Changes in live weight, body condition, milk production and composition were monitored for single- and twinsuckling cows in two experiments. In experiment 1, all cows were managed for 3 months before calving to achieve condition scores at calving 2·1 (low) and 3·1 (moderate). Subsequently, these animals were placed on a fixed level of energy intake (80 MJ metabolizable energy (ME) per day) for 16 weeks. Post-partum diets contained differing amounts of digestible undegradable protein (DUP), but similar amounts of effective rumen degradable protein (ERDP). In experiment 2, single- and twin-suckling cows were fed to achieve a condition score at calving of around 2·75. These animals were subsequently placed on two levels of energy intake (80 MJ and 130 MJ ME per day) during the first 16 weeks of lactation. The diets providing these energy intakes also contained differing amounts of DUP. Neither the level of DUP in the diet nor the level of feeding significantly altered the digestibility of the dry matter or organic matter in the dry matter of the experimental diets.

Body condition at calving had little effect on daily milk yield (9·5 kg) and composition (33·5 g/kg fat; 30-0 g/kg protein) in experiment 1. Cows with higher body condition scores at calving lost 13·5 kg more live weight and 0·26 units more body condition (P < 0·05) during the 1st month post partum but remained in better condition throughout the trial period than cows with lower body condition scores. Cows on 130 MJ ME per day produced higher (P < 0·001) daily yields of milk, milk fat and protein (+2·0 kg; +151 g; +85 g) and lost 57 kg less live weight and 0·52 units less body condition (P < 0·001) than cows on 80 MJ ME per day. Elevated dietary levels of DUP promoted (P < 0·05) daily milk yields (by 1·25 kg) and reduced live-weight and body condition-score loss during the 1st month post partum in both experiments. Twin-suckling cows produced the highest average daily yields of milk throughout the 16-week trial period in both experiments but yields were not significantly higher after early lactation in experiment 1. Cow body condition score at calving and the level of DUP in the post-partum diet of cows had no effect on the post-partum performance of calves. Twin calves had slower growth rates (P < 0·001) than single calves (0·65 and 1·07 kg/day respectively) in experiment 1. However, twin calves sucking cows on 130 MJ ME per day in experiment 2 grew similarly to single sucking calves from week 8 of lactation (0·8 kg/day).

Thirty-four Dorset and 34 Suffolk rams were slaughtered in pairs within breed at birth, weaning, at monthly intervals to 420 days and then at bimonthly intervals until 600 days to compare factors used to estimate mature size and to determine how genetic size-scaling affects the comparison of body composition between breeds. At equal empty body weights there were no detectable differences between the breeds in amounts of water, protein or ash but Dorsets contained more lipid. Dorset and Suffolk asymptotic values of water (40 and 61 kg), water-plus-protein-plus-ash (51 and 77 kg), metacarpal bone length (127 and 147 mm) and metacarpal bone weight (59 and 101 g) were used to estimate mature size. Regardless of the estimator of mature size, there were no significant differences between breeds in degree of maturity at a given metabolic age. At the same degree of maturity within the estimators of mature size, there were no differences between breeds in the degree of maturity of water, protein or ash, with the exception of small differences in protein for two of the mature size indicators. Because lipid appeared to accumulate without limit in these sheep, the mature body was assumed to contain 300 g lipid per kg. Using this proportion to estimate amount of mature lipid, there was a higher degree of maturity of lipid for Suffolk than for Dorset rams at equal degrees of maturity of water, water-plus-protein-plus-ash or metacarpal bone weight. Thus, Dorsets with greater lipid content at equivalent empty body weights were leaner than Suffolks when adjusted for mature size.

Changes in certain blood metabolites, serum progesterone concentration and pregnancy rate, were monitored for single- and twin-suckling cows in two experiments. In experiment 1, cows were managed to achieve condition scores at calving of 2·1 (low) and 3·1 (moderate). Subsequently, these animals were placed on a fixed level of energy intake (80 MJ metabolizable energy (ME) per day) for 16 weeks. Post-partum diets contained differing amounts of digestible undegradable protein (DUP), but similar amounts of effective rumen degradable protein (ERDP). In experiment 2, all cows were fed to achieve a condition score at calving of around 2·75. These animals were subsequently placed on two levels of energy intake (80 MJ and 130 MJ ME per day) for 16 weeks. The diets providing these levels of energy intake also contained differing amounts of DUP. Pregnancies were established by embryo transfer following induced oestrus.

