Linearity of offspring–parent regression in barley
A study was conducted to test the linearity of offspring–parent regression in barley. Forty-two F2 familes and 398 F∞ (i.e., doubled haploid) lines were derived from 7 × 7 diallel crosses. The F2 families, F∞ lines, and their parents were evaluated in hill plots with five replications. Three types of regression analysis (F2 family on midparent, F∞ family on midparent, and F∞ line on midparent) were carried out for grain yield, heading date, plant height, and number of spikes per hill. The linear regression mean square was significant in most cases, but the quadratic regression mean square was not significant in all cases. These results suggest that the true regression of offspring on midparent is linear, although dominance, additive × additive epistasis, and maternal effects were detected for the four traits of the same materials in previous studies.Key words: barley, Hordeum vulgare L., curvilinearity, regression, heritability.