COMPARISON OF HONEYCOMB AND INDEPENDENT CULLING OF SPACED PLANTS FOR YIELD POTENTIAL IN FABA BEAN

The honeycomb design of Fasoulas (1973) and the independent culling methods of McVetty and Evans (1980) were used to select among spaced plants of two generations (F2 and F4) of two crosses of faba bean (Vicia faba L.) for yield potential. Single plants were planted in a hexagon pattern with 60 cm interplant spacing for the honeycomb selection method and in a grid pattern with 60 cm interplant spacing for the independent culling selection method, on Riverdale silty loam soil at Winnipeg in 1983. Three selected groups per cross and generation were developed using honeycomb, independent culling and random selection. The randomly selected population served as a check against which to evaluate the effectiveness of the honeycomb and independent culling selection methods. Response to selection, as measured by the differences in mean yields among the three selection groups determined by replicated yield trials of derived F3 or F5 families indicated that neither selection method was effective. Response to selection, as measured by the number of high yielding families retained by either honeycomb or independent culling methods compared with the number retained by random selection also indicated that neither selection procedure was effective. Correlations between F2 and F4 single plant characters and plot characters on corresponding F3 or F5 families were low and generally nonsignificant. Heritability of single plant yield ranged from zero in most cases to 0.44 for one cross, generation and selection method. Heritabilities for both total dry matter and harvest index were not significantly different from zero. It is concluded that the response to selection is too small for either selection method to warrant use in faba bean breeding programs.Key words: Vicia faba L., early generation selection, heritability, faba bean