Heritability or the genetic regression of one trait on another, for example live-weight gain in field v. performance test station conditions, can be estimated efficiently by rearing offspring only from high- and low-ranking individuals, and estimating the regression of offspring on parent performance. Formulae are given for the optimum proportion of available animals to select as parents when the relative numbers of individuals in the parent and progeny generation are fixed. Although the optimum depends on the value of the parameter to be estimated, the designs are very robust to poor estimates.

1. Analyses were based mainly upon entries in Volume 101 of the Hereford herdbook (horned section) for 1971, but also used details published between 1900 and 1970.

2. During the calving period July 1969 to June 1970, 2705 herds in Great Britain and Eire registered 17 297 calves. It was estimated that there were some 21 900 cows in pedigree herds; 2664 bulls were used for pure breeding, more than half in one-sire herds. During the period 1964 to 1970 approximately 60% of all bull calves registered by about 6 mo of age were later licensed for pure- or crossbreeding.

3. From a sample of calves, 87% of their sires and 37% of their dams had not been bred in the same herds where the calves were registered. More than half the herds had been established for fewer than 10 yr and the average age was 12·2 yr. The mean generation interval was 5·25 yr.

4. Pedigree analyses revealed a particularly well developed hierarchy within the breed; the effective number of herds at the great-great-grandparental level was only four. The dominant herd (Vern) had made a total genetic contribution to the breed of over 24% judged by the number of independent occurrences of Vern animals in the four-generation pedigrees of sample animals. The average increase per generation in inbreeding over the past 12 generations had been only 0·48%, and there was no evidence of any subdivision of the breed into separate strains.

A theoretical analysis of open nucleus breeding systems, in which there is some introduction of breeding females to the sire breeding nucleus, is presented. Numerical analysis of a discrete generation model shows that the rate of genetic gain may be increased by 10 to 15% by opening the nucleus when selection intensity in females is low. In sheep and beef cattle breeding the optimal structure would be to have about 10% of the population in the nucleus, to get half of the nucleus female replacements from the base population, and to use all nucleus-born females not needed as nucleus replacements for breeding in the base population. The genetic gain, however, is not very sensitive to variation in these parameters. The rate of inbreeding in such an open nucleus would be about half that in a closed nucleus of the same size.

1. Analyses are presented of 1628 calvings from 397 pedigree Hereford cows kept in three herds whose records were made available by their owners. In addition a further 541 calvings were examined from a sample of 100 cows taken at random from the herdbook entries.

2. Mean age at first paternity for 71 homebred bulls was 25·6 mo. Mean age at first calving varied between the three herds from 29·5to 32·8 mo, with 39·7 mo for the random cows.

3. Random herdbook cows produced an average of 3·3 calves per cow compared with 4·2 from those in the three herds; these latter cows had an average calving interval of 13·1 mo.

4. The pattern of cow and calf disposal varied partly according to the differences in herd age and culling policy. While one herd retained more than 80% of heifers and subsequently sold half of them after two or three calvings for breeding, another retained only one-third but kept these for their full working life. Similarly the percentage of bull calves sold for breeding varied between 43 and 83%.

5. The net reproductive rate, or number of calves weaned per 100 cow-herd-years was 74, 88 and 67 for the three herds, and 66 for the random herdbook cows. The mean generation intervals were 4·1 yr for sires and 5·8 yr for dams whose breeding life in the herds had been completed.

6. The biases in reproductive statistics estimated from herdbook entries were examined, and it was concluded that such data can be of use where better herd records are unavailable.

1. Three experiments have been carried out to study the effect of adding trichloroethyl adipate, a compound known to prevent the production of methane both in vitro and in vivo, to the diet of sheep on fermentation in the rumen and on the growth of lambs. In Experiment 1 the additive was incorporated at various levels for 6 weeks, in Experiment 2 it was given at one of four levels for 12 weeks and in Experiment 3 it was included in the food of growing lambs for 15 weeks.

