The effects of spineless cactus (Opuntia ficus indica var. inermis) supply on digestion of wheat straw was studied n rumen cannulated sheep. In addition to urea (10 g) and mineral and vitamin mixture (30 g), the animals — received wheat straw alone or supplemented with graded levels of spineless cactus (150, 300, 450 or 600 g dry natter (DM) per head per day). The diets were studied in five successive 25-day periods from February to June 992 in Tunisia. "Water consumption, food intake, digestibility, diurnal variations of ruminal fermentation (pH, mmonia nitrogen, volatile fatty acids, protozoal concentrations in rumen fluid) and cellulolytic activity in the umen were measured.

Spineless cactus had high contents of ash (260 g/kg DM) and water (926 g/kg fresh weight) and low neutral-detergent fibre content (185 g NDF per kg DM). Crude protein (CP) content of spineless cactus was 2-5 times reater than that of wheat straw (58 v. 23 g/kg DM, respectively). Drinking water consumption was substantially educed (P < 0·001) as the level of spineless cactus increased. When spineless cactus supply exceeded 300 g DM peray, sheep consumed practically no drinking water. The voluntary intake of straw increased significantly with vineless cactus level. Total diet apparent digesibilities of DM, organic matter and CP tended to increase with mneless cactus supply. Such increase was significant only with the 450 and 600 g levels of cactus. NDF and acidetergent-fibre apparent digestibilities were not affected by spineless cactus supply. Addition of spineless cactus up i 300 g DM significantly increased ruminal ammonia nitrogen (P< 0·001) but there was no additional effect with irther supplementation. When sheep were supplemented with spineless cactus, total rumen volatile fatty acid icreased and acetate: propionate ratio decreased significantly. Moreover, spineless cactus supply increased total wtozoa number and reduced significantly rumen cellulolytic activity measured as DM and NDF disappearance of heat straw from incubated nylon bags.

A total of 897 pigs were used in a study to investigate the relative effects of terminal sire genotype (lines Av.Bv. C), sex (castrate v. gilt), slaughter weight (80 v. 100 v. 220 kg), feeding regimen (ad libitum v. restricted, 0·82 ad libitum intake) and slaughter-house (HI v. H2 v. H3) on growth performance, carcass and meat quality characteristics and the eating quality offresh pig meat. Sire line A was a pure Duroc population, and B and C were European-type experimental lines where C contained Pietrain and B did not. In total, 26 sires from line A, 42 sires from line B, and 21 sires from line C were mated to females from the same crossbred dam line and progeny were reared under standard conditions to slaughter. Following slaughter and carcass evaluation, samples of longissimus dorsi were investigated for a range of meat quality and organoleptic characteristics. Line A produced fatter carcasses (C fat depths = 15·6 v. 24·0 v. 14·0 mm for lines A, B, and C, respectively, average s.e. 0·39) with higher killing-out proportions (g/kg) (790 v. 779 v. 786 respectively, average s.e. 1·4) and higher visible marbling, less tissue separation, firmer backfat, and juicier (3·81 v. 3·67 v. 3·72 respectively, average s.e. 0·044: on a scale 1 (extremely dry) to 8 (extremely juicy)) and more acceptable meat (4·54 v. 4·37 v. 4·42 respectively average s.e. 0·037: on a scale 1 (dislike extremely) to 8 (like extremely)) with a lower shear force (5·35 v. 5·78 v. 5·67 kg respectively, average s.e. 0·078) than lines B and C which were similar in most respects. Increases in slaughter weight were associated with a reduction in growth rate (785 v. 769 v. 725 glday for 80, 100 and 120 kg slaughter weight respectively, average s.e. 8·5), increases in backfat (Cfat = 13·3 v. 24·2 v. 26·3 mm respectively, average s.e. 0·34) and longissimus muscle area (34·6 v. 40·7 v. 44·6 cm2 respectively, average s.e. 0·59) and a deterioration in tenderness (4·72 v. 4·40 v. 3·95 respectively, average s.e. 0·062: on a scale 1 (extremely tough) to 8 (extremely tender) and overall acceptability (4·65 v. 4·44 v. 4·25 respectively, average s.e. 0·045) and an increase in shear force (5·37 v. 5·58 v. 5·87 kg respectively, average s.e. 0·085). Slaughter-house had a significant impact on pork odour scores but not on other organoleptic properties. Pigs reared under ad libitum feeding grew faster (840 v. 678 g/day respectively, average s.e. 3·7), were fatter (Cfat = 15·8 v. 23·2 mm respectively, s.e. 0·28), had lower carcass yields (780 v. 790 g/kg respectively, average s.e. 1) and produced more tender, juicier meat than those reared under restricted feeding. Differences between castrated males and gilts in growth and carcass trait were in line with other studies and there were no significant differences between the sexes for eating quality. There were relatively few significant interactions (P < 0·05) for eating quality traits and most of these involved slaughter-house and were for pork odour intensity, which are of limited practical significance. This suggests that the effects of sire genotype, slaughter weight and feeding regimen on eating quality identified in this study are likely to be additive.

