Introduction
The commercial turkey is the result of a cross of a commercial sire line (or sire line cross) and a commercial dam line (or dam line cross). In general, early research indicated that nonadditive genetic variation was not a major contributor to total genetic variation in growth traits. In earlier studies, heterosis in body weight was observed in some crosses (Asmundson, 1942, 1948; Jerome et al., 1960; Friars et al., 1963) but not in other crosses (Knox and Marsden, 1944; Kondra and Shoffner, 1955; Jerome et al., 1960; Clark, 1961; Nestor, 1971). In some cases, heterosis was observed in body weight only at certain ages (Asmundson and Pun, 1954; Friars et al., 1963). Using diallel crosses, McCartney and Chamberlin (1961) concluded that additive genetic variance was much more important than non-additive genetic variance in body weight and body conformation measurements.
More recently, Emmerson et al. (1991) observed significant heterosis in body weight, particularly at older ages, and shank length in F1 and F2 reciprocal crosses of an experimental line (F) selected long-term for increased 16-week body weight and a commercial sire line (N) no longer used commercially. No heterosis in body weight was observed in a cross of an experimental egg line and the F line at 8, 16, and 20 weeks of age and heterosis was observed only at 20 weeks of age in a cross of the egg and N lines (Nestor et al., 1997). In crosses of an experimental line (FL) selected for increased shank width and an unimproved commercial sire line, Ye et al. (1997) observed significant heterosis in body weight and shank length. When the FL line was crossed with two improved commercial sire lines (A and B), significant heterosis in body weight was observed for females of the cross involving Line A and for males of the cross involving Line B (Nestor and Anderson, 1998). The results of the most recent studies suggest that nonadditive genetic variation may be evolving into a more important source of variation in body weight than was observed in earlier studies as would be the case if additive genetic variation in body weight is decreasing. In the F line, Nestor et al. (2000) found that the realized heritability, a measure of additive genetic variation, of 16-week body weight declined slightly over 30 generations of selection. The purpose of the present study was to study the genetic variation of growth traits in the F, A, and B lines and in reciprocal crosses among them.