2. Initially, in Experiments 1 and 2, the additive caused a large reduction in methane production and in the molar proportion of acetic acid in the rumen liquor with a corresponding increase in propionic acid. These effects were much reduced after the first 2 to 3 weeks of administration.

3. The total food intake of the animals receiving the additive in Experiment 3 was 4% less than that of the control animals. This difference was negligible at the start of the experiment but increased throughout the feeding period.

4. On average, the relative live-weight gain and the changes in the weights of carcass, viscera and empty body weight of the animals receiving the additive were also 4% less than those of the control animals.

5. The observed differences in weight gain resulted from the differences in food intake and the additive had no beneficial effect upon lamb growth. This lack of effect on growth was due to rapid adaptation by the rumen fermentation to the presence of the additive.

1. Gestation diets formulated to contain A, 9; B, 11; C, 13 or D, 15% protein, and lactation diets with C, 13 or E, 17% protein provided eight treatment combinations. They were given at seven centres at 2 kg/day in gestation and approximately 5·7 kg/day in lactation to sows which remained on experiment for three to five parities and produced in total 468 litters. Mean results for parities one to four are given in this summary.

2. For gestation diets A to D mean litter size at birth was 9·7, 10·1, 10·3 and 10·4 live pigs, and at weaning (at 6 weeks), 8·2, 8·7, 8·8 and 8·8. Sows which received lactation diets C and E subsequently produced 10·4 and 10·3 pigs at birth and 8·9 and 8·3 at weaning.

3. Gestation diets had no consistent effects on piglet weights. During lactation, however, diet E gave heavier pigs than C, by 0·2 kg at 3 weeks and 0·5 kg at weaning.

4. Extra protein for either gestation or lactation had a positive effect on sow weight at that stage, but a negative effect in the following stage. There were no direct treatment effects on the time required for re-conception.

5. Twenty experimental sows at the Rowett Institute, together with an additional group given diet A for both gestation and lactation, showed strong positive relationships between lactation protein intake and milk yield or composition.

6. The treatment combination AC (i.e. 9% protein in gestation and 13% in lactation) gave the lowest weight of weaned pigs per litter, and CC gave the highest. The remaining treatments gave similar weights; of these BC (nominally 11 with 13, but actually 11·5 with 13·5% protein) was concluded to provide the lowest protein intake consistent with satisfactory performance.

Data for 432 litters of pigs were collected from 1965 to 1975 inclusive, at the Fasola Stock Farm, near Oyo in the Western State of Nigeria. The breeds represented were Large White, Duroc and Hampshire, and their crosses. The main objective was to investigate the effects of certain identifiable environmental sources of variation on litter performance characteristics such as litter size, mortality and weight at birth, 3 weeks and 8 weeks of age. Breed and year effects were significant for the litter size traits, but parity of dam, season and breed × season effects were significant only for litter size at birth which increased up to the fifth parity and then declined steadily in subsequent parities. Postnatal pig mortality was affected significantly by breed, litter size at birth and breed × season interaction. It was found that the larger the size of litter at birth the higher the postnatal mortality. Average pig weight at birth was reduced by approximately 0·01 kg for each additional pig in the litter.

Data collected on the Meat and Livestock Commission's pedigree recording scheme have been analysed by least-squares procedures to measure the effect of environmental factors on calf performance. In total 5524 birth weights, 16 484 200-day weights and 6248 400-day weights were available for 765 herds and seven breeds. The model adopted included terms representing type of cattle within breed, sex of calf, age of dam, year and month of birth, pre-weaning management, herd average level of performance and herd size as fixed effects with sires as random effects. The model accounted for between 19 and 36% of the variation in birth weights and for 36 to 49% and 39 to 69% of variation in 200- and 400-day weights respectively, depending upon breed. For birth weights and 200-day weights age of dam was a significant effect in all breeds. Sex of calf was the main source of variation at all ages, its relative importance increasing with age. Calves from 2-yr-old dams had 200-day weights 5 to 11% lower and calves out of 3-yr-old dams 2 to 9% lower than calves from 5- to 8-yr-old dams. Month of birth was a significant source of variation in most breeds. Pre-weaning management system was confounded with sex and breed. Correction factors for use in the MLC recording scheme were derived from the data.

