RESOLUTION OF GENETIC VARIABILITY IN SELF-POLLINATED SPECIES WITH AN APPLICATION TO THE SOYBEAN

ONCEPTS for the partitioning of genetic variability for a mating population 'were developed by FISHER (1918) and have been extended under varying assumptions by numerous authors. Quantification of the nonallelic gene interactions was presented by COCKERHAM (1954) and HORNER, COMSTOCK and ROBINSON (1955). The comments in this paper are modifications of COCKERHAM (1954). Partitioning of genetic variability involves the identification of a population, such as an F, population or a randomly mating population, and the definition of a set of genetic parameters relative to this population. Inconsistencies develop when the partitioning concepts are extended to self-pollinated crops: (1) Because of problems in estimation, the partitioning of genetic variability for selfed progeny can be made only with the assumption that the gene frequencies are essentially .5; (2) experimental designs for measuring the variance partitions (if the required number of hybridizations was experimentally feasible) require plots of heterogeneous individuals; however, if competition between rows of soybean genotypes, for example, can create biases in the range of 10 bu/A (HANSON, BRIM and HINSON, 1961), then the interpretation of the yield of a heterogeneous plot certainly would be questioned; and (3) since the natural state for a selfpollinated species is the homozygous condition and selection in most self-pollinated species is normally practiced when the dominance variance is negligible, the definition of the partition of genetic variability in terms of say an F, distribution leads to a quantification for the genetic variability which has limited utility. The contribution of this paper is considered original only with respect to the redefinition of concepts as relating specifically to most self-pollinated crops. For a crop such as tobacco which is relatively easy to hybridize, the development would have limited applicability. The concepts will be illustrated with a set of quantitative genetic data from soybeans. An interpretation of the results also will be considered.