The eyes of eighteen female sheep (Ovis aries) were refracted and details of inter ocular distance, pupil size, shape and fundus presence recorded. The sheep eyes generally possessed very low hyperopia with little astigmatism, such physiological optics being expected to produce a well focused retinal image for objects in the middle and long distance. No evidence was found for accommodation, which would have produced a well focused ocular image for near objects. A further 10 sheep had their monocular and binocular visual fields measured. The estimated visual field suggests the existence of at least binocular vision, if not the presence of stereopsis. Given the lack of accommodation and a wide inter-ocular distance, it is likely that some degree of stereopsis exists in the animal's middle and long distance vision, but is absent in near vision. These findings support those taken from the animal's neurophysiology and observations of its visually guided behaviour.

Genetic and phenotypic relationships between performance test and reproduction traits were estimated, after five generations of divergent selection for components of efficient lean growth, in a population of Large Wltite pigs. On ad-libitum feeding, a total of 4334 pigs were performance tested, of which 884 selected gilts had measurements of reproduction traits. On a restricted feeding regime, 1558 pigs were tested, which included 336 selected gilts with reproduction records. For pigs given food ad libitum, genetic correlations between litter weights at birth and weaning with daily food intake (0·48 and 0·42, s.e. 0·16) and with growth rate on test (0·65 and 0·52) were positive, but correlations with backfat depths were not significantly different from zero. For pigs given food at a restricted level, litter birth weight was positively genetically correlated with growth rate (0·50, s.e. 0·18) and negatively correlated with backfat depths (-0·48, s.e. 0·16). Phenotypic and environmental correlations between performance test and reproduction traits were all less than 0·10 in magnitude, for pigs tested on either feeding regime. The variation in backfat depth enabled detection of a non-linear relationship between predicted breeding values for litter weight at birth with predicted breeding values for average backfat depth of farrowing gilts performance tested on ad-libitum feeding, but not for gilts tested on restricted feeding. The positive genetic correlations between growth rate and daily food intake with litter traits suggested that selection strategies which change growth and daily food intake may result in relatively greater genetic changes in piglet growth rate than in litter size.

Confinement has been shown to affect bone strenth in poultry but this weakness has not been documented in other species housed in confinement. The objectives of this experiment were to compare muscle weight and bone strength in non-pregnant sows, of similar age and parity, housed throughout eight or nine pregnancies in two different dry sow systems: (1) individually in stalls and (2) communally in a large group. Following slaughter, the left thoracic and pelvic limbs were dissected and 14 locomotor muscles removed and c. ???lied. A proportional muscle weight was then calculated by dividing individual muscle weight (g) by total body weight (kg). Where there were significant differences, stall-housed sows had lower absolute and proportional muscle weights than group-housed sows. The left humerus and femur were also removed. The bones were broken by a three-point bend test using an Instron Universal Tester. Both bones from stall-housed sows had breaking strengths that were about two-thirds those of group-housed sows. The results indicate that confinement of sows, with a consequent lack of exercise, results in reduction of muscle weight and considerable reduction of bone strength.