Cows in moderate body condition at calving subsequently had higher (P < 0·05) serum non-esterified fatty acid (NEFA) concentrations. Cows on 130 MJ ME per day had lower serum NEFA concentrations than cows on 80 MJ ME per day. Cows given higher DUP diets had higher (P <0 ·05) serum NEFA concentrations when in negative energy balance and lower serum NEFA concentrations when close to energy equilibrium than cows given lower DUP diets. Twin-suckling cows had higher (P < 0·05) serum NEFA concentrations than single-suckling cows, but only during the first 3 weeks post partum. Plasma glucose concentrations were below 3 mmol/lfor twin-suckling cows until week 7 post partum when they became similar to the values for single-suckling cows. Serum D-3-hydroxybutyrate concentrations mirrored those of plasma glucose. Serum urea-nitrogen concentrations were high for twin-suckling cows on 80 MJ ME per day and single-suckling cows on 130 MJ ME per day.

Neither body condition at calving nor post-partum energy intake significantly influenced the time taken for cows to overcome anovulation nor did they significantly influence the proportion of cows which eventually resumed normal oestrous cyclicity. DUP concentration in the post-partum diet had no overall effect on the various measures of fertility assessed but thin cows at calving subsequently placed on the high DUP diet took longer to ovulate (55 days; P < 0·05) and had the lowest pregnancy rate (0·30). The post-partum anovulatory period was longer (P < 0·01) by 27 days in experiment 1 and 12 days in experiment 2 for hvin-suckling cows compared with singlesuckling cows. A lower proportion (P < 0·001) of twin-suckling cows regained oestrous cyclicity and became pregnant in experiment 1. Fertility traits were improved when twin-suckling cows were in moderate body condition at calving and/or offered 130 MJ ME per day during the post-partum period.

Fleece weight and shedding score, a measure of premature fleece loss, were examined at first shearing and in older female sheep of three hilt breeds and their reciprocal crosses. Starting from a non-inbred base, the sheep were mated, mostly by younger-parent × offspring, for four generations, to produce inbreeding coefficients from 0·25 to 0·59. Crosses of inbred lines were also produced within breed type. The experimental design allowed the effects of inbreeding of the individual to be separated from the effects of maternal inbreeding.

Inbreeding of the individual significantly and linearly reduced fleece weight. This effect was still apparent after adjustment for body weight. Maternal inbreeding significantly reduced only the weight of first fleeces, but the trends were similar at the later ages, especially among the purebreds. There was no significant interaction of purebred/crossbred status with level of inbreeding. Inbreeding did not significantly affect shedding score. The pure breeds and the crosses did not differ, on average, in fleece weights or shedding scores, but within the purebred and within the crossbred classes breed variation was significant. Fleece weight declined and the incidence of shedding increased with increasing age. The larger the number of lambs born in the year of shearing the lower the fleece weight and the greater the extent of fleece shedding.

The effects of inbreeding could not be fully explained, statistically, in terms of dominance alone. Therefore, it seems probable that epistasis also plays a role in producing the observed changes with inbreeding in these traits.

A combined crossbred and purebred selection (CCPS) method, i.e. using crossbred and purebred information, was proposed to achieve genetic response in crossbred animals. Selection index theory was applied to establish a CCPS index. The CCPS was compared with pure-line selection (PLS) and crossbred selection (CS) methods. The genetic correlation between purebred and crossbred performance (rpc) and crossbred heritability (hc2) are crucial factors in the comparison. The CCPS is always better than PLS or CS when a fixed number of purebred progeny is tested. With a fixed total number of purebred and crossbred tested progeny, CCPS is only worse than PLS for very high values of rpc (>0·8). Superiority of CCPS over PLS increases and over CS decreases with decreasing rpc. The larger hc2 is, relative to purebred heritability (hc2 the more response CS and CCPS will achieve. The robustness of CCPS against inappropriate assumptions on rpc and hc2 values was investigated. The expected response is always an overestimate, and the actual response is smaller than the optimal response when rpc is assumed one but the true rpc is smaller. The difference between actual and optimal response increases as rpc decreases but it is small for large rpc values (e.g. <3% for rpc >0·7). The expected response is smaller than the actual response when rpc is large and hc2> hp2 Finally, the actual response to CCPS is larger than the optimal response to PLS for positive values for rpc. The main conclusions are: (1) CCPS method is optimal for obtaining genetic response in crossbreds; and (2) CCPS with inappropriate assumptions on rpc and hc2 values (e.g. recognizing crossbreds as purebreds) achieves more genetic response than PLS for common values of rpc and crossbred heritability.

