Estimates of Genetic Variance in ‘Cherokee’ Alfalfa (Medicago sativa, L.)1

Four blocks of clones (17, 18, 19, and 20 clones per block) taken at random from ‘Cherokee’ alfalfa (Medicago sativa L.) were evaluated in a partial diallel mating design. Vegetatively propagated clones and four cross progeny from each clone were evaluated in three replications at two locations for 2 years in the upper coastal plain of North Carolina. Variance components estimated were interpreted in terms of autotetraploid inheritance with alpha = 0. Significant estimates of total genetic variance (δ2c), general combining ability variance (δ2g), and covariance of parent‐offspring (δpo) were obtained for yield, recovery after cuts 1 and 2, spring growth, and fall growth. Estimates of specific combining ability variance (δ2s) were not significant for any character. However, estimates of δ2c were significantly larger than estimates of 2 δpo or 4 δ2g for forage yield, recovery after cut 1 and spring growth indicating that either trigenic, quadrigenic, or epistatic genetic variance was important. Predicted gain in yield from selecting the upper 10% of the clones and recombining them into a synthetic was 9.3%. Because both additive and non‐additive genetic variance are important in Cherokee, additional gain might be made by producing hybrid varieties.