Four litters produced by father-daughter matings (back crosses) resulting in 35 animals with a theoretical inbreeding coefficient of 25% were typed with 21 independent informative markers. The differences between the two founder animals were estimated, based on the marker information, and it was found that the founder boar had higher genetic potential for proportion of lean meat and lower genetic potential for groivth than the founder sow. The proportion of the genome of each offspring which was identical by descent was investigated. On the basis of these markers the realized inbreeding was found to vary between 7 and 47%. The linear decrease in weight at days 1, 26 and 136, average daily gain and proportion of lean meat regressed on the realized inbreeding were estimated to 0·6 kg, 2·4 kg, 18 kg, 95 g/day and 15 g/kg, respectively. For weight at day 88 a corresponding linear increase of 11 kg was observed. The joint effect of founder differences and realized inbreeding were as expected negative and statistically significant for all growth traits.

Two hundred and ten piglets weaned at 30 days of age and weighing 9 kg were allocated to a 3 × 2 factorial arrangement of treatments to examine the influence of feeder type ('wet and dry' single-space feeder (SSF), 'dry' SSF, or multi-space feeder) and method of group allocation (piglets from different litters were mixed together, or piglets were weaned with their littermates) on performance after weaning. The use of a 'wet and dry' SSF did not enhance growth rate or voluntary food intake (P > 0-05) in the 28 days after weaning. Piglets feeding from 'wet and dry' SSF grew proportionately 0-06 slower in the 28 days following weaning as a result of a 0-45 proportional decrease (P = 0.013) in growth in the 1st week. Piglets mixed together from different litters at weaning grew proportionately 0.34 faster (P = 0.010) in the first 14 days after weaning than piglets weaned as littermates. This initial difference resulted in a 0-14 proportional increase in growth rate from day 1 to 28 (385 v. 339 glday, P < 0.001). Piglets from different litters mixed at weaning consumed proportionately from 0.06 to 0.16 more food (P = 0.096) than piglets weaned as entire litters. No interactions between feeder design and the method of group allocation at weaning occurred for any of the production parameters measured. Single-space feeders failed to increase the production of weaner piglets, whilst mixing non-littermate piglets into a new social group seemingly enhanced performance from weaning to 56 days of age.

Ultrasound imaging is widely used in animal breeding to provide in vivo estimates of the carcass composition of candidates for selection. Although the technique is less accurate than more recent medical imaging methods, such as X-ray computed tomography and magnetic resonance imaging, it is relatively cheap and mobile. Therefore large numbers of animals can be measured. Most current ultrasound scanners require some degree of manual interpretation of images, which is time consuming and liable to vary both between and within individuals. Hence, this study investigated the automatic interpretation of ultrasound scans from sheep. A computer algorithm is proposed for identifying tissue boundaries. Estimates of tissue depth are shown to be comparable with those obtained by manual interpretation, for images of 72 sheep scanned twice at the position of the 13th thoracic vertebra. The root-mean-square errors of subcutaneous fat depth and m. longissimus muscle depth ivere 0·7 mm and 1·7 mm, respectively.

