SUMMARYSeventy-nine Finnish Landrace x Dorset Horn ewes in lamb to Suffolk rams were used in a comparative slaughter experiment to assess the effects of numbers of foetuses on the growth of the products of conception. The mean number of foetuses was 2·7, with a range from 1 to 5. The ewes were individually penned and given a standard diet with metabolizable energy concentration 7·7 MJ/kg and N concentration 21 g/kg, rationed at 2 kg/day during the first month and 1·25 kg/day during the second and third. Allowances beyond that time were on two scales and depended on the number of foetuses being carried as diagnosed by radiography. The ewes were slaughtered between 50 and 145 days of gestation. The gravid uterus was dissected into foetal, placental, foetal fluid and empty uterus components. Equations were fitted to the weights of each component to describe the effects of stage of gestation, litter size and ewe weight. For the mean weight per foetus (Y, kg) the preferred equation isIn (Y)= 2·419–17·574e-0.01976t–00079ft+0–0046w,where tis the time in days from conception, / is the number of foetuses and wis the weight (kg) of the ewe at mating. This is a version of the Gompertz equation, with additional terms to express the effects of / and w.The weights of the placenta and of the empty uterus were similarly fitted by versions of the Gompertz equation but the mean weight of fluids per foetus (Z, kg) or, rather, its natural logarithm was best described by a third degree polynomial, which isIn(Z) = –11·518 + 0–326t;–0·00316t2+0·0000102t3.None of the weights was significantly affected by the level of feeding in late pregnancy.Estimates of mean weights from the equations and of daily rates of gain in weight from the first differentials of the equations are tabulated against stage of gestation and litter size, and the forms of the weight, growth rate and specific growth-rate curves are illustrated graphically.The equation for foetal weight estimates that at the end of pregnancy the mean weight per foetus is reduced by a factor of 0–89 for each additional foetus being carried. The mathematical model implies that the differences originate in early pregnancy, when the factor is very close to unity, and that the mean weights gradually diverge. In the absence of direct evidence this would appear to be the simplest hypothesis, rather than the assumption in most of the earlier literature that the effect is entirely confined to the last 4 or 5 weeks of gestation.Just before parturition the total daily weight gain of quadruplet foetuses was about 250 g and was associated with a similar gain in weight of foetal fluids, the ratio of fluid weight to foetal weight appearing to increase with litter size. The ewes were clearly under considerable physical stress. It is suggested that this aspect must be closely considered when greater prolificacy is sought.

1. Measurements were made of the losses of nitrogen in the faeces and urine of sixty pigs of approximately 33 kg, given dietary regimens comprising twenty-nine combinations of fish flour (0-800 g/d) and maize starch (0-1200 g/d).2. The results were used to develop a generalized equation describing N retention as a joint function of N intake and starch intake.3. The protein-sparing effect of starch was exerted in all circumstances. It was greatest when protein intake exceeded 220 g/d but some effect persisted with protein-free diets. With a high protein supply, the increase in N retention per unit increase in dietary starch decreased from 36 mg/g with the first increment to 3 mg/g with the highest attainable starch intake.4. From the generalized equation the relationships between net protein utilization and protein concentration and food intake could be described as continuous functions. The equations may be of use in comparing the protein values of diets measured under non-standardized conditions.One of the difficulties of estimating nutrient requirements arises when theefficiency with which a nutrient is utilized is not constant. This difficulty seems to arise to some extent with every nutrient. In some instances it may be overcome for practical purposes by delineating a range over which a constant efficiency is a reasonable approximation. In studies of energy metabolism, for example, the convention has commonly been adopted of treating efficiency for maintenance and for growth as two constants, although this implies a distinct discontinuity at maintenance, whereas the over-all relationship may be better described by a single exponential equation (Blaxter & Boyne, 1970) with a continuously diminishing efficiency.This difficulty is particularly evident in estimating protein requirements, and it has long been realized that the efficiency of protein utilization is not determined simply by the amino acid composition of the protein but by the amount and composition of the diet in which the protein is supplied to the animal.It has been recognized that there are, in the absence of vitamin and mineral deficiencies, three major dietary variables which jointly determine the efficiency of protein utilization : the quality of the protein, the amount of protein given and the amounts of lipid and carbohydrate, the non-protein energy, given with it.Although these are the primary unconfounded variables, the effects on protein utilization of the amounts of protein and non-protein energy have usually been examined in terms of the variables derived from them : protein concentration (protein energy as a proportion of total energy) and daily food intake. Measurements of biological value (BV), or of net protein utilization (NPU) are by convention, therefore, made at a constant protein concentration, commonly 0.1.In more practical studies, there is a need to estimate the protein value of diets fed unaltered; in such instances a comparison of the protein quality of two diets can be made only if the effects on protein utilization of pr...

