Comparison of Stability Statistics as Criteria for Cultivar Development in Peanut1

Abstract

Abstract Means of yield and qualitative traits are assessed in multi-location tests in the final stages of breeding line evaluation. Due to large environmental variation and genotype x environment interactions, it is often desirable to compare stability of lines over a range of environments. The objective of this study was to use various stability parameters to try and determine the stability of experimental breeding lines. Using data from regional advanced peanut (Arachis hypogaea L.) breeding line yield trials conducted over 3 years and four locations in Virginia and North Carolina, the stability of peanut cultivars and breeding lines was compared. Stability variance was found to be highly correlated (0.91–1.00) with covariate adjusted stability variance. In many instances, the stability-variance parameters produced similar results to pairwise regressions and dissimilarity measures when compared with standard cultivars. However, the distance parameters and regressions provided more precise information on relative responses in varying environments of two advanced breeding lines being considered for release. This allowed for direct comparison to cultivars targetted for replacement. NC 18411 had equal mean yields and qualitative traits but better stability than breeding line NC 18423. Regression analysis indicated that NC 18423 performed best in good environments but worse than most other cultivars in poor environments. Means and stability of value per kilogram and value per hectare were highly correlated with percentage of sound mature kernels and yield (0.92–0.99), respectively. A comparison of means and stability parameters was effective in discerning superiority of peanut breeding lines for potential release and use by growers.