Breeding schemes for ivool sheep based on natural service (N), MOET (multiple ovulation and embryo transfer) in adults (A) and MOET in juveniles (J) are compared using Monte Carlo simulation allowing overlapping generations and selection across age groups. In all schemes parents can be a maximum of 6 years old when their offspring are born, and can be a minimum of 2 years in N or A schemes or 1 year in f schemes. Embryo transfer success rates follow realistic distributions (only 70% of flushes resulting in live lambs) with a mean of 2·13 or 345 6-month-old lambs per donor per flush for current or ‘future’ situations. Under natural mating an average of 1·09 6-month-old lambs per ewe mated is assumed. The trait under selection has a heritability of 0·4 and is measurable on both sexes by 18 months of age, which is before selection in N and A schemes but after the initial opportunity for selection in f schemes where selection is based on a pedigree index. Two culling policies for males are considered in J schemes: all sires culled if not selected (J), or all sires available for selection after their record is collected even if they were not selected on their pedigree index in the previous year (f o ). Comparison of schemes is made on the basis of response depressed by the effect of inbreeding and cumulated and discounted over a medium time horizon. For a given population size (constant expected number of lambs per year and therefore numbers of dams dependent on fertility rates) different numbers of sires were utilized. The optimum number of sires was determined as the number which maximized the response criterion. N, A, J and J° schemes have different optimum numbers of sires (5, 8,30 and 20 respectively) which occur at different rates of inbreeding and which are approximately independent of population size. Advantages in the response criterion of A, J and J° over N schemes are 23%, 33% and 47% respectively for ‘future’ embryo transfer success rates and mass selection. Increasing the number of lambs born per donor ewe from 2·13 to 13·8 (by increasing embryos per flush and number of flushes) increases the response criterion by up to 25% in A and up to 11% in J schemes (assuming donor ewes are mated to a different sire for each flush) with 345 lambs born per year. Use of best linear unbiased prediction (BLUP) estimated breeding values increases response by 8%, 3% and 3% in N, A and J schemes respectively. The advantages off over A and N schemes, of increased lambs per donor and of BLUP over mass selection all increase as population size increases. Insemination of the poorer ewes with semen which gave only female offspring gave 5% increase in the N scheme and no benefit in A or f schemes.

Two experiments were conducted to investigate the effect of supplementation with fodder tree leaves on the utilization of teff straw. This ivas aimed at testing the hypothesis that forage supplements (e.g. sesbania) that disappear faster from the rumen compared with those that disappear slowly (eg. leucaena) may elicit higher intakes of the basal diet. Experiment 1 investigated the effect of level (i.e. 0,200,300 and 400 g/kg dry matter (DM) intake) of leaves of Sesbania sesban or Leucaena leucocephala on the utilization of teff straw fed ad libitum. Seven sheep were given the seven diets in four periods following a partially balanced (7 × 4) cross-over design. In experiment 2, teff straw alone or supplemented with graded levels of sesbania (250, 350 and 450 g/kg DM intake), were fed to 16 sheep in a completely randomized block design. Intake, degradability, nitrogen (N) utilization, particulate and liquid passage rates were estimated.

In experiment 1, protein source × protein level interaction was not significant. Supplementation increased intakes of total DM (P < 0-05), organic matter (P < 0·01) and N retention (P < 0·01) but had no effect on the intake of teff straw. Supplementation with sesbania, but not with leucaena, increased (P < 0·05) the apparent digestibility of DM. Both supplements increased (P < 0·05) the fractional rate of passage from the rumen. In experiment 2, graded levels of sesbania increased DM intake (P < 0·05), fractional rate of liquid outflow (P< 0·05) and tended to increase the rate of degradation. It was concluded that the effect of fodder tree leaves on roughage intake is a function of their disappearance rate from the rumen.

