Plans for Diallel Crosses and their Analyses

Abstract

The role of partial diallel crosses in plant and animal breeding experiments is well-known. The plans for partial diallel crosses hitherto have been generated by using partially balanced incomplete block deigns in two plot blocks and two associate classes. But these plans are not free from certain restrictions which the above two plot-block designs impose. In the present investigation new methods of evolving plans for partial diallels are suggested which prove to certain extent more flexible than the earlier ones. Making use of PBIB designs with any block size and any number of associate classes, the plans’ have been generated to suit analysis when reciprocal crosses are also performed. Methods of analysis for estimating general and specific combining abilities and maternal effects are presented along with standard errors for comparison. Expectation of mean squares due to general and specific combining abilities and errors have been calculated. Different cases according to whether parental inbred with/without the reciprocal F1’s are included or not have been discussed in chapter IV through chapter VII. The necessity for shorter blocks in the layout of the experiment becoming evident, methods for growing partial and complete diallel crosses in incomplete blocks are indicated. The analyses of such incomplete blocks of crosses are shown in chapter VIII and they include the case when reciprocal crosses are also present.