The aim of this experiment was to reduce piglet stress and improve performance following moving and mixing at weaning by use of the anti-aggressive drug amperozide and the practice of co-mingling piglets during lactation. Thirty-six piglets weaned at 29 days of age were allocated to a 2 × 2 factorial arrangement of treatments as follows: (1) injection of 1 ml amperozide or placebo (0.9% saline) at the time of weaning, and (2) co-mingling or no comingling of piglets from day 10 of lactation to weaning. At weaning all piglets were injected with pliytohaemagglutinin or saline, their change in skin-fold thickness was measured 24 h later and behaviour was recorded every 5 min for the first 120 min after weaning. At 4 and 24 h after weaning the depth and extent of injuries were recorded. Daily live-weight gain, apparent voluntary food intake and food conversion ratio were not improved (P > 0.05) by administration of amperozide or co-mingling of piglets during lactation. Cell-mediated immunity, as evidenced by a 0.17 proportional decrease (P <; 0.001) in skin-fold thickness following the injection of phytohaemagglutinin, was decreased in piglets not given amperozide at weaning. Injection of amperozide caused a 0.26 proportional increase (P < 0.001) in the amount of time piglets spent lying and caused piglets to display no interest in the feeder (P < 0.001) in the first 2 h after weaning. Consequently piglets not given amperozide were more active, spent more time at the feeder and participated in more aggressive encounters (P < 0.01). A significant interaction occurred between co-mingling and amperozide for mean injury score at both 4 (P < 0.01) and 24 h (P < 0.05) after weaning. When measured 4 h after weaning, co-mingling of piglets during suckling reduced mean injury score regardless of whether piglets were given amperozide or not at weaning. When piglets were not allowed to co-mingle during lactation, amperozide reduced mean injury score proportionately by around 0-44 (P < 0.01) compared with piglets administered with saline. Despite the administration of amperozide and the practice of comingling causing a reduction in stress and an improvement in welfare in this experiment, there was no beneficial effect on performance in the first 2 weeks after weaning.

The development and moulting of the birth and first adult coat of two genotypes of cashmere-bearing goats, a highyielding Siberian (S) and low-yielding Icelandic × Scottish feral (IF), were examined. Samples of fibre from the neck, mid side, belly and rear-leg areas and of mid-side skin were obtained monthly from five male and five female kids of each genotype from birth to 13 months of age. Guard hair and undercoat length, secondary/primary (S/P) skin follicle ratio and the proportion of skin follicles containing a non-growing fibre with a brush end were recorded. S/P ratio reached an asymptote between 4 and 5 months of age and was higher fP < 0·01) in S (10·7) than in IF (6·7) kids. In both genotypes peaks of brush formation in primary follicles occurred between 2 and 4 months and again between 8 and 11 months of age. In contrast to primary follicles, secondary follicles had only a single peak of brush formation occurring at 10 and 8 months of age for S and IF kids respectively. Guard hair moulted from the birthcoat, but there was no loss of undercoat prior to 12 months of age and harvesting of cashmere prior to this time is unnecessary.

The secondary follicles of S kids had a longer growing period and fibre length increased at a greater rate than that of IF kids, resulting in more than a two-fold difference in the maximum undercoat length. Assessments to estimate the final weight and proportion of cashmere, made prior to completion of growth of the fleece, are likely to underestimate these parameters for high-yielding goats in comparison with their low-yielding counterparts. However, it is unlikely to alter the ranking, and hence the selection, of animals for retention in the breeding herd.

Two experiments investigated the welfare of pigs during transport. In experiment 1, 12 groups offour 90-kg pigs were transported to slaughter in a commercial livestock lorry for 1·5 h. Half the animals were transported in their social groups (unmixed condition) and half were transported with groups of previously unfamiliar pigs mixed together (mixed condition). Behaviour was recorded, a general activity index scored and saliva samples taken at different stages of the journey for analysis ofcortisol. Pigs spent most of their time standing in both conditions. The journey was very rough (as revealed by characterization with an accelerometer) and in the unmixed condition the pigs appeared to stand to reduce travel sickness. In contrast, in the mixed condition, this preference for standing seemed to be due to fighting which stressed and exhausted the animals (the general activity index was three times the unmixed condition). Levels of salivary cortisol were higher in the mixed condition at the beginning, middle and end of the journey. In experiment 2, six 35-kg pigs, prepared in advance with jugular vein catheters, were loaded onto a commercial livestock lorry (09.30 h) where they were individually penned. The vehicle remained stationary with the engine off and blood samples were taken at 30-min intervals during the next 8 h (control). Two days later this procedure was repeated while the vehicle was driven for 8 h (on main roads and motorways). Plasma concentrations of cortisol and beta-endorphin increased markedly in both conditions immediately after loading. Cortisol levels were greater (relative to control) at the beginning, in the middle and at the end of the journey. Concentrations of beta-endorphin did not differ between control and experimental conditions except during the final 180 min of the journey when the control levels were higher.

