Genotype ✕ Environment Interactions and Stability Analysis for Herbage Yield of Tall Fescue Synthetics1

Abstract

The development of forage grass cultivars which maintain a high level of performance over a wide range of environments is a goal of most breeding programs. Twenty‐five synthetics and two cultivars of tall fescue (Festuca arundinacea Schreb.) were evaluated for 3 years at two locations. Entries differed significantly for reproductive, vegetative, and total herbage dry matter yields. Significant entry ✕ environment (linear) interactions occurred for reproductive and vegetative regrowth, indicating that there were significantly different environmental responses among the 27 entries. Entry mean yields were linearly regressed on the average of all entries La each environment. Regression coefficients of the synthetics ranged from 0.71 to 1.24, 0.76 to 1.17, and 0.84 to 1.15 for reproductive, vegetative, and total herbage dry matter yield, respectively. Linear regression accounted for 95 to 99%, 78 to 99%, and 89 to 99% of the variation in yields of the 27 entnes for reproductive, vegetative, and total herbage dry matter yield, respectively. Mean square deviations from regression were found to be homogeneous, using the Bartlett test, for the three yield measurements. In the regression analysis of variance, the pooled deviations were not significant for all three yield measurements. The mean square deviations, coefficient of determination, and ecovalence stability indexes were highly and significantly correlated. We concluded that the mean herbage yield was a sufficient statistic for the selection of synthetics which have high and stable total yield. The regression coefficient was a useful statistic, in addition to mean herbage yield, if the goal was selection for desirable stable herbage regrowth yield.