To date all reported attempts to hybridize maize and sorghum have failed (E. E. Gerrish, 1967; J. J. Mock and W. H. Loescher, 1973). At CIMMYT 12,844 sorghum pollinations have been made onto shortened silks of dehusked, isolated, detasseled maize plants. Because of seed breakdown immature embryos were excised and cultured on artificial nutrient medium. Seven hybrids have been retrieved, five from diploid (2n = 20) and two from an artificially induced stable tetraploid (2n = 40) maize, pollinated with diploid (2n = 20) sorghum. Hybrids were identified mainly on their cytology at mitosis. All hybrids possessed 20 maize chromosomes and from 10 to 2 of sorghum, whether the maize parent was diploid or tetraploid. To produce viable hybrids therefore it may be necessary that 20 maize chromosomes are present so that only fertilizations of unreduced gametes from diploid maize survive.
Although sorghum chromosomes are much smaller than those of maize in their own cytoplasm, this difference was obvious in only very few hybrid cells, and was minimal in the majority. It appears that maize and sorghum chromosomes condense differentially so that the period during which the size difference may be observed is extremely short. The presence of knobs on maize chromosomes and their absence on sorghum also helped to distinguish the two genomes.
Chromosome elimination occurred in all hybrids. Most cells observed from seedlings contained more chromosomes than from older plants, and at flowering very few cells contained more than 20.
F1 plants had a slow erratic growth and were morphologically more maize-like than sorghum. All produced a reduced tassel. In one no tassel was produced, five gave unbranched linear spikes with tassel seed at the base, and the seventh produced a tassel with two very short lateral branches. No hybrid gave viable pollen. All produced at least one ear but successful pollinations were made on only two. Thirty-three BC1 plants have been obtained with maize and will be described in a later paper.
Three major phenomena important to the production of hybrids have been identified:
1) Seed stimulation was very low, and on average occurred in less than one percent of the pollinations made. This seemed to depend on the genotype and condition of the plant as well as environmental conditions.
2) Visible large scale endosperm breakdown occurred from seven to seventeen days after pollination, depending on the environment and the condition of the plant. In only very few seeds the endosperm appeared to develop normally after this critical stage, but these seeds were always smaller than control comparisons at all stages.
3) Embryo breakdown may occur at any stage throughout seed development. No living hybrid embryos have been observed at seed maturity. Embryo development proceeded at different rates in individual seeds, and was not necessarily related to the stage expected in normal development. Type of development was also different, and hybrid embryos were often so abnormal in shape that they were not recognizable as embryos, until further development in culture. Most were not able to survive because of endosperm breakdown. However in seeds with apparently normal endosperm, embryo breakdown was independent of endosperm development.
In order to examine the potential value of sorghum germplasm in maize improvement, it is necessary to produce more hybrids of wide genetic diversity. Because of the low degree of seed stimulation large numbers of crosses must be made to give a sufficient number of hybrid embryos. Also, because of the abnormal development and complete breakdown of most hybrid embryos it is essential that embryo culture is used routinely for hybrid production. The results obtained at CIMMYT have encouraged us to expand our program, and we hope others will be encouraged similarly. The techniques used and the results obtained will be published in more detail shortly.
Return to the MNL 52 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page