Cultivar cross environment interactions for kernel yield in Virginia type peanuts (Arachis hypogaea L.) in Queensland

Abstract

Twelve Virginia type peanut cultivars were evaluated for kernel yield at 10 sites in 1979 and nine sites in 1980 in southern and northern Queensland. Two trials, harvested 9-14 days apart, were grown at each site in 1979, resulting in a total of 29 test environments. In 1979 the cultivar x site (c x s) interaction was significant, although the cultivar x harvest date within-site interaction was not. Over all 29 environments the cultivar x environment (c x e) interaction was significant. In orthogonal sets of five sites and 2 years, cultivar x year x site interactions were significant, whereas cultivar x year and c x s interactions generally were not. Pattern analysis revealed that cultivars of dissimilar genetic origin had different productivity responses across environments. An environmental classification based on c x e interactions indicated that there were no temporal or closely related regional environment groups with similar c x e interactions. Lower critical percentage differences between new and established cultivars in a pre-release testing program would be obtained by adding environments rather than replicates, or by accepting a greater risk of misjudging a new cultivar's superiority (higher probability of type I error).