Benefits of collection of annual versus a single fleece record are investigated for breeding programmes of Australian Merino sheep. Breeding scheme alternatives are based on natural mating (N) or multiple ovulation and embryo transfer (MOET). The breeding objective is assumed to be an adult fleece trait which is expressed annually from a minimum of age 2 years, with a unity genetic correlation between annual records. Initial selection of parents can occur prior to the availability of their adult records, based on records collected at 6,12 or 18 months which have a maximum genetic correlation of 0·8 with the adult trait. Monte Carlo simulation is used to compare breeding programme alternatives, which allows overlapping generations and selection across age groups. MOET schemes select parents of minimum age 19 months in adult (A) schemes, 7 months in juvenile (J) schemes or 13 months when there are two breeding seasons per year. Natural mating results in an average of 1·09 lambs per ewe mated, whilst embryo transfer success rates follow realistic distributions (only 70% of donors giving live lambs) with an average of 3·45 lambs per donor flushed. Estimated breeding values, on which selection is based, are calculated by multiple trait selection index of an individual's own records or by multivariate best linear unbiased prediction (BLUP). In J schemes 7-month-old lambs may be selected on the basis of a pedigree index if no individual record has been collected at 6 months of age. Breeding scheme alternatives are compared on the basis of genetic response, rate of inbreeding and economic ‘income’ which is calculated as genetic response reduced for inbreeding depression, cumulated and discounted over a medium time horizon (14 years). The quantitative benefits of the collection of annual records are dependent on the choice of genetic and phenotypic parameters between traits, but these were chosen from the literature to be representative of annual fleece weight records in Merino sheep. Genetic response to selection and ‘income’ associated with the adult trait is increased by at least 15% in N, A and J schemes compared with when only a single record is collected at 18 months. The advantage of MOET over N schemes and the optimum number of sires to use is unchanged, but the average age of sires is increased. Both annual rate of inbreeding and average age of dams are unchanged. In N schemes, 80% of the benefits of collecting annual records on both sexes can be achieved by collecting records on males alone, whilst in MOET schemes measurement of dams contributes up to 50% of the benefit of the measurement of both sexes. Since MOET schemes have fewer dams to be recorded, the cost-effectiveness of collecting records on females is much higher than in natural mating schemes. Breeding programmes operating two mating seasons each year are found to be a viable alternative to J schemes: although genetic response is less, rate of inbreeding is also less, so that ‘income’ response is approximately equal.

The effects of inbreeding on the gross financial returns from ezues have been studied in a flock of sheep comprising three hill breeds and Ihe crosses among them. Each of these groups mis represented In non-inbred sheep (purebred and crossbred, designated O2, and F2, respectively) and four classes of inbreeding (inbreeding coefficients (F): 0·25, 0·375, 0·50 and 0·59). A second non-inbred class ums represented by crosses of inbred lines.

Gross lifetime income per eive derived from the sale of lamps for slaughter, cast ezves and wool was reduced by inbreeding in an almost linear fashion (between the O2/F2 and the F 0·50 class) at a rate of £1·27 (at 1992 prices) for ench 0·01 (percentage point) increase in the level of inbreeding — an overall reduction to 0·35 of the O2/F2 level. Line crosses gave a better return than the other non-inbred sheep. When expressed in terms of ewe metabolic body weight (M0·73) the decline with inbreeding was less —an overall reduction to 0·4 of the O2/F, income. Inbreeding of Ihe lamb caused a further decline in output independent of the inbreeding of the ezve.

A study was carried out to determine if sheep could be used as a model for predicting the intakes of silages by cattle. Fifty ewes and 45 cows were offered five grass silages unsupplemented in two parallel randomized-block experiments over a 4-week period. Each silage had been made using the same equipment over a range of cutting dates (May, July, August 1991). The silages differed in fermentation characteristics, with lactic acid concentrations varying from 15·3 to 90·7 g/kg dry matter, and ammonia-nitrogen concentrations from 77 to 198 g/kg nitrogen. When expressed on a metabolic live-weight basis, the mean intakes of silages by cows were 2·25 times those of the sheep, with the mean intakes being 35·1 and 79·1 g/kg M0·75 for the sheep and cows respectively. However, there was no significant correlation between recorded intakes during the 4th week of the study by the two species (y (cows) = 0·62× (sheep) + 58·40; r2 = 0·03). It was concluded that sheep were not a useful model system to predict the intakes of a range of grass silages of similar chop length by dairy cows.

