Reciprocal Diallel Crosses Impact Combining Ability, Variance Estimation, and Heterotic Group Classification

Questions such as the following often arise: Should reciprocal crosses be included in a diallel? and Would their inclusion in a diallel impact grain yield (GY), estimates of general (GCA) and specific combining ability (SCA) effects, and heterotic group classification in maize (Zea mays L.)? We evaluated a 12-parent maize diallel cross (Griffing's Method 3 and Method 4) in three environments to determine (i) if reciprocal crosses impact GY of crosses and GCA and SCA effects, (ii) if reciprocal crosses influence the GCA and SCA and residual variance estimates in a diallel analysis, and (iii) if reciprocal crosses impact maize heterotic group classification. The results showed that inclusion of reciprocal crosses in a diallel greatly impacted GY and estimates of GCA and SCA effects. Under the assumption of a random-effects model, the inclusion of reciprocal crosses caused the residual and GCA variances to decrease and the SCA variances to increase as the number of parental lines increased in a diallel cross. Because inclusion of reciprocal crosses impacted GY and SCA estimates, reciprocal crosses would have great impact on maize heterotic group classification. The maize heterotic groups might be classified differently with and without the inclusion of reciprocal crosses. Based on our dataset from southwest China, three heterotic groups seemed to be an ideal number for improving maize-breeding efficiency.