Testcross performance of Doubled Haploids versus S2, S4, and S6 selfing generation in a maize population for the subtropics in México

Abstract

Maize (Zea mays) hybrid breeding programs are increasingly using Double Haploid (DH) technology; however, in developing countries like in the subtropics, its use is still limited. DH is considered to provide 100% homozygous lines faster than the pedigree selfing method, and since production costs are presumably decreasing, it would be necessary to consider incorporating this technology into new and established breeding programs in these countries. This investigation starts with an S1 population from the same heterotic group the belongs to the CIMMYT Genebank, and it was divided into three parts. First, it describes the DH process, then the pedigree selfing method is evaluated for early generation testing, and finally, there was a comparison of both methodologies through a line by tester evaluation. The testcross performance of 15 DH lines and 14 lines from each of the S2, S4, and S6 selfing generation were tested in multilocation yield trials in the subtropics region of Mexico. It would take three years from the S1 population to identify the DH 100% homozygous lines with highest GCA, one year before the S6 lines (98.4% homozygous). For this data set, early generation testing was not found appropriate due to low correlation values. The DH lines reviled to have more GxE interactions than S2 testcrosses and the same heritability than the S6. There was no difference between the S2 evaluations; however, there were statistical differences found with the S4 and the S6 testcrosses. Thus, the DH technology could be implemented gradually in hybrid breeding programs, but the pedigree method should not be abandoned just yet, and testing in selfing generations should be done in S6.