RFLP variation and genealogical distance, multivariate distance, heterosis, and genetic variance in oats

Patterns of restriction fragment length polymorphisms (RFLPs) have been proposed as estimators of genetic diversity among breeding lines and as predictors of heterosis and genetic variance. We evaluated these proposals by using a set of nine elite oat lines crossed in a diallel mating design without reciprocals. RFLP analysis was conducted using HindIII-digested DNA and a total of 107 probes from three different sources: 14 heterologous wheat cDNA clones, 17 oat genomic clones, and 76 oat cDNA clones. Of the 77 probes that produced high-quality autoradiographs, 26 detected polymorphisms among this set of lines, with an average of 2.6 variants per probe. RFLP-based genetic distance (FD) was calculated from these data by using Nei and Li's measure of genetic similarity, and was compared with two other measures of genetic divergence. Genealogical distance (GD (*)) was obtained from the coefficients of parentage based on known parental pedigrees, and multivariate distance (DI) was calculated by using the first five principal components of the parental correlation matrix for 12 agronomic traits. FD was significantly correlated with GD (*) (r=0.63, P<0.01), but not with DI (r=-0.05). Cluster analysis based on these three distance estimates did not produce equivalent groupings, but the FD and GD (*) clusters were more similar to each other than to the DI clusters. These results indicate that: (1) sufficient variation exists for further application of RFLP technologyto oats, (2) RFLPs could provide accurate estimates of genetic divergence among elite oat lines, and (3) it is unlikely that dispersed markers can predict heterosis or population genetic variance in oats. Further investigations will require more parental lines, a larger set of markers, and more information on the linkage relationships between RFLP markers and loci controlling the trait of interest.