Early selection of induced genetic variability in yield components based on M₂-variances of an easily measurable trait

G. ROBBELEN and R. TRUJILLO-FIGUEROA^*

Section of Cytogenetics, Institute of Agronomy and Plant Breeding University of Gottingen, Federal Republic of Germany

Like any other breeding method, mutation breeding includes two basic procedures: (1) the production of variability and (2) the selection according to the final objective. Since FREISLEBEN and LEIN (1944), GUSTAFSSON (1947) and some other authors have shown certain mutants of crop plants to be of practical use, very efficient methods for the induction of mutations have been elaborated. Today there is no problem to produce any amount of genetic variability by ionizing radiation as well as by chemical mutagens. Certainly, the detection of recessive mutants in wheat is impeded by polyploidy. On the other hand, simple inactivation or deletion of genes, the dose of which exceeds the functional optimum may be beneficial in polyploid species. Moreover, RAO and SEARS (1964) pointed towards the interesting possibility, that the "unemployed" homologues of the duplicated genes can be altered in a way to create new functional activity, even within metabolic processes very different from the original one. In this case recessive mutations are manifested phenotypically, not only when homozygous for all four or six alleles, but already in the duplex condition. Such mutants, which arise from the transformation of accessory alleles, my be expected more likely to preserve the fitness of the initial genotype. Therefore, hexaploid wheat appears to be specially suited for mutation experiments which aim at a high percentage of mutants with positive and practically useful potentials.

Selection, the second procedure of mutation breeding, includes few problems as long as drastic mutants are desired. But there is at present no conclusive method of how to select for "small mutations", which probably are the main component of quantitative variation. The question is certainly almost the same for the selection in populations produced by crossing. After mutagenic treatment, the situation may even be less complicated, since the mutant is theoretically expected to differ from the original line only in single loci. However, the frequency of mutations leading to a positive shift in the particular characteristic may be low Thus the progeny size of the mutagen treated population is generally the factor most seriously limiting the success of any selection.