The possibility that embryo survival in a breed of sheep may be related to the number of embryos normally present after natural mating was investigated by transferring large numbers of embryos to sheep of breeds with an approximately two-fold difference in ovulation rate. When the same number of embryos (14) was transferred to each of 12 Scottish Blackface and 11 Finnish Landrace ewes survival was greater in the Finn ewes (10%) than in the Black-faces (4%). The survival of 28% of embryos when transferred in groups of 7 to each of 14 Scottish Blackface ewes indicated that survival was not similar in the two breeds when the number of embryos present was a similar multiple of their natural ovulation rates. When the number of embryos transferred to Scottish Black-face ewes increased from 7 to 14 the decline in survival rate was so great that the number of lambs born also declined, indicating an intermediate optimum ovulation rate for maximum litter size.

The reproductive performance of 51 Finnish Landrace × Dorset Horn (FD) ewes was compared with that of 50 Border Leicester × Scottish Blackface or Greyface (GF) ewes. Ewes were mated without hormone synchronization in November 1973 and with hormone synchronization in August 1974 and again in February 1975. Conception rates at the three matings were 88, 82 and 73% respectively in FD ewes and 96, 56 and 26% respectively in GF ewes. Mean litter size over three lambings was 2·34 in FD ewes and 2·15 in GF ewes. Stillbirths accounted for 12·0 and 10·2% of all lambs respectively and further lamb deaths during and after birth were 15·2 and 4·3% respectively; 11·6% of live FD lambs and 7·7% of GF lambs were fostered or reared artificially. The losses resulted in mean numbers of lambs weaned per ewe lambing of 1·75 from FD ewes and 1·85 from GF ewes (including lambs from those ewes that were artificially reared). Annual productivity was 2·13 and 1·67 lambs respectively.

1. Of 80 gilts reared on a restricted plane of nutrition, 75 were allocated to one of four groups. Group 1; 17 gilts mated at puberty were slaughtered after farrowing their first litter. Group 2; 20 gilts mated at puberty were observed for three parities. Group 3; 18 gilts mated at second oestrus were observed for three parties. Group 4; 18 gilts mated at third oestrus were observed for three parities. Group 5; nine additional gilts reared on feed ad libitum and mated at first oestrus were observed for three parities. All animals received the same ration after the first mating.

2. The mean age (±SD) of the restricted gilts at puberty was 186 ± 17 days at a weight of 89 ± 9 kg. The gilts fed ad libitum were significantly (P<0·01) older and heavier at puberty (203 ± 15 days and 120 ± 27 kg).

3. Conception rates were 83, 86, 94 and 82% for Groups 1 and 2 combined, 3, 4 and 5 respectively.

4. Gilts mated at first, second and third oestrus produced, respectively, 7·8, 9·8 and 10·4 ± 2·2 piglets in the first parity. Group 5 gilts had 11·0 ± 2·2 piglets. All significant differences were confined to the first parity.

5. By the end of the third parity there were no differences between the restricted gilts mated at first, second or third oestrus in total number of piglets born or reared.

6. The gilts slaughtered (at 135 kg) after producing one litter yielded carcasses which were heavier but leaner than those from contemporary maiden gilts grown to 120 kg on a semi-restricted feeding scale.