Responses to four generations of divergent selection for lean groivth rate with ad-libitum feeding (LGA), for lean food conversion (LFC) and for daily food intake (DFI) in Large White pigs were studied. The LGA (LFC) selection criterion was designed to obtain equal correlated responses in growth rate (food conversion ratio) and carcass lean content, measured in phenotypic s.d. The selection criteria had phenotypic s.d. of 27, 29 and 253 units, respectively, and results are presented in s.d. units. There was a total of 3537 pigs, with an average of 40 boars and 40 gilts performance tested in each of the high, low and control lines per generation and the lines consisted of 10 sires and 20 dams. The generation interval was equal to 13·5 months. Animals were performance tested in individual pens with mean starting and finishing weights of 30 kg and 85 kg respectively.

Cumulative selection differentials in the three selection groups were 5·8, 3·6 and 3·3 phenotypic s.d. for LGA, LFC and DFI respectively. Direct responses to divergent selection were 1·7, 1·3 and 1·2 (s.e. 0·17) for LGA, LFC and DFI. The correlated response in LFC (1·6 (s.e. 0·18)) with selection on LGA was greater than the direct response in LFC. Conversely, the direct response in LGA was greater than the correlated response (1·1 (s.e. 0·18)) with selection on LFC. The response in LFC (–1·1 (s.e. 0·17)) with selection on DFI was similar in size but opposite in sign to the direct response in LFC. Responses were asymmetric about the control, as the high LGA and LFC responses were proportionately smaller (0·74 and 0·58) than low line responses. In contrast, the difference between the high DFI and control was four times greater than the difference between low line and control.

Heritabilities of LGA, LFC and DFI were 0·38, 0·35 and 0·29 (s.e. 0·03), when estimated by residual maximum likelihood, with common environmental effects of 0·09 (s.e. 0·02). Genetic correlations for LGA with LFC and DFI were positive, 0·76 (s.e. 0·03) and 0·23 (s.e. 0·07), but the genetic correlation between DFI and LFC was negative, –0·45 (s.e. 0·06). The experiment demonstrated that substantial responses to selection can be achieved in LGA, LFC and DFI. Selection on LGA resulted in larger direct and correlated responses than selection on LFC.

Carcass composition was measured on 133 Blackface ram lambs from a flock divergently selected for predicted carcass lean proportion. Prediction equations for different carcass components were developed using combinations of live weight and ultrasonic backfat and muscle depth. Both carcass lean and carcass fat proportion were best predicted using only live weight and fat depth, and a genetic transformation of the equation predicting carcass lean proportion was highly correlated (genetic correlation = 0·97) with the index on which the sheep were selected. Weights of carcass tissues were more accurately predicted than proportions. Lean weight was best predicted using live weight and muscle depth, and the weights of different fat components were best estimated using live weight, muscle depth and fat depth.

The equations predicting carcass lean proportion, carcass fat proportion, lean mass and fat mass had heritabilities of 0·29, 0·27 0·20 and 0·23, respectively. Heritabilities for carcass lean and carcass fat proportions, and the subcutaneous and intermuscular fat components were 0·43, 0·48, 0·24 and 0·49, respectively. Genetic correlations of the equation predicting carcass lean proportion with lean and fat proportions were 0·52 (s.e. 0·21) and –0·45 (s.e. 0·22), respectively. The same correlations for the equation predicting carcass fat proportion were –0·47 (s.e. 0·22) and 0·57 (s.e. 0·21). The equations predicting carcass lean and fat proportions were strongly correlated with subcutaneous fat proportion but weakly genetically correlated with intermuscular fat proportion.

The aim of this work was to investigatethe effect of stockingrate (0·6 m2 per pig or ≥1·2 m2 per pig) and faeces plus urine deposition on skatole concentration (boar taint)in subcutaneousfat. The study utilized pens with concrete flooring and different environmental temperatures.

The experimental material consisted of 144 cross bred male and female pigs in each of two experiments. One experiment was carried out in the summer (temperature ≥ 22°C) and one in the winter (temperature approx. 17°C in the pig house). In both experiments, stocking rate and faeces deposition were kept constant up to first delivery to the abattoir. Then stocking rate and faeces deposition were changed to the opposite treatment a week before the last three of the four deliveries to the abattoir.

