Honeycomb Design for Selection among Homozygous Oat Lines1

Abstract

The honeycomb design has been proposed for evaluating the yielding ability of single plants of small grains. The objective of this study was to assess the effectiveness of selection for grain yield via grain, adjusted grain, and biomass yields of single plants grown in a honeycomb design. Materials were 236 F10-derived lines of oats from a broad-based bulk population. They were tested as individual plants in two replicates of the honeycomb design in each of five environments and as lines in a hill-plot experiment with three replicates in each of three environments. Mean yields from the latter experiment were used as the standard for measuring selection effectiveness. Pooled correlations for grain, adjusted grain, and biomass yields of single plants with grain yield of oat lines measured in hill plots were 0.29 ± 0.06, 0.27 ± 0.06, and 0.37 ± 0.06, respectively. With a 10% selection intensity, the actual gain in grain yield (expressed as a percentage of the hill-plot population mean) of oat lines averaged 9.8 ± 0.05, 8.3 ± 0.06, and 11.0 ± 0.05, respectively, when grain, adjusted grain, and biomass yields of single plants were the selection criteria. Discarding of the lowest 10, 25, and 50% of plants for grain, adjusted grain, and biomass resulted in the average retention of 97, 90, and 70%, respectively, of the top decile of oat lines for grain yield. Our results suggest that selecting for grain or biomass yield of oat plants grown in a noncompetitive field arrangement can identify higher yielding oat lines. Selection for biomass gave about one-seventh more gain than did selection for grain yield. Adjustment of grain yield of a plant according to the mean of six neighboring plants in the honeycomb design reduced the predicted gains from selection.