RECURRENT PHENOTYPIC SELECTION FOR FORAGE YIELD IN RYE

Little progress has been made in the past 20 years in improvement of rye (Secale cereale L.) forage yields. In 1985, a recurrent phenotypic selection procedure was initiated to improve forage yield of 'Wrens Abruzzi' (WAL-Cycle 0) rye. The selection procedure included visual selection of spaced plants, grid selection, maintenance of a relatively large population size, bi-parental selection, a one-year cycle interval, and intermating of selected plants in isolation. Cycle four of selection in the WAL population was completed in 1989. Experiments were conducted from 1987 to 1990 to evaluate progress from selection over the first four selection cycles and quantify the association between visual forage potential ratings and fresh plant weight. Wrens Abruzzi rye and Cycle 1, 2, 3, and 4 populations were evaluated over a 2 year period in space-plant population progress trials at Tifton, GA. Wrens Abruzzi was evaluated in comparison to WAL-C1 (9 environments), WAL-C2 (18 env.), WAL-C3 (16 env.), WAL-C4 (5 env.) and rye check cultivars in small-plot clipping trials in a set of 36 southern U.S. environments. Spaced-plant evaluation indicated forage yield gais of 6 to 7% per cycle with little change in population variation. In seeded small-plot clipping evaluations WAL-C1, WAL-C2, and WAL-C3 produced similar or slightly higher forage yields as WAL-C0, but WAL-C4 produced 9 to 12% higher forage yields than WAL-C0. Available data from WAL-Cycles 1 to 4 indicates that forage yield gains in small-plot evaluation are 2 to 3% per cycle. High positive correlations between visual plant potential ratings and plant fresh weights indicate that visual selection is effective in identifying high-yielding rye phenotypes.