The experiment confirmed the hypothesis that the pigs which lay in their faeces plusurine in pens with the high stocking rate for at least a week had a higher skatole level in subcutaneous fat than pigs kept clean in pens with the low stocking rate. It was feasible within a week before slaughter to increase or lower the skatole level by changing the treatment of the pigs.

Male and female pigs showed similar differences. Air temperature (season) had a significant influence on skatole level. The level was significantly higher in subcutaneous fat at hight emperatures in the summer experiment compared to the winter experiment despite the fact that the pigs with high stocking rate in the winte rexperiment were as heavily fouled with faeces and urine as those in the summer experiment.

Skatole concentrations in faeces of individual pigs were not influenced by different temperatures or treatments, and there was no significant correlation between the skatole level in faeces and in subcutaneous fat.

High skatole levels in subcutaneous fat seemed to have a genetic component.

It is hypothesized that skatole from faeces plusurine (excreta) especially at temperatures of 30° Cor more may pass through the skin of the pig and/or that skatole in a gaseous form may be absorbed through the lungs.

An experiment was made to examine the effects of food intake and an orally active progestagen on plasma steroid concentrations and subsequent reproductive performance of multiparous sows. Thirty-five multiparous Landrace × Large White sows were assigned to one of four treatments immediately after parturition. Treatment H-AT sows were offered a high level of food intake throughout lactation: 3·5 kg of a diet containing 160 g/kg dry matter (DM) of crude protein and 13 MJ digestible energy per kg DM given twice daily. Treatment H+AT sows were offered the same food intake as H-AT sows and in addition they were given 20 mglday of allyl trenbolone (AT) mixed with the morning feed. Treatment L-AT sows were given 1·5 kg of the same diet offered twice daily during lactation and treatment L+AT sows were also offered this lower level of food intake and given AT. The respective levels of food intake were offered to sows from the 1st day of lactation onwards until weaning at 21 days post partum. Litter weights at weaning were influenced significantly by feeding level (P < 0·001) and also by AT administration (P <0·05). Sows on the high level of feeding had the heaviest litters and food-restricted sows had the lightest litters. AT treatment depressed litter weight at weaning. There was no effect of food level or AT on the plasma concentrations of progesterone or oestrogens during lactation. There was a highly significant (P <0·001) difference in mean plasma oestrogen concentration between high (H-AT and H+AT) and low (L-AT and L+AT) fed groups during early pregnancy in the subsequent cycle. Sows given a combination of high energy in lactation and AT (H+AT) exhibited shorter intervals from weaning to oestrus than both groups of food-restricted (L-AT and L+AT) sows fP <0·01) but treatment had no significant effect on either the farrowing rate or on the subsequent litter size. It is concluded that despite significant changes in the live weight and condition of sows and changes in steroid hormone concentrations due to food intake and the administration of allyl trenbolone, there were no significant effects on reproductive performance.

Responses to divergent selection for lean growth rate with ad-libitum feeding (LGA), for lean food conversion (LFC) and for daily food intake (DFI) in Landrace pigs were studied. Selection was practised for four generations with a generation interval ofl year. A total of 2642 pigs were performance tested in the high, low and control lines, with an average of 37 boars and 39 gilts performance tested per selection line in each generation. The average within-line inbreeding coefficient at generation four was equal to 0·04. There was one control line for the DFI and LFC selection groups and another control line for the LGA selection group. Animals were performance tested in individual pens with mean starting and finishing weights of 30 kg and 85 kg respectively with ad-libitum feeding. The selection criteria had phenotypic s.d. of 32, 29 and 274 units, for LGA, LFC and DFI, respectively, and results are presented in phenotypic s.d.

Cumulative selection differentials (CSD) were 5·1, 4·5 and 5·5 phenotypic s.d. for LGA, LFC and DFI, respectively. Direct responses to selection were 1·4,1·1 and 0·9 (s.e. 0·20) for LGA, LFC and DFI. In each of the three selection groups, the CSD and direct responses to selection were symmetric about the control lines. The correlated response in LFC (1·1, s.e. 0·19) with selection on LGA was equal to the direct response in LFC. In contrast, the direct response in LGA was greater than the correlated response (0·7, s.e. 0·18) with selection on LFC. There was a negative correlated response in DFI (-0·6, s.e. 0·18) with selection on LFC, but the response with selection on LGA was not significant (0·2, s.e. 0·16).

