Pest linear unbiased prediction of the performance of crosses between untested maize inbreds

In some situations, maize (Zea mays L.) inbreds may not have testcross data available for best linear unbiased prediction (BLUP) of single-cross performance. The objective of this study was to evaluate the effectiveness of BLUP when the parents of a single cross have not been tested in hybrid combination within a given heterotic pattern. Yield, moisture, stalk lodging, and root lodging data were obtained for 4099 single crosses evaluated by Limagrain Genetics in multilocation trials in 1990 to 1994. For each of 16 heterotic patterns, the performance of an i x j single cross was predicted as follows: (i) with all available testcross data for both i and j and their relatives; (ii) disregarding all testcross data for i; (iii) disregarding all testcross data for j; and (iv) disregarding all testcross data for both i and j. Correlations between predicted and observed performance, obtained with a cross-validation procedure, were highest when testcross data for both parents of a single cross were utilized. Except for moisture, these correlations were severely reduced to <0.50 when both i and j were assumed untested. Prediction of performance of the cross between two untested inbreds seems unwarranted, but this situation is rare because new inbreds are usually crossed to extensively tested inbreds. The performance of the cross between an untested inbred and a tested inbred was predicted effectively when the number of tested single crosses in the heterotic pattern was large. In this situation, the highest correlations were approximate to 90% of the corresponding correlations obtained when data were available for both parents of the single cross.