Using height to adjust for interplot interference in spring wheat yield trials

Abstract

Interplot interference can occur in yield trials when genotypes vary for height. We investigated the possibility of adjusting yield for interplot interference by using height measurements in an analysis of covariance. Computer simulation showed that genetic correlation between height and yield would have no impact on analysis of covariance, as well as showing that analysis of covariance would reduce precision in the absence of interference due to height difference. Sixty-five data sets from yield trials conducted in Saskatchewan in 1992, 1993 and 1994 were analyzed to see if analysis of covariance would improve precision of yield estimates. Forty percent of the historical data sets showed evidence of significant interplot interference. In those trials in which interference occurred, yield decreased an average of 0.39% for every centimetre by which the average height of the flanking plots exceeded that of the test plot. Using final height to adjust for interplot interference is effective when genotypes vary for height and when the regression coefficient is significant. Key words: Interplot interference, yield adjustment, height difference, covariate