Partitioning of variation derived from tissue culture of winter wheat.

The presence of somaclonal variation is well documented in wheat, but information is needed which identifies point(s) during the tissue culture process at which variation is most likely induced. Field experiments were designed to partition the total somaclonal variation among three potential sources: single embryo-derived calli, regenerant (R0) plants of a common embryo-derived callus, and spike-derived lines of a common R0 plant. Three populations of winter wheat ('TAM 105', 'Vona', and 'KS75210') totaling 72 lines were evaluated in replicated drilled plots in the R2 and R3 generations. The principal source of variation was influenced by parent genotype. Considering all traits, somaclonal variation in the 'TAM 105' and 'Vona' populations was predominantly attributed to tillers from the same regenerant plant. This source, as well as the original R0 plant source, contributed to variation in the 'KS75210' population, depending on the trait measured. Embryos did not consistently provide a significant source of variation. The presence of somaclonal variation was not always associated with a downward shift in population mean compared to parental controls. Significant population increases were noted for spike density and biomass, and some lines showed significantly higher grain protein content without a yield reduction, but these responses were again genotype-specific.