TESTS FOR CROSSOVER GENOTYPE-ENVIRONMENTAL INTERACTIONS

Abstract

Two statistical tests for crossover genotype-environmental interactions are illustrated. One test, suitable for investigating two-factor interactions in factorial experiments, was used to show that interactions between wheat cultivars and temperatures for median germination time did not involve significant changes in rank order. A second test was used to compare yields of three cultivars of spring wheat evaluated in nine environments and to demonstrate that the differences between HY320 and Fielder involved significant crossover interactions, whereas the differences between HY320 and Neepawa, and between Fielder and Neepawa, did not involve crossover interactions. Calculations required for the two tests are shown. The importance of the tests for investigating genotype-environmental interactions and related topics is discussed.Key words: Genotype-environmental interactions, qualitative interaction, statistical tests, Triticum aestivum L., wheat (spring)