Data preparation
Data analyses
Sheep
Individual records were built up for each ewe related to each parturition. These gave number, date of birth, origin and type of birth (single or twin), of the ewe herself, and the current parturition date, previous parturition date, type of birth, sex of lamb(s), viability of lamb(s) over 4 stages to 12 months of age, lamb weights and three linear measurements at birth, 2, 4, 6, 8, 10 and 12 months, and ewe's own weight and three linear measurements at parturition and 4 months after parturition. From these data, the age at first lambing, lambing interval, ewe mean weight and three productivity indices were additionally computed. Information on sires and dams of ewes and sires of progeny were noted whenever available. Eventual disposal or death of the ewe was recorded. All available causes of death were obtained from periodical reports and treatment records.
Cattle
Individual records were built up for each cow related to each parturition. These gave number, date of birth and origin of the cow herself and the current parturition date, previous parturition date, sex of calf, viability of calf over 4 stages to 12 months, calf weight and three linear measurements at birth, 3, 6, 9, 12, 15, 18, 21, 24 and 30 months, and cow's own weight and three linear measurements at parturition and 6 months after parturition. From these data the age at first calving, calving interval cow mean weight and three productivity indices were additionally computed. Information on sires and dams of cows and sires of calves were noted when available. Eventual disposal or death of the cow was recorded. All available causes of death were obtained from periodical reports and treatment sheets.
All characters were analysed by least-squares procedures (Harvey, 1972) using both fixed and mixed models. In the analyses, some parameters were measured for further study in their own right and some were measured to account for variation arising, in order that less biased estimates of other parameters could be obtained. Unequal and disproportionate subclass numbers gave unbalanced factorial designs for which conventional analyses of variance techniques were not applicable.
With both sheep and cattle, a considerable number of foundation animals had no sire or dam records. In addition the use of individual sires was often confounded with month and year of use, so use could not be made of sire group in analyses at this stage in time.
Typical models used included the fixed effects of origin (purchased or born on station); parturition number within origin; year of birth or parturition; month of birth or parturition; sex of calf; type of birth (single or twin); and the random effects of individual cow or ewe, nested within origin. The specific factors included in the model used will be evident when results are presented for each character analyzed.
The residual mean square was used as the error term to test the significance of all differences evaluated, except origin, when cows or ewes within origin was used. Linear contrasts of least-squares means were computed to determine the significance of differences between groups for all characters where the difference among groups was significant in the analysis of variance. More comparisons were made using the least-squares means than there are independent degrees of freedom for the characters where the group differences were significant in the analysis of variance. Therefore, all of the comparisons are not independent and the error rate over the entire set of comparisons may be different from that indicated by the level of probability. Tests of significance associated with the linear contrasts, although not independent, can be taken as guides as to whether the observed values could have occurred by chance. Repeatabilities were calculated using the variance components between and within cows or ewes.
