Six years of data collected from two shifts of Cornell random bred control matings each year are analysed. Approximately 900 progeny were tested per year from 50 sires mated to 250 dams.
Means for the important economic traits are presented by years and regressions for nine such traits, together with their standard deviations, are shown. The only significant regressions are those for both egg production and specific gravity score in shift II. An abnormally low egg production the first year for this shift is considered responsible for this result and the reality of the regression is discounted.
Genetic variances are presented by trait and year. Year to year variability is evident, but no trends are indicated.
Estimates of heritability and genetic correlations between traits are presented. These estimates are considered to be quite reliable and will be useful as estimates of genetic parameters when this population is used in selection experiments. Estimates of heritability from the sire components of variance were as follows: 8-week body-weight, 0·29; 32-week body-weight, 0·48; 55-week body-weight, 0·44; age at maturity, 0·09; egg production to 500 days, 0·07; 32-week egg-weight, 0·51; 55-week egg-weight, 0·53; USDA albumen score, 0·29; blood spot score, 0·13; and specific gravity score, 0·34.
Important maternal effects were found for 8, 32 and 55-week body-weight, sexual maturity and egg production. Selection programmes should take into consideration the fact that maternal effects accounted for 13% of the total variation in egg production. Genetic correlations utilizing combined sire and dam components revealed no serious antagonisms between egg production and other economic traits. However, this conclusion must be considered in the light of what influence maternal effects may have, since correlations computed from sire and dam components separately may have opposite signs and be important.
The many negative estimates of interaction components (SH, genetic by environment) and (I, sire by dam) indicate that these elements of the model contributed little to the variation of the traits studied.