Heritabilities for LGA, LFC and DFI ivere 0·25, 0·25 and 0·18 (s.e. 0·03), when estimated by residual maximum likelihood, with common environmental effects of 0·12 (s.e. 0·02). Genetic correlations for LFC with LGA and DFI were respectively positive (0·87, s.e. 0·02) and negative (-0·36, s.e. 0·09), while the genetic correlation between DFI and LGA was not statistically different from zero, 0·13 (s.e. 0·10). Selection on components of efficient lean growth has identified LGA as an effective selection objective for improving both LGA and LFC, without a reduction in DFI.

An investigation was conducted to determine the effect of mating at puberty or third oestrus or after pre-mating oestrogen and progestagen therapy, designed to mimic the changes that occur in plasma oestrogen and progesterone concentrations during the 1st, 2nd or 3rd oestrous cycles, on fertility in ewe lambs. Twenty-five ewe lambs were mated at puberty on mean date of 18 October and 21 were mated at third oestrus on 3 December. Of these, 28% and 81% of animals mated at puberty and third oestrus lambed, respectively (P < 0·05). A similar number of ewe lambs were mated within 24 h of ram introduction in the groups given one, two or three consecutive applications oestrogen and progestagen (25 μg oestradiol benzoate followed 48 h later by a progestagen pessary left in situ for days). Of the 43 corpora lutea, 21 were represented by viable embryos in the 37 animals holding to service after one application compared with 34 out of 46 corpora lutea in 40 animals given three applications (P < 0·05). Furthermore, embryo crown-rump length, embryo weight and amniotic sac width was greater (P < 0·05), on day of pregnancy, in animals given one compared with three applications of oestrogen and progestagen. These results demonstrate that fertility is improved in ewe lambs mated on third oestrus or after progestagen and oestrogen therapy designed to simulate three oestrous cycles.

The effects of altering the glucogenic potential of the diet of wether lambs (initial body weight 56·4 kg) offered an energy allowance of 0·25 that required for maintenance on body-weight loss, nitrogen retention and blood and rumen fluid parameters were examined. Whole barley was offered alone or substituted on an isoenergetic basis with 40 or 120 g of a blend of glucose precursors (polyhydric alcohols, propylene glycol, sodium propionate) absorbed on to palm kernel meal (acetona meal), to 36 lambs (no. = 12 per treatment) for 8 weeks. Substituting barley with 40 g acetona meal had a small but non-significant effect in decreasing body-weight loss and plasma aspartate amino transferase activity and in increasing plasma insulin concentration. Increasing the glucogenic potential of the diet by substituting barley with 120 g acetona meal decreased body-weight loss in the early phase of the study, decreased plasma concentrations of non-esterified fatty acids and increased the proportion of propionate in volatile fatty acids in rumen fluid, plasma aspartate amino transferase and gamma-glutamyl transferase activities and plasma insulin concentrations. Increasing the glucogenic potential of the diet did not affect nitrogen retention or plasma glucose concentration. It is concluded that increasing the glucogenic potential of the diet of lambs resulted in an increase in metabolic efficiency at the onset of negative energy balance which was mediated through a change in plasma insulin concentration. The absence of a significant difference in body-weight loss between the diets in the latter part of the study reflects the lack of sustained ketosis.

Responses to four generations of divergent selection in pigs for lean growth rate (LGS) with restricted feeding were studied. The selection criterion was designed to obtain equal correlated responses in growth rate and carcass lean content, measured in phenotypic s.d. Animals were to be performance tested in individual pens with a mean starting weight of 30 kg for a period of 84 days. Daily food intake was equal to 0·75 gig of the daily food intake for pigs offered food ad-libitum. In the high, low and control lines, there ivere 1250 Large White-Edinburgh (LW) pigs and 875 British Landrace-Wye (LR) pigs. Each selection line consisted of 10 sires and 20 dams, with a generation interval ofl year.

