Phenotypic and genetic parameters for fertility in sheep, in terms of number of lambs born and number weaned, have been estimated in a flock of medium Peppin Merinos. Repeatability of fertility traits over all ages is low, but there is an age effect, the record at 3 years of age having a higher value than those at 2 or 4 years. The regression of subsequent performance on a difference of 1 lamb at the initial lambing was higher for the difference between 1 and 2 lambs than between 0 and 1 lamb, which indicated that selection for twins is likely to raise fertility in the current flock more rapidly than selection against barrenness. The estimate of heritability for 2 years of age was negligible for each fertility trait, but the estimates for the 3-year-old record were higher, the value for lambs born being over 0.3. Mass selection for number of lambs born at 3 years of age would thus be expected to lead to appreciable genetic progress, while there would be at least some progress in number of lambs weaned. Phenotypic and genetic correlations were also estimated between fertility at different ages and 10 sheep and wool traits measured at 15–16 months of age. Phenotypically, fertility is positively correlated with body weight and negatively with skin wrinkle score. Genetically, it is suggested from the estimates that fertility is positively correlated with body weight and staple length and negatively correlated with fibre diameter, clean scoured yield, and wrinkle score. No phenotypic or genetic correlation was found between greasy or clean wool weight and either measure of fertility. The application of the findings to breeding for higher fertility is discussed. In an appendix, a maximum likelihood method of estimating heritability for all-or-none traits in half-sib data is presented.

Estimates of death rates are given for four mating groups in a flock of approximately 1000 Merino ewes, for each year of age from 1½ to 10½. Three of the groups are under selection for clean wool weight and other characters, and the fourth is an unselected control. The flock is run under extensive grazing conditions, and death rates are based on ewes dead or missing between one pen-mating and the next. Between 1951 and 1957, the average death rate for ewes 1½ to 7½ years old was 2.2 per cent. per annum, while at the older ages the average was no higher than 7.3 per cent. During the drought year 1957-58 losses averaged 3.8 per cent. for ewes up to 6½ years old, then rose steeply with age to 45.6 per cent. for ewes 9½ years old. In neither period did the selected groups differ from the unselected control. Starting with the 1953 drop, all rams in the unselected control group have been retained for the study of age effects on them. Average losses to date in each age group have been less than 5 per cent.

Repeatabilities of 10 traits, namely, greasy wool weight, clean scoured yield, clean wool weight, body weight, wrinkle score, face cover score, fibre population density, fibre diameter, staple length, and crimps per inch, were estimated in ewes and rams. Estimates ranged from 0.4 to 0.7 for the 10 traits in both sexes. In general, body weight was the most repeatable trait and face cover the least. There is no evidence of sex differences in repeatability of the traits studied. Correlations between pairs of records obtained at different ages were high in both rams and ewes. This finding, together with the high repeatability values for all the traits, suggests that selection of sheep based on a single record is probably sufficiently accurate for lifetime performance of these 10 traits in Merinos. The correlations between yearling records and subsequent records in ram were found to be in agreement with correlations between records taken at older ages. Early selection of rams for lifetime performance therefore is probably almost as efficient as selection at a later age, and it is suggested that this finding might also be found to apply to ewes.

Three genes have now been identified as affecting horn growth in the Merino. Test matings have shown that these are alleles at one locus, or comprise two pairs of alleles at closely linked loci. The genes are named P, P', and p, and their effects on horn growth in each sex are tabulated.

Estimates were made of the effects of the following factors on 10 fleece and body characteristics measured on breeding ewes aged 1½ to 10½ years in three mating groups over a period of 15 years: age of ewe, single or twin birth, age of dam, the ewe's own lambing performance, the year in which measurements were made, and the year in which each set of ewes was born. Two groups (S and MS) were under selection for high clean wool weight at 15–16 months, with a ceiling on wrinkle score and fibre diameter, while the third (C) was a random control. Changes with age were present in all characteristics and were similar in the three groups. The finding that selection on wool weight at an early age had no effect on subsequent age changes in any characteristic is of considerable importance. Greasy and clean wool weight reached a maximum at 34 years, then declined by 0.3–0.2 1b per year. Percentage clean yield, fibre diameter, body weight, and wrinkle score had maxima at 5½ to 6½ years. Staple length fell consistently by approximately 0.2 cm per year, while face cover rose consistently but slightly. Crimp number rose, fell, and rose again, while fibre number rose, fell, and remained constant from 4½ years. The chief source of increase in wool weight from l½ to 3½ years was an increase in the total number of fibres. The chief source of the subsequent fall was a decrease in fibre volume, with a minor contribution from a fall in total fibre number after 6½ years. Twin-born ewes cut 0.21 lb (4.2% of the mean) less clean wool per year over their lifetime than single-born ewes, while the progeny of 2-year-old ewes cut 0.32 lb (6.4%) less than the progeny of adults. The main source of lower weight in each case was a lower total fibre number. Pregnancy lowered clean wool weight more than lactation, the separate effects being 0.87 and 0.38 lb respectively (17.4 and 7.7% of the mean) and the combined effect 1.25 1b or 25.1%. Pregnancy lowered total fibre number but lactation had no further effect. Mean clean wool weights over all ages in the C group varied from year to year, the range being from 1.08 lb (21.6%)below the mean to 0.97 lb (19.4%) above. Differences in total fibre number contributed between one-third and two-thirds of the variation. Ewes born in consecutive years in the S and MS groups showed marked upward trends in clean wool weight, fibre number, and staple length, with a marked downward trend in crimp number and a slight upward trend in body weight. These trends demonstrate direct and correlated responses to the strong selection for high clean wool weight at 15–16 months of age, and the associated slight selection against fibre diameter and wrinkle score. The mean annual increases in clean wool weight were 0.15 and 0.11 Ib (3.0 and 2.2%) in the S and MS groups, approximately 40% of the increase arising from increased total fibre number and 40% from increased staple length. The effects of age and lambing performance can be used to predict productivity in flocks of differing age structures. As the casting age rises to 54 years changes in productivity are negligible. With a rise in casting age to 7½ years the average clean wool weight of the flock would fall by 0.14 lb, with a slight decrease in staple length and crimp number. These changes need to be balanced against any increased lambing percentage or decreased annual genetic gain due to increased generation interval. Comparison with other available figures indicates that age changes may vary from one area to another.