Utilizing Genotype-by-Weather Interactions for Selecting Wheat Cultivars
Crop cultivars should be widely adapted to varying climatic regions for optimum performance, but genotype by weather interactions often limit adaptation to subregions. Our objective was to quantify genotype by weather interaction effects between pairs of cultivars and thereby identify subregions where new cultivars could be expected to outyield established ones with prescribed confidence. Genotype by weather interaction effects were quantified by regressing yield differences (D) between a new and an established cultivar on weather variables using cultivar performance trial data for winter wheat (Triticum aestivum L.). A predicted value (D̂) was assumed to estimate the mean, E(D), of a hypothetical population of performance trials for a specified location-year. If the 95% confidence interval for E(D) covered only positive values, it is highly likely that E(D) was greater than zero and the new cultivar outyielded the established one in that location-year. For a specific location, inferences were extended to years not included in model development using the regression equation and associated confidence intervals for E(D). Results indicated that a new cultivar Karl would have outyielded a popular cultivar Newton in more than 50% of the years from 1950 through 1989 at locations in Kansas with mean annual precipitation exceeding 711 mm. Further, the mean yield of Karl would have exceeded that of another popular cultivar Arkan in over 50% of the years over all of Kansas. The methodology for quantifying genotype by weather interactions and extending results to years not included in model development provides important information for recommending new cultivars for specific subregions.
Copyright © . .