Abstract
Genetic variation in the reproductive performance of heifers and cows was studied in AI [artificial insemination] data, comprising 57,058 records from open heifers and 45,649 records from cows in first and later lactations representing 3 AI units and 2 breeds, Swedish Red and White cattle (SRB) in the Skara and Malmen AI units and Swedish Friesian cattle (SLB) in Skane AI unit. Heritability estimates of reproductive performance for heifers were generally very low. The additive genetic variability in cow fertility results was of the same magnitude as in heifers. Estimates of heritability for interval measures were almost at the same level as measures of pregnancy rate, the calving interval being one exeption with a heritability estimate near zero. Despite the low heritability estimates, the genetic standard deviation for pregnancy rate may be of the order 8-10%, expressed as coefficients of variation. Genetic correlations between the various measures of pregnancy rate were close to 1, but corresponding phenotypic correlations were lower. The interval between calving and 1st insemination (CFI) was not genetically related to pregnancy rate. Correct conclusions about the efficiency of the different fertility measures as selection criteria in Swedish AI recording cannot be drawn without considering the whole selection program. Measures of pregnancy rate may be more efficient than other measures as selection criteria for the improvement of fertility in cattle by selection. The reproductive performances of heifers and 1st calvers may be closely related genetically. When progeny testing bulls for female fertility it is possible to use heifer records only. It is possible to use 1st calver records before progeny group results for milk yield become available, increasing accuracy in selection.