Heritabilities and Genetic Correlations from a Selection Experiment in Broiler Breeders using Restricted Maximum Likelihood.

Abstract

POLLAK, 1980) avoids bias due to selection (SORENSEN and KENNEDY, 1984). BESBES et al. (1991) have noted that the application of these methods in poultry breeding is not as widespread as in other species such as cattle. Especially in Japan there are no published reports on genetic parameters estimated by REML for broiler breeders even though their breeding has become important for the development of specialized broiler breeder lines. The objective of this study was therefore to estimate heritabilities and genetic correlations by REML with an animal model for some body weight and egg production traits in nine selected lines of broiler breeders. Materials and Methods 1. Data Data were obtained from selection programs in five female lines (13, 14, 16, 18 and 19) of White Plymouth Rock breed and four male lines (52, 53, 54 and 55) of White Cornish breed at the National Livestock Improvement Center, Hyogo Station, from 1985 to 1991. Each line of both breeds had approximately 1500 male and 2000 female chicks replicated in three houses each year. The selection procedure involved culling of birds on body weight at 7 weeks of age. At this age about 30% of males and 72% of females were selected in the female line White Plymouth Rock breed. For the male line White Cornish breed about 28% and 57% of males and females, respectively, were selected in each house. Chicks were then separated according to sex in each house. At 17 weeks of age the birds were also culled based on visual appraisal of the shape of toes and shanks and color of feathers. Measurements were also taken on body weight at 30 weeks of age, egg weight in the 34th week and age at first egg in all lines of both breeds. Any particular type of measurement was taken on the chicks in the same season of the year. Parents for the next generation were selected between 40 and 41 weeks of age. With reference to the initial population the proportions of sires and dams selected at this age were about 4% and 23%, respectively, for the White Plymouth Rock and 3% and 23%, respectively, for the White Cornish. Sires were selected on the basis of their previous body weight measurements and full-sib egg production while dams were selected on the basis of the dam's own body weight and egg production. One sire was mated to about 10 dams chosen in such a way as to minimize inbreeding. Lines 16, 52, 53 and 54 were selected for 6 generations each, while lines 13 and 14 were each selected for 5 generations and lines 18,19 and 55 were selected for 4 generations each. Feeding was ad libitum up to 7 weeks of age but restricted feeding was applied thereafter. 2. Statistical procedures Univariate analyses were carried out on measurements of body weights at 7 and 30 weeks of age, 34th week egg weight and age of hen when her first egg was laid in each line of both breeds. Single records were available per bird. The mixed linear model included breeding value and residual random effects, inbreeding coefficient as a covariate and three fixed effects-year, house and sex. For lines 13, 14 and 16 there were a maximum of six levels of year and three levels of house. For lines 18, 19, 52, .53,