Agronomic Trait Variation in Oat Lines Derived from Tissue Culture

Tissue culture may generate agronomic variation that is potentially useful in plant improvement. This somaclonal variation, if extensive, could pose problems when tissue culture is a component of genetic transformation or when in vitro selection procedures are used. The objective of this study was to analyze agronomic trait variation in R 4 and R 5 generation oat ( Avena sativa L.) lines derived from R 1 progeny of tissue‐culture regenerated plants (R 0 ) that had been previously classified as having no karyotypic aberrations. Fifty‐six R 4 and R 5 ‘Lodi’ lines tracing to 11 R 0 plants, and 213 R 4 and 147 selected R 5 ‘Tippecanoe’ lines tracing to 45 R 0 plants were evaluated. Lines were planted in hill plots in radomized complete‐block designs with six replicates and evaluated for 2 yr at two locations. Traits analyzed were height, heading date, seed protein, flag leaf area, seed weight, seed number, grain yield, and bundle weight. Significant variation ( P < 0.05) was detedted for all traits both within families (lines derived from the same R 0 plant), and when lines and families were compared with the parental cultivars. Eighty‐two percent of the Lodi families and 93% of the Tippecanoe families had significant changes for at least one of the eight traits measured. Although agronomically less desirable changes were more frequent than desirable changes, liens with increases and lines with decreases were found for each trait. One cycle of successful bidirectional selection among culture‐derived lines demonstrated that the trait changes were heritable and that individual lines could be selected for use in plant improvement programs. The variation observed also demonstrates that materials produced by using directed genetic manipulation involving tissue culture should be tested extensively for possible quantitative trait variation.