After four generations of selection, cumulative selection differentials were 5·9 and 4·8 phenotypic s.d. for LW and LR populations, with similar responses, 1·8 (s.e. 0·17) phenotypic s.d. Mean weight at the end of test, growth rate and backfat depths at the shoulder, mid back and loin were 89 kg, 712 g/day, 26,13 and 13 mm for LW and for LR pigs were 87 kg, 683 g/day, 28,10 and 10 mm. High line pigs were heavier at the end of test (4·3 (s.e.d. 1·4) kg and 4·0 (s.e.d. 1·6) kg) for LW and LR populations, with corresponding responses in growth rate (54 (s.e.d. 16) g/day and 47 (s.e.d. 18) g/day). Responses in the three backfat depths were −4·1 (s.e.d. 1·2) mm, −2·6 (s.e.d. 0·7) mm and −2·9 (s.e.d. 0·7) mm for LW and −2·2 (s.e.d. 0·05) mm, −2·2 (s.e.d. 0·4) mm and −2·4 (s.e.d. 0·5) mm for LR populations. Responses in weight off test and backfat depths were symmetric about the control lines.

Heritabilities for LGS were 0·34 and 0·28 (s.e.d. 0·5) for the LW and LR populations, when estimated by residual maximum likelihood. Common environmental effects for LGS were 0·11 (s.e. 0·03) for LW and 0·17 (s.e. 0·04) for LR. Heritabilities for growth rate and average backfat depth were similar for LW and LR populations (0·17 and 0·29, s.e. 0·05), as were common environmental effects (0·10 s.e. 0·04). Average phenotypic and genetic correlations between growth rate and backfat, for LW and LR populations, were small (0·15 (s.e. 0·03) and −0·06 (s.e. 0·16), respectively).

Responses to selection and genetic parameter estimates demonstrate that there is substantial genetic variation in growth and fat deposition when pigs are performance tested on restricted feeding.

Measurements were carried out on 200 selected carcasses in order to investigate whether single or multiple site measurements could provide an accurate estimation of the lean meat proportion in the carcass and the major cuts i.e. ham, shoulder, loin and belly. Fat measurements were taken with the Hennessy Grading Probe II at 17 sites on the left side of the carcass, and muscle measurements at four of these sites. The day after selection 200 left sides were dissected according to the simplified European Community (EC) reference method.

Fifty of these left sides were further dissected according to the full EC reference method. The use of multiple site measurements, compared with a single site measurement, only slightly reduced the residual s.d. of the estimated lean meat proportion in the carcass and the major cuts, e.g. addition of a second site measurement reduced the residual s.d. values by 0·3 to 1·3 gjkg only. The measurement between the 13th and 14th thoracic vertebrae, 7 cm off the dorsal mid line, appeared to be the best single estimator, of the sites probed, for lean meat proportion in the carcass and in the major cuts. Prediction of lean meat proportion in the belly appeared to be problematic (residual s.d.> 29 g/kg).

During the experiment two problems emerged. Firstly, the algorithm, which calculates fat and muscle thickness from the optically measured tissue profile, was not appropriate for every site. Secondly, the definition of the probing position did not always coincide with the anatomical position of the intended superficial layer of muscle underneath the subcutaneous fat.

Two randomized-block experiments were carried out with growing pigs housed in groups of 10 from about 36 to 92 kg live weight and offered non-pelleted diets from single-space feeders containing separate food and water dispensers both operated by the pigs. The dispensing rate of food could be adjusted and rates of 1·4, 2·7 and 5·3 g per press of the nose plate were compared in experiment 1 and 4·8, 6·9 and 9·3 g per press compared in experiment 2. As dispensing rate increased in experiment 1 there were significant increases in food intake, growth rate and backfat thickness and a significant improvement in food conversion ratio on a carcass gain basis. There was no effect on the variation in growth rate between pigs within pens nor were there treatment effects in experiment 2 on any of the above measurements. The amount of food accumulating in troughs increased with feeder setting in both experiments.

Behaviour at feeders was observed in each pen for 24 h during the 5th week using time-lapse video recording. The number of visits to feeders was not significantly affected in either experiment but time spent feeding was significantly increased at the lowest setting in experiment 1. Both the number of withdrawals from feeders accompanied by aggressive encounters and the number of pigs queuing decreased up to the dispensing rate of 6·9 g per press. There were tendencies for all parameters of behaviour to continue changing up to the highest dispensing rate.