Data collected in the Meat and Livestock Commission's pedigree recording scheme have been analysed by least square procedures to assess the relationship between, and heritability of, birth, 200-day and 400-day weights. In total 5524 birth weights, 16 484 200-day weights and 6248 400-day weights were available for 765 herds, in which were represented 2031 sires of seven breeds. Heritability estimates were calculated using the complete data and a set restricted to the larger herds and sire progeny groups. The influence of single effects on the heritability estimates was computed. Linear relationships were established between birth and 200-day weights in the Aberdeen-Angus and Charolais breeds but in the Hereford, Devon and Sussex breeds they were curvilinear. The data from the Hereford breed indicated an optimum range of birth weights. The partial regressions of 400-day on 200-day weights were positive; they were not linear for the Charolais, Devon and Sussex breeds. In these cases there was a more than proportionate increase in 400-day weights for an increase in 200-day weights. The heritability estimates varied widely and were dependent on the method and underlying model used. The most reliable estimates were as follows: birth weight 0·23, 200-day weight 0·38 and 400-day weight 0·27.

Three mature rams in each of four naturalized breeds (Romney, Border Leicester, Cheviot and Dorset Down) and two or three 15-mo-old rams in each of four recently imported ‘exotic’ breeds (Finnish Landrace, East Friesian, German Whiteheaded Mutton and Oxford Down) were mated to between 14 and 32 Romney ewes in a pedigree pen-mating system at Mana Island, New Zealand. Two or three 60-min observations of mating behaviour were made on each ram.

There were no significant differences in mating behaviour scores between groups of breeds, but the mature rams of the naturalized breeds were more fertile than the young rams of the exotic breeds (76% and 57% ewes not returning to first mating respectively). Within the young exotic breed group there were significant differences between breeds in number of serves per test, interval to first mount and fertility. Fertility was negatively correlated with time to first mount.

Herbage intake and grazing times were measured in four experiments in which adult sheep grazed short and long pastures. The digestible organic matter (DOM) intake of adult wooled Merino and Corriedale wethers was higher, both absolutely and per kg live weight, when they were lower in live weight due to prior under-nutrition. The relative differences were similar on both abundant and scant pasture and were proportional to the relative differences in live weight. The higher intakes were achieved by differences in the time spent grazing and in the rate of intake per hour of grazing. DOM intake of shorn sheep was increased by up to 44% per kg live weight within 5 weeks of shearing in two experiments when the mean temperature was 8°C but no increase in DOM intake was observed in four other experiments when the mean temperature was 11°C. Shorn sheep achieved higher intakes on both abundant and scant pastures by increasing intake per hour of grazing.

The effects of supplementary feeding of licks containing non-protein nitrogen on the reproductive performance of beef breeding cows and on the live-weight gain of growing stock in Botswana is described. The trials involved 1375 breeding cows at five artificial insemination centres, and 360 breeding cows and 269 growing stock under ranch conditions. These licks increased the percentage of pregnancies in cows under lactation stress by up to 20%, but had no effect on the reproductive performance of dry cows or cows with a calf older than 5 mo. In growing animals the provision of rumen stimulatory licks increased live-weight gain by an average of 12%, but was of doubtful economic value.

Voluntary intake of a range of feeds by cows is predicted on the assumption that the cows will satisfy their energy requirements unless physical limitations intervene. Predicted patterns of intake throughout lactation agree generally with observed changes, including an eventual plateau in body weight due to a predicted physical equilibrium between abdominal fat and gut contents. An attempt to incorporate a depression in milk yield when physical limitations prevented energy requirements being met by food intake was successful only with good quality feeds.

The three indigenous beef cattle breeds in Botswana, the Tswana, Africander and Tuli, were evaluated for reproductive performance, viability and growth. Over 3 yr the calving percentages of 1389 Africander cows, 305 Tswanas and 357 Tulis were 64·5, 70·6 and 85 respectively. These differences were significant (P<0·05). Mortality of the Africander calves was higher to 2 yr of age, being 11% compared with 7-5% for Tswana calves and 7·4% for Tulis. The 18 mo weight of 219 Tuli cattle was 284 kg, and of 454 Tswanas 279 kg, both superior to the 270 kg for the 786 Africanders. Productivity estimates combining these three traits demonstrate the value of the Tuli breed for Botswana conditions.

