Results of one cycle of recurrent selection in rye

Abstract

A sufficiently variable rye population was subjected to one cycle of recurrent selection for combining ability in order to ascertain whether this complex property can be repatterned. Some hundreds of clones derived from this population were toperossed to an other unrelated variety used as tester. Through visual selection, more than half of these clones were rejected. The topcross progenies of the remaining clones were subjected to a comparative test whereby, as was to be expected, a considerable degree of variability was demonstrated. From the best 10% and the poorest 10% that is from those which surpassed the general average by at least 1.5 standard deviation units or remained below that average by at least 1.5 σ, narrowed populations were built up. This was done with the aid of reserve seed which in the year in which the top cross was made had been obtained from these clones, either through compulsory selfpollination or by allowing the clones to fertilize each other in isolation. The two subpopulations obtained through selection in opposite directions further indicated as “high” and “low” were crossed “en masse” with the same tester variety as the first time. In addition a large number of randomly taken individuals from the “high” synthetic were cloned and toperossed to this common tester. On testing the top crosses carried out “en masse”, the “high” synthetics clearly yielded more than the “low” ones. In one trial it was even found that the “high” synthetic surpassed our standard variety Petkus by approximately 20%. It is possible, therefore, to segregate a population into two fractions of different combining ability. The optimistie pieture to be derived from the foregoing trials unfortunately is shaded by the results of the individual test crosses pointing to a shift of the frequency distribution in a negative sense in regard to the original population. The influence of the year may be mentioned as a possible cause of this contrast and also the fact that in the original population a close visual selection had been practised while those obtained from the synthetic resulting after one cycle were derived from individuals which had been taken entirely at random.