ABSTRACT
Summary This research was undertaken to evaluate the effectiveness of within-half-sib family recurrent selection for seedling vigor, seed yield, and forage yield in birdsfoot trefoil (Lotus corniculatus L. cv. Leo). The effectiveness of selection was measured by comparing progeny performance, genetic variance, and heritability in selected and unselected populations. In addition, experimental synthetics (Syn-2 generation) from the selected population were compared to Leo for forage yield and seed yield in several trials. Seedling vigor and yield in 1979 and forage yield in 1980 of polycross progeny in the selected population were significantly higher than those of similar progeny in the unselected population. Forage yield in 1979 of polycross progeny was similar in the selected and unselected populations. Seedling vigor of selfed progeny in the selected progeny was at a level similar to that of the polycross progeny of the unselected population. Although selfed-progeny performance in the selected and unselected populations was similar in magnitude for seed yield and forage yield, relative decline was greater in the selected population. Estimates of genetic variance indicated significant and similar levels of additive variation for forage yield in row plots and nonsignificant dominance variation in both populations. Heritabilities were similar in the populations tested. Both the cycle-3 and cycle-4 populations advanced to the Syn-2 generation were significantly higher yielding in the year of seeding (a measure of seedling vigor), similar in the second year, and significantly higher in the third year when compared with 219Leo. Seed yield was below that of Leo for the cycle-3 population and above that of Leo for the cycle-4 population. Selections from cycle-4 for seedling vigor and seed yield were based on progeny testing. The seedling-vigor selection had significantly higher forage yield in the year of establishment than Leo, while the seed-yield selections had about 40% higher seed yield than Leo. Lack of observed significant levels of dominance variation and greater inbreeding depression in the selected population suggest that additional populations of diverse genotypic background would be necessary to obtain optimum performance when elite genotypes are used as parents for new cultivars. Performance of experimental synthetics from the selected population, which was significantly above that of Leo for forage yield in the establishment year and much better than Leo's for seed yield, is evidence of the success of this selection method in birdsfoot trefoil.
