The traditional process of obtaining maize hybrids involves the generation of inbred lines through successive generations of selWng and subsequent testcrosses in order to identify the best combining ability by allelic complementation. A fast alternative to obtain inbred lines is to induce the formation of haploids followed by chromosome doubling. However, even with the aid of haploid-inducing genetic sources, this strategy has not been widely used in maize breeding programs, partly due to diYculties inherent to haploid generation and identiWcation. In order to evaluate the possibility of using dihaploids to generate homozygous maize tropical lines, we used the androgenetic haploid inducer line W23 as a female parent in crosses with the tropical single-cross hybrid BRS1010.Within the progeny of these crosses, 462 seeds were phenotypically selected as putative haploids by the purple-colored endosperm and colorless embryo conditioned by the R1-nj gene. Among these, only four individuals were conWrmed as being haploids using SSR markers, chromosome counting and Xow cytometry, showing that the phenotypic marker was not eYcient in detecting haploids in the tropical maize genotype used. All four haploids as well as some diploid plants presented reduced size, corroborating the diYculties for haploid identiWcation by phenotypic evaluation. Genetic diversity analysis revealed by SSR markers divided the haploids in two groups represented by Xint and dent maize inbred lines, which could be helpful in identifying complementary dihaploid lines. The present article demonstrates that a combination of haploid production and SSR Wngerprinting is a feasible strategy for maize hybrid development in tropical germplasm.

No abstract

No abstract