Stability of Harvest Index and Grain Yield in Winter Wheat1

Abstract

Cereal breeders seeking yield improvement have considered that selection for harvest index (HI) might be more effective than selection for yield per se. However, no information is available on the stability of HI across diverse environments. This study was conducted to determine the stability of HI and grain yield and to study the correlation between the 2 in 10 adapted winter wheat (Triticum aestivum L.) genotypes. The genotypes were grown in replicated tests at six locations in Oklahoma in each of 2 yr. Two stability parameters were obtained as linear regression coefficient (b) of an entry mean on the average of all entries in each environment and deviations from regression (s2d), and stability was tested with the hypotheses b = 1.0 and s2d = 0. Genotypes differed significantly for HI and grain yield, and significant genotype ✕ environment interactions occurred for both traits. Regression coefficients ranged from 0.65 to 1.46 and 0.75 to 1.17 for HI and grain yield, respectively. Deviations from regression were significant for most genotypes for both traits. On the basis of the estimates of stability parameters, all genotypes were unstable for both traits. Correlation coefficients between HI and grain yield were inconsistent in good and poor environments. This study suggested that both HI and grain yield are significantly affected by environmental changes. Consequently, selection for yield per se might be as effective as selection for HI for improving wheat grain yields in the diverse environments in the Southern Great Plains.