PREDICTED AND REALIZED GAINS FROM SELECTION FOR INVITRO DRY-MATTER DIGESTIBILITY AND FORAGE YIELD IN SWITCHGRASS

Improved forage yield and quality, which can lead to more efficient li vestock production, are important goals in switchgrass (Panicum virgat um L.) breeding. Objectives of this study were to determine the effect iveness of multiple cycles of recurrent restricted phenotypic selectio n ( RRPS) in improving forage yield in switchgrass and in improving in vitro dry matter digestibility (IVDMD), and to compare predicted and realized gains from selection. Six switchgrass populations were develo ped by selecting a base population for one cycle of low and three cycl es of high IVDMD and two cycles for high yield. Half-sib progenies of parents selected from the base population for high and low IVDMD were used to estimate heritability for IVDMD by parent-progeny regression. Realized gains from selection were determined in a seeded sward trial that included the base and six selected populations. Estimated and rea lized heritabilities for IVDMD were 0.40 and 0.31, respectively. Predi cted pins and (in parenthesis) realized gains from selection for IVDMD in grams per kilogram per cycle were - 21 (- 15) for the Low IVDMD Cy cle 1, 16 (15) for the High IVDMD Cycle 1 ('Trailblazer), 14 (4) for t he High IVDMD Cycle 2, and 19 (18) for the High IVDMD Cycle 3 populati ons. There was a low but positive genetic correlation between forage y ield and IVDMD of 0.10. Forage yield was not improved by RRPS, probabl y because of lack of genetic variability for forage yield in the base population. It is concluded that RRPS is effective for improving IVDMD in switchgrass, but was not effective in improving forage yield in th e populations evaluated, and that predicted gains exceeded realized ga ins from selection for IVDMD due to limited selection for traits other than IVDMD.