I . Thirty-three diets were made by adding to ground barley combinations of L-lysine (0-6.0 g/kg) and L-threonine (0-3.0 g/kg), together with vitamins and minerals. Each was given to two female and two castrated male pigs during their growth from 25 to 60 kg.2. Growth rate increased from 0.36 kg/d to a maximum of 0.65 kg/d with additions of 3.8 g L-lysine/kg and 1.8 g L-threonine/kg; these values and the maximum gains achieved were slightly higher for castrates than for females. Minimum values for food conversion ratio of 2.84 for castrates and 3.06 for females were achieved with similar amino acid additions to those giving fastest growth, 3. Carcass fat, estimated by specific gravity, was least with the addition of 5.9 g L-lysine/kg, but continued to decrease up to the highest threonine concentrations. Backfat thickness was also reduced by amino acid additions but failed to reach a minimum.Cereal grains form the basis of diets for pigs in many parts of the world, and particularly in industrialized countries. Such diets are completed by the addition of foodstuffs supplying vitamins and minerals and additional protein. The purpose of the additional protein is twofold; to increase the total protein supply and to remedy the amino acid deficiencies of the cereal protein. Ideally, the supplementary protein should bring the total amount of protein and its amino acid balance to the point where the animal's needs are exactly met. The success of this is limited in two ways; first by limitations of the available foodstuffs and secondly by uncertainties about the amino acid requirements of pigs.We have determined the additions of amino acids required to optimize the utilization of barley protein by growing pigs in a series of experiments in which urinary nitrogen excretion was used as the criterion of protein utilization (Fuller eta!. 1978). It was found that addition of the first-and second-limiting amino acids, lysine and threonine, accounted for most of the attainable improvement. In these experiments, addition of 4.0 g L-lysine/kg andlof 1'2 g L-threonine/kg, raising the total concentrations to approximately 7.2 and 4.2 g/k g, minimized urinary N excretion.The present experiment was designed to examine two further aspects: first, whether the commercially-important measures of pig growth could be improved to the same extent as N metabolism by the same amino acid additions; second, to examine the form of the response to determine what improvements in pig performance could be obtained with smaller additions of amino acids. E X P E R I M E N T A L

SUMMARY1. In a comparative slaughter experiment, 12 female pigs (six at 80 kg and six at 100 kg) were allocated at first oestrus to each of five treatments: Treatment 1, initial slaughter, or Treatments 2, 3 and 4, mated and given 19·5, 25·8 or 32·1 MJ ME/day for the last 100 days of pregnancy, or Treatment 5, not mated (virgin) and given 25·8 MJ ME/day over a similar period. Pigs on Treatments 2, 3, 4 and 5 were given the same amount of protein and were killed about 123 days after first oestrus. Piglets were removed at birth.2. There was no evidence of any special effect of pregnancy in the stimulation of permanent maternal growth. The average live weight, ingesta-free body and carcass gains of the pair-fed, mated and virgin gilts (±SE of difference) were, respectively, 32·3 and 36·9 + 2·8, 27·5 and 27·6 ± 3·4, and 26·0 and 26·7 + 2·1 kg. There were no statistically significant differences between these two treatments in carcass composition, specific gravity or backfat thickness. The mated pigs had lighter livers (P < 0·01) and heavier reproductive tracts (P < 0·05), and lost about 9 kg within a week of parturition.3. Increasing energy intake increased piglet birth weights (P<005) but had no effect on the number of piglets born. The pigs that were initially heavier (100 kg v. 80 kg) had 1·5 more piglets, though this was not statistically significant (P<0·1).4. Although there was no special effect of pregnancy on permanent maternal growth, the conversion of food by the once-mated pig was very efficient if an allowance was made for the food cost of producing the piglets.

Plasma prolactin concentrations during the first 2 months after lambing, at oestrus, and during early pregnancy were investigated in 2 experiments in which Finn x Dorset Horn ewes were mated at an induced oestrus approximately 9 weeks after lambing. Mean prolactin concentrations between lambing and mating were dependent on seasons, being greater than 260 ng/ml plasma in lactating ewes mated in July and less than 150 ng/ml in those mated in October. Within 8 days of weaning of the lambs at 50 days post partum values declined to 122 and 30 ng/ml respectively. Plane of nutrition had little effect on prolactin levels. Higher prolactin values were recorded during oestrus in ewes mated in March or July, the normal period of anoestrus, than in December, the normal breeding season, mean values being approximately 200 ng/ml and 35 ng/ml respectively. The mean increases in the concentrations of prolactin during oestrus were smaller in lactating than non-lactating ewes. It is suggested that these differences in prolactin levels may be responsible for the effects of season and lactation on ewe fertility.