The work described was carried out to study the expression of insulin-like growth factor 1 (IGF-1) and growth hormone-receptor (GHR) mRNA in liver, skeletal muscle and adipose tissue from three breeds of pig with varying growth characteristics. The three breeds studied were the Large Wlrite, noted for its lean tissue; the Duroc, characterized by its high intramuscular fat content; and the Meishan × Landrace (0·5 Meishan), noted for its fat, poorly conformed carcass and slower growth rate. The probes used were designed to monitor promoter usage for IGF-1 expression and also expression of the extra-cellular domain of the GHR. Eighteen gilts, six of each breed, were given a barley/wheat diet (158 g crude protein, 10·7 g lysine and 13·9 MJ energy per kg), to appetite, for 1 to 2 weeks until they reached about 85 kg. Samples of liver, longissimu s dorsi (LD) muscle and three adipose tissue depots (subcutaneous (SC), perirenal (PR) and omental (OM)) were collected immediately after slaughter and frozen in liquid nitrogen (total time of sample collection to plunging of sample into liquid nitrogen was <3 min), prior to extraction of total RNA and ribonuclease protection assays. Individual serum samples collected at exsanguination were frozen prior to IGF-1 radioimmunoassay. There were no breed differences in the serum IGF-1 concentrations (range 49 to 134 μg/l), or in expression of the GHR gene or either class of IGF-1 transcript in the liver. However, there was a significant difference between the breeds in expression of IGF-1 mRNA in the LD muscle (P < 0·001), the order being Duroc > Wliite > Meishan, with only class 1 transcripts of IGF-1 being found. GHR expression in LD muscle was lower in White than in the other tivo breeds (P = 0·022). There was a significant difference between the breeds in expression of IGF-1 mRNA (only class 1 transcripts present) in the adipose tissue (P = 0-006), the order being Wliite · Duroc · Meishan, and also a significant depot difference, with expression being highest in the SC depot (P < 0·001). There were no differences between the breeds or depots in expression of GHR mRNA in adipose tissue. The observed differences in muscle and adipose tissue IGF-1 expression may relate to the overall growth of the animal.

Thirty-two piglets weaned at 28 days of age were used to test the hypothesis that maintenance of nutrition after weaning would prevent the normal decline in villous height and increase in crypt depth and hence preserve the structure and function of the small intestine. Piglets were allocated to one of four treatments at weaning: (1) control group killed at weaning; (2) piglets offered a dry starter diet ad libitum; (3) piglets offered ewes' fresh milk; and (4) piglets offered ewes' fresh milk plus 20 g t-glutamine per I. Piglets in treatments (3) and (4) were offered ewes' fresh milk every 2 h in a feeding schedule that increased from 1·2 I per piglet on the 1st day after weaning to 2·4 I on days 4 and 5. On the 5th day all piglets were killed and samples of small intestine were taken for histological and biochemical examination. Feeding ewes' milk or ewes' milk plus 20 g L-glutamine per I maintained (P > 0·05) villous height and crypt depth compared with piglets killed at weaning. In contrast, piglets given a dry starter diet had shorter villi (P < 0·001), deeper crypts (P < 0·001), and proportionately 0·21 to 0·28 less protein (P > 0·05) in their intestinal mucosa. Piglets given the starter diet proportionately grew from 0·49 to 0·62 more slowly (P < 0·01), ate the same amount of dry matter (DM; P > 0·05), but consumed proportionately 0·30 less energy (P < 0·001) than their counterparts given the milk diets. No treatment differences in the specific activity of lactase and sucrase were observed (P > 0·05). Significant correlations existed between voluntary food intake and villous height at the proximal jejunum for piglets given the starter diet and ewes' milk (P < 0·05 and P = 0·073, respectively). In turn, villous height was significantly correlated (r = 0·78 to 0·87, P < 0·05) with the rate of body-weight gain after weaning in these two groups. For piglets offered ewes' milk plus glutamine, an increase in DM intake was associated only with increases in crypt depth (P < 0·01). These data show that the structure and function of the small intestine can be preserved when a milk diet is given after weaning, and suggest an association between food intake and villous height in determining post-weaning weight gain.

