Maximum Heterozygosity in Alfalfa: Results Using Haploid‐derived Autoteraploids 1

Abstract

The theory of maximum heterozygosity in autotetraploids (i.e., four different alleles at a locus) was tested in Medicago sativa L. sensu lato. Autotetraploid parents (HD4x) which could be only mono or di-allelic duplex were produced by chromosome doubling 2x haploids of cultivated alfalfa and their diploid derivatives from crosses with wild diploid relatives. HD4x parents were mated to produce single crosses (SC), which were in turn used to produce double crosses (DC). Gene frequencies should be the same in HD4x, SC, and DC populations, although their genotypic structures were expected to differ with the DC population having the highest frequency of tetra-allelic loci. Thus, any performance differences among the populations should reflect differences due to different genotypic structures. SC and DC populations were evaluated for forage yields in the greenhouse and field; self and eross-fertility and seed weight of HD4x, SC, and DC groups were tested in the greenhouse. Performance was consistently DC > SC for forage yield, although it was significant at P = 5% level in only one of four harvests. Performance for fertility and seed weight were DC > SC > HD4x, and differences were much larger and significant at P = 1% level. The consistent ranking of DC > SC > HD4x when compared with the theoretical genotypic structures of these populations provides evidence for both the existence of multiple allelic series and the importance of tri and tetra-allelic loci in conditioning maximttm performance. Methods of maximizing heterozygosity were discussed, whereby component lines or strains could be increased separately and then synthesized in two final steps, analogous to single and double crosses.