Abstract
A planting of 3,919 M(1) kernels from normal ears crossed by EMS-treated pollen produced 3,461 M(1) plants and 3,172 selfed ears. These plants yielded 2,477 (72%) total heritable changes; the selfed ears yielded 2,457 (78%) recessive mutants, including 855 (27%) recessive kernel mutants and 8 (0.23%) viable dominant mutants. The ratio of recessive to dominant mutants was 201:1. The average mutation frequency for four known loci was three per 3,172 genomes analyzed. The estimated total number of loci mutated was 535 and the estimated number of kernel mutant loci mutated was 285. Among the 855 kernel mutants, 432 had a nonviable embryo, and 59 germinated but had a lethal seedling. A sample of 194 of the latter two types was tested for heritability, lethality, chromosome arm location and endosperm-embryo interaction between mutant and nonmutant tissues in special hyper-hypoploid combinations produced by manipulation of B-A translocations. The selected 194 mutants were characterized and catalogued according to endosperm phenotype and investigated to determine their effects on the morphology and development of the associated embryo. The possibility of rescuing some of the lethal mutants by covering the mutant embryo with a normal endosperm was investigated. Ninety of these 194 mutants were located on 17 of the 18 chromosome arms tested. Nineteen of the located mutants were examined to determine the effect of having a normal embryo in the same kernel with a mutant endosperm, and vice versa, as compared to the expression observed in kernels with both embryo and endosperm in a mutant condition. In the first situation, for three of the 19 mutants, the mutant endosperm was less extreme (the embryo helped); for seven cases, the mutant endosperm was more extreme (the embryo hindered); and for nine cases, there was no change. In the reverse situation, for four cases the normal endosperm helped the mutant embryo; for 14 cases there was no change and one case was inconclusive.