ISOLATING BETTER FOUNDATION INBREDS FOR USE IN CORN HYBRIDS

Abstract

It was shown that the yields of S3 and S4 inbreds are better criteria of their value in hybrids than generally has been recognized. A re-analysis of JENKINS’ (1935) data indicates that changes in the rank of 12 families for combining value from S2-S3 to S648 were associated with successful selections among S4 progenies. Furthermore, that discarding the poorer families on the basis of top-cross tests in Sz and S3 would have eliminated three families whose average combining value in S648 practically equalled that of the best families which would have been retained. A working hypothesis is advanced as to the operation of genes determining yield in corn. The essentials of this hypothesis are that genes determining quantitative characters lack dominance; that genes in the dominant-recessive category are concerned with vigor; that individual genes differ in effect; and that the frequency of a recessive in a population will be in inverse relation to its deleteriousness. The expectations under this hypothesis are explored, with the following conclusions Test crosses are a good criterion of combining value at any stage in the program, as of that time. They are not good indicators of prospective combining value, however, until tixation has reasonably been achieved. Selfed performance on a progeny basis is not a good criterion of ultimate combining value until selection has eliminated recessives of larger individual effects and lower frequencies. With effective selection against recessives of major effect and progress toward fixation, progeny performance of selfs and of crosses will tend to tell more nearly the same story. Selfed performance then will be satisfactory for selecting within families, while crossed performance will be more suitable for selecting among families A program based on these principles is outlined and designated as Cumulative Selection. This is a single, integrated system, comprising: (I) an initial period of visual selection for individual plants in desirable progenies; ( 2 ) testing for combining value; (3) crossing among the high combining families; and (4) a second cycle of selection based on selfed progeny performance.