Fast-myosin in frozen histological sections of eight, 10, 11 and nine muscles of the upper forelimb, lower forelimb, upper hindlimb and lower hindlimb, respectively, of goats was quantified by an immunohistochemical micromethod based on the enzyme-linked immunosorbent assay. The structure of the muscles is well preserved during the immunohistochemical measurement. High fast-myosin levels (more than 201 mg/g total protein) were observed in the triceps brachii (lateral head), rectus femoris, vastus lateralis, semitendinosus, semimembranosus, gastrocnemius (lateral head) and long digital extensor muscles. In contrast, low fast-myosin levels (less than 50 mg/g) were found in the triceps brachii (medial head), superficial digital flexor, vastus intermedialis, and soleus muscles. Fast-myosin-positive fibres (type II or fast-twitch type) were distributed more in the superficial regions than in the deeper regions in the triceps brachii (lateral and long heads), biceps brachii, brachialis, biceps femoris, vastus lateralis, vastus medialis, semimembranosus and gastrocnemius (lateral and medial heads) muscles. In contrast, type IIfibres were distributed more in the deeper regions than in the superficial regions in the extensor carpi radialis, deep digital flexor, cranial tibial, deep digital flexor and superficial digital flexor muscles. When the results obtained by the immunohistochemical micromethod were compared with those obtained by biochemical techniques and by histomorphometrical analyses, high correlations were noted. This technique could be used in research projects to study the muscle characteristics that determine meat quality.

A study was made to investigate the influence of pregnancy feeding level (ad libitum (AL) and restricted (R); 2·0 kg food per day) and lactation length on growth, carcass, meat quality and processing characteristics of pregnant gilts (P) slaughtered after weaning on day 0 (P·0), day 10 (P·10) or day 35 (P·35) after farrowing. Non-pregnant (NP) control gilts were slaughtered at live weights of approximately 95 kg (NP-L), 135 kg (NP-M), and 160 kg (NP-H). P gilts were mated at an average age of 172 days and an average live weight of 97 kg. Pregnancy per se reduced live weight and lean growth rates compared with NP-H animals which were slaughtered at a similar age to the once-bred animals. Weight loss during lactation tended to increase with lactation length and was higher for gilts given food ad libitum during pregnancy compared with restricted animals. The weight of closely trimmed lean in the carcass was similar for the P·0, P·10 and P·35 groups (59·5 v. 62·2 v. 58·8 kg respectively; average s.e. 2·42) and all once-bred gilt groups virtually doubled trimmed lean yields during pregnancy and lactation. Curing yields were similar for P and NP gilts. As lactation length increased, the fat content of the longissimus dorsi and of the overlying subcutaneous fat layers tended to decrease and the moisture content of these two tissues increased. The major impact offeeding level was on fat levels which were significantly higher for AL animals. Trimmed lean yields were also higher for AL gilts suggesting that lean growth rates were not being maximized on the R feeding regime. Subjective muscle colour scores indicated that the longissmus dorsi from P gilts was darker than from NP-L animals and muscle colour tended to become darker with increasing lactation length. There were few significant differences between treatment groups for eating quality. Meat from once-bred gilts was generally judged to be tougher than that from NP-L animals and toughness generally increased with lactation length. This study confirms the potential of the once-bred gilt as an efficient producer of pig meat but raises issues regarding meat quality that require further research.

