Analysis of a Series of Yield Trials with Common Checks

Abstract

Sometimes a large number of new selected lines, although evaluated with replications, are not repeated over trials. However, there are a number of checks common across trials. The new selected lines from different trials are compared with the help of the common checks. Some limitations of an existing analysis has been described, and ways to overcome them are presented. An assessment of genotype X environment interaction has been made by check X trial interaction. A general case of incomplete block design at each trial and the connected nature of check vs trial classification considered to obtain the adjustment for comparing selected lines from different trials. A set of cereal trials has been considered for illustration.