Comparisons of the crossbreds produced by the use of Simmental, Brahman, Bonsmara and Tuli sires on Tswana cows showed an advantage in growth to 18 mo through crossbreeding. The 18 mo weights of Simmental and Brahman crosses at 324 and 304 kg respectively were significantly different, and superior to the weight of pure Tswana at 279 kg. Bonsmara and Tuli crosses at 294 and 290 kg also showed a significant increase over the pure Tswana. The use of the Simmental is now recommended for better management conditions, with Brahman crossbreeding on a wider scale.

1. Oxygen consumption was measured in two experiments using castrated Romney male sheep both shorn and unshorn. The sheep were fed either ground hay or dried grass. An ambient temperature of 13°C ± 1°C was maintained at all times. In Experiment 1, oxygen consumption was measured with eight sheep using a respiration chamber and in Experiment 2, oxygen consumption was measured with four sheep using a ventilated hood. The times spent standing, lying, eating and ruminating were measured in Experiment 2. Heat production was calculated from the measured values for oxygen consumption.

2. Heat production increased following shearing in both experiments. In Experiment 1, heat production of the sheep receiving dried grass was significantly greater (P<0·05) on days 1, 8 and 16 following shearing, compared with pre-shearing values. For sheep receiving ground hay, heat production was greater on day 1 (P<0·05) and day 8 (P<0·01) following shearing, compared with pre-shearing values. The mean maximum increase in heat production following shearing was 2·3 MJ/24 hr (25% increase) for sheep receiving dried grass and 1·2 MJ (17%) for those receiving ground hay.

3. The patterns of change in heat production following shearing in Experiment 2 were similar to those observed in Experiment 1 although the maximum increases were smaller: 21% for the sheep receiving dried grass and 13% for those receiving ground hay.

4. The time spent standing increased considerably after shearing and this change in behaviour made a large contribution to the increase in energy expenditure after the sheep were shorn.

Teat canal stretchability was measured on 119 lactating cows from two dairy farms, using a probe. Diameter of one hind teat was measured by calipers at the proximal end of the teat and 20 mm above its distal end. Teat canal stretchability, teat diameter at the proximal end and teat diameter gradient were highly correlated with milk yield. Cows with higher teat diameter gradient (i.e., funnel-shaped teats) had higher teat canal stretchability values compared with the cows with lower teat gradient values (i.e., cylindrical teats). Stretchability of the hind teat canal was significantly greater than the stretchability of the fore teat canal.

The patterns of seasonal variation in reproductive activity were observed over a period of 15 mo for approximately 15 females of each of three breeds: Finnish Landrace (Finn), Tasmanian Merino (Merino) and Scottish Blackface (Blackface). The incidence of oestrus was measured by teasing with vasectomized rams, and the incidence and rate of ovulation were determined frequently by laparoscopy. Luteal function was assessed from peripheral venous blood progesterone concentration on days 7 and 11 of the oestrous cycle. Nutritional status was monitored by recording body weight and plasma-free fatty acid levels throughout the study.

The breeding seasons differed significantly: Finn, October to May; Merino, September to February; and Blackface, October to February. Variation in the incidence of ovulation was similar to that in the incidence of oestrus for each breed. The incidence of silent ovulation varied with the breed, being greatest in the Merino and least in the Finn. The ovulation rate varied among breeds (Finn, 2·99; Merino, 1·08 and Blackface, 1·30), and during the breeding season (e.g. Finn: November, 3·5; March, 2·6). Follicles were observed in each breed throughout the period of study.

The pattern of variation in progesterone concentration was similar for each breed despite their different breeding seasons. In addition to breed differences in ovulation rate and in onset and end of the breeding season, the sensitivity to oestrogen was apparently such that, with the Finn if oestrogen secretion was high enough to stimulate ovulation it would usually also stimulate oestrus, whereas with the Merino ovulation often occurred without oestrus; this suggests that in the Merino the centres controlling ovulation are more sensitive to oestrogen than those controlling behaviour.