLL, AL, and GNS are discriminant score, flag leaf length, awn length and grains number per spike, respectively. This index could explain 72% of grain yield variation and had significant positive correlation with grain yield in water stress condition (r = 0.85**). Also it could well separate genotypes with the accurate classification rate of 90%. Discriminant function revealed that flag leaf and awn length were the most important effective traits on grain yield under drought stress condition, respectively. This index can be used as criteria for simultaneous selection of the mentioned traits in the future breeding programs.ConclusionAwn length, flag leaf length and grain number per spike that entered to the discriminant function had high correlation with grain yield, high heritability and easy evaluation. Therefore, selection based on these traits is a good approach to improve grain yield in drought prone environments. Discriminant function obtained in this study could be an appropriate technique to selecting high yield genotypes under drought stress condition.
Keywords