The hypothesis tested in this experiment was that the structure and function of the small intestine of piglets given a milk liquid diet after weaning depends on their level of energy intake. At weaning (28 days), 42 piglets were allocated to one of five treatments: (1) control group killed at weaning; (2) piglets offered a dry starter diet ad libitum; (3) piglets given cow's fresh milk at maintenance energy intake (Ma); (4) piglets given cow's fresh milk at 2-5 Ma; and (5) piglets given cow's fresh milk ad libitum. On the 5th day all piglets were killed and samples of gut were taken for histological and biochemical examination. Piglets given milk ad libitum grew faster (P < 0·001) than piglets on all other treatments. Piglets offered the dry starter diet ingested similar quantities of dry matter and energy, and grew at the same rate as piglets given cows' milk at 2·5 Ma. As predicted, piglets given milk at maintenance energy intake grew slower (P < 0·001) and consumed less food (P < 0·001) than piglets in all other treatments. For piglets given both cows' fresh milk and the dry starter diet, there were significant linear relationships (r = 0·72 to 0·82, P < 0·05) between villous height and crypt depth with dry matter (energy) intake after weaning. In turn, mean villous height in both milk-fed and starter-fed piglets was significantly correlated (r = 0·68 to 0·79, P < 0·05) to empty body-weight gain in the first 5 days after weaning. Estimates of digestive enzyme activity and absorptive capacity of the gut did not corroborate the large differences seen between treatments in villous height and crypt depth, and possible reasons are discussed. These data illustrate the interdependence between voluntary food intake and mucosal architecture in determining piglet performance after weaning.

Effects of a two-diet system combined with ad libitum lactation feeding of sows on food consumption, sow live weight (LW), backfat depth, condition scoring, rebreeding interval, symptoms of agalactia and culling were studied on 60 sows followed for four parities. Feeding regimes were: CR, conventional i.e. restricted during gestation and lactation; SA, a simplified diet offered at a restricted level during gestation and a conventional diet ad libitum during lactation; CA, conventional during gestation and the same diet ad libitum during lactation. During lactation CR sows consumed 5·9 kg food daily (71 MJ metabolizable energy (ME)), whereas SA and CA sows consumed on average 7·0 kg (85 MJ ME), the difference being larger in multiparous than in primiparous sows and larger in the first than in later weeks of lactation. From the second farrowing onward sows on CR treatment were lighter and had thinner backfat and lower condition scores than sows on CA treatment, with sows on SA treatment in between, i.e. approaching the CR sows in the gestation periods and the CA sows in the lactation periods. At first service LW was 130 kg and backfat thickness 14 mm. During the fourth lactation LW averaged 190 kg in CR sows and around 220 kg for SA and CA sows, and backfat thickness was 11 mm in CR sows and around 16 mm in SA and CA sows. SA sows tended to rebreed earlier after weaning than did CR or CA sows. CA sows and in particular SA sows had higher incidences of agalactia symptoms and higher rectal temperatures post partum than CR sows. The culling rates did not differ between feeding regimes, although the predominant causes for culling did, being shoulder lesions and abscesses in thin CR sows and leg disorders in SA and CA sows. At the end of the experiment, however, more SA sows than others were retained in the experimental herd. It was concluded that the SA feeding regime was superior in terms of sow performance, despite an increased incidence of agalactia symptoms.

