RECURRENT SELECTION FOR INCREASED GRAIN-YIELD IN OAT

Grain yield is a primary consideration in choosing oat (Avena sativa L.) cultivars. This research examines gain from five cycles of recurrent selection for grain yield in oat, and compares the yield responses when evaluated in hill and row plots. A secondary goal was to examine the yield components that contributed to the yield gains in the two plot types. Parents of all five cycles of recurrent selection and the checks 'Don', 'Hazel', 'Moore', 'Noble', 'Ogle', 'Starter, and 'Steele' were grown in hill and row plots in 1988 and 1989 at St. Paul and Rosemount, MN. Experimental design for both experiments was a randomized complete block and data were recorded on grain yield (g m-2), seed weight (g 100 seed-1) and seed number (seed m-2). Grain yield gain per cycle, as measured by weighted linear regression, was 15.1 g m-2 in hill plots and 12.8 g m-2 in row plots. Examination of the yield components seed weight and seed number indicated that the grain yield gain resulted from an increase in both traits. The relationship between seed weight and seed number was numerically different in the two plot types, with seed weight showing a larger increase per cycle in row plots and seed number a larger increase per cycle in hill plots. Phenotypic correlations of parent performance in the two plot types were 0.85 for grain yield, 0.81 for seed weight and 0.69 for seed number (P < 0.01). The large gain in grain yield per cycle and the excellent agreement in performance of the parents of all five cycles in hill and row plots indicate the effectiveness of recurrent selection in oat.