Growth modelling was used to characterize the response to digestible lysine in two experiments (114 gilts in experiment 1 and 96 gilts in experiment 2) from 34 to 72·5 kg and 72·5 to 136 kg, respectively. Maize-soya-bean meal diets were formulated to assure that lysine (5·4 to 10·4 and 5·4 to 9·4 g digestible lysine per kg for experiments 1 and 2, respectively) was the first limiting amino acid. Analysis of variance was used to test linear and quadratic responses in cumulative weight gain on test as digestible lysine increased. A time × digestible lysine interaction (linear, P < 0·001) was detected, indicating that a separate regression equation for each lysine level was necessary. In experiment 1, average daily gain (ADG) and carcass crude protein (CP) accretion were maximized for gilts given 10·4, 9·4 and 8·4 g digestible lysine per kg from 34 to 44 kg, 44 to 54 kg, and 54 to 72·5 kg, respectively. Lipid accretion was minimized for gilts given 7·4 to 8·4 g digestible lysine per kg. In experiment 2, ADG was maximized by feeding 8·4 g/kg from 72·5 to 92·5 kg and 7·4 g/kg from 92·5 to 136 kg. Carcass CP accretion was maximized by feeding 9·4 g digestible lysine per kg, whereas lipid accretion was minimized for gilts given 8·4 g digestible lysine per kg from 72·5 to 136 kg. If feeding graded levels of digestible lysine resulted in parallel lines for protein accretion, mean values would result in accurate data evaluation. However, responses to digestible lysine changed over the feeding period. Therefore, the use of body weight and compositional growth curves offers an approach to more accurately characterize the growing pig's response to increased digestible lysine.

Sixty gilts were allotted to one of the following feeding regimes: CR, a conventional diet offered at a restricted level according to Swedish feeding standards; SA, a simplified gestation diet (mainly oats and barley) offered at a restricted level and a conventional diet offered ad libitum during lactation; CA, a conventional gestation diet offered at a restricted level and the same diet offered ad libitum during lactation. The sows were followed for four parities. Piglet birth weight was unaffected by sow feeding, but differences in piglet live weight were found from 3 weeks of age to the end of the experiment when the piglets were 9 weeks old. CA piglets were heavier than SA piglets at 3 and 9 weeks of age and heavier than both CR and SA piglets at weaning (35 ± 3 days). Litter size was unaffected by treatment as was piglet mortality, whereas the cause of death varied with treatment. Mortality due to trauma was higher among CA piglets than among others, while mortality due to weakness was higher among SA piglets than others and there was also a tendency for higher mortality due to diarrhoea among CR piglets. It was concluded that the CA feeding regime was superior with regards to piglet performance.

A total of 315 first parity sows were used in two experiments to establish the lactational and subsequent reproductive performance of first-litter sows in response to varying levels of dietary amino acids and digestible energy (DE) offered during lactation. In the first experiment, sows were offered five diets ranging in lysine content from 6·2 to 15·1 g/kg. In the second experiment sows were offered five diets with a similar lysine: DE ratio (0·88 to 0·85 g/MJ) but ranging in DE content from 12·6 to 15·1 MJ/kg. Voluntary food intake during lactation was unaffected by lysine or DE content of the diets. Piglet pre-weaning growth was similarly unaffected by dietary DE content but responded in a curvilinear manner with increasing dietary lysine. Sow body-weight loss during lactation was unaffected by dietary lysine but declined with increasing DE content up to 13·8 MJ/kg. Backfat loss was lowest for sows offered the lowest lysine diet and tended to reach a plateau at a dietary level of 10·6 g lysine per kg. In response to dietary DE content, backfat loss tended to be highest for sows offered the diet of lowest DE content and declined with increasing DE content up to 15·1 MJ DE per kg. The subsequent weaning to remating interval was unaffected by dietary DE content. Subsequent litter size was similarly unaffected by dietary DE content but was significantly higher for sows offered the two higher, compared with the two lower, lysine diets (10·7 v. 9·6 born alive, s.e.d. 0·5, P < 0·05). The results suggest that current amino acid recommendations for lactating first-litter sows are inappropriate and may be constraining both milk production and subsequent reproductive performance.