Annual surveys of fertilizer practice have been conducted since 1942. Since 1985, nitrogen application rates to grass have progressively declined and phosphorus application has been reduced since the mid 1970s. Increasing environmental issues and the present interest in organic farming and low input systems indicate that these trends will continue. Present overall fertilizer use for grazing on dairy farms is about 170 kg N, 10 kg P and 20 kg К per ha. Higher rates are used for intensive silage production, 200 kg N, 15 kg P and 50 kg К per ha. In contrast, recommended applications are 340 kg N, 18 kg P and 25 kg К per ha for grazing and 380 kg N, 40 kg P and 260 kg К per ha for intensive silage. Herbage yield is controlled by the amount and timing of nitrogen fertilizer applications. The major mineral content depends on herbage maturity within the growth cycle; both phosphorus and potassium contents fall with declining crude protein concentrations.

Nitrogen fertilizer generally increases both magnesium and sodium concentration whilst potassium application decreases both. The overall effect of combined applications is to minimize changes. In the critical spring and autumn periods, herbage dry-matter intakes may be as important as magnesium content and availability when assessing the risk of hypomagnesaemia in the grazing cow.

Fertilizers, including phosphorus, even in the long term, have little effect on the calcium and phosphorus contents of herbage. Nitrogen and phosphorus concentrations are significantly ocrrelated in grazed herbage, silage and hay. If environmental or economic policies were to result in a reduction in the crude protein content of grazed herbage from ca. 250 to ca. 200 g/kg dry matter, the phosphorus intake of a cow giving 30 kg milk per day would fall by 0·5 g/kg dry-matter intake. This, together with any desirable additional magnesium, should preferably be given as an oral mineral supplement rather than by attempting to modify the mineral content of herbage.

With a view to comparing previous findings from experimental journeys, with observations during commercial transport of sheep, hormonal and other physiological data were gathered during three long journeys (866 to 1178 km), all of which involved a sea crossing (1 h 45 min to 8 h 55 min). One was a commercial journey from the United Kingdom to continental Europe and the other two, which were simulated commercial journeys under the control of the investigators, were from a Scottish island to Cambridge. In all journeys there was a break in lairage (duration 13 to 24 h). The stocking rate on the commercial journey (0·17 m2 per sheep for shorn sheep of 35 to 37 kg body weight) was less generous than previously used in experimental journeys. In all journeys there was an apparent increase in plasma concentration of cortisol soon after loading and commencement of travel, with subsequent decline. Transient increases in beta-endorphin and prolactin were also observed but changes in creatine kinase were not obvious. Haematocrit showed slight evidence o f a decline during the journeys. These findings are generally similar to those obtained in experimental journeys and there was no evidence of a hormonal response to sea conditions during the ferry crossings. When transport was resumed after a period of rest with food and water off the vehicle, the physiological responses were less marked than they had been during the first stage of the journey. Under the circumstances observed, long journeys of the type commonly practised commercially probably do not of themselves pose a major welfare challenge.

This study was conducted to determine whether carotenoid pigments can act as biomarkers of grass feeding to trace lamb production systems. Three production systems were compared: G = grazing (72 lambs), S = stall-feeding (26 lambs), and GS = grazing period followed by a stall-feeding period (27 lambs). Presence of carotenoids in tissues was evaluated by plasma concentration and reflectance spectrum of subcutaneous caudal adipose tissue. Plasma carotenoid content was measured during the grazing period for G and GS lambs, at the end of the grazing period for GS lambs and at slaughter for G, S and GS lambs. Reflectance spectrum of adipose tissue was measured at slaughter for 38 G and 26 S lambs. We collected 135 G and 26 S blood samples. Carotenoids were detected in 97% of the G blood samples, whereas they were not detected in 93% of the S blood samples. Plasma carotenoid content of all the GS lambs decreased during the stall-feeding period. Mean reflectance spectra of adipose tissue of G and S lambs differed between 450 and 510 nm, which corresponds to light absorption by carotenoids. We performed a mathematical analysis of the spectrum in order to quantify absorbance in this zone and to propose an index that can be used in the meat industry to trace animal production systems. This method was less discriminating than plasma carotenoid analysis, as there was some overlapping in the frequency distribution of the traceability index for G and S lambs that concerned 19% of the lambs. Sensitivity of the two methods to bias and applicability to cattle are discussed.

Effects of heat load and of photoperiod on lactation performance were evaluated using milk test data of three Israeli Holstein herds over a period of 3 years, from 1994 to 1996. All together 2209 cows, with 28029 milk records, were included. Photoperiod effects were examined as associated with day length and daily changes in day length and heat load index was formulated as a function of the seasonal day and night ambient temperatures, to account for the heat load effect. The regression model included effects of cow, herd, year, lactation number and days in milk in addition to the seasonal effects. The dependent variables were milk yield and fat, protein and lactose concentrations. Milk yield was affected by both photoperiod and heat load, with the peak photoperiod effect in May and amplitude of 3·1 (s.d. 0·9) kg/day and negative heat load effect of-1-8 (s.d. 0·4) kg/day at its peak. Protein concentration was affected by photoperiod, with the peak effect in January (amplitude of 1·7 (s.e. 0·5) g/kg) but not by heat load (-0-2 (s.d. 0·6) glkg). Fat concentration was affected primarily by heat load (-3-4 (s.d. 0·7) g/kg), with a photoperiod effect which peaked in October (amplitude of 1·8 (s.d. 0·8) g/kg). Lactose concentration was affected by both environmental factors to a lesser extent (photoperiod amplitude of 0·6 (s.d. 0·2) g/kg and heat load effect of-0-03 (s.d. 0·16) g/kg). Comparison of the predicted seasonal effects on milk yield and composition with the annual fluctuation in the national herd showed a good match of the predicted effects with the national observations. It is concluded that while heat load relief may be beneficial, manipulation of the photoperiod may induce adverse effects on milk yield and composition.

Economic values for average ad libitum food intake capacity (FIC) were estimated with two different linear-plateau models. One model based on a constant minimum fat to protein deposition ratio (Rmin). The other model relied on a constant minimum marginal ratio between fat and protein deposition (MR). Economic values were used to define aggregate genotypes where the driving variables of the growth models were considered as traits. The consequences of a possible implementation of such aggregate genotypes were evaluated with a comparison of the expected selection responses for the different traits.

Economic values for the traits under the two models indicated the same direction of selection and differed only slightly in their magnitude. For a FIC below its optimum, FIC had a positive and Rmin or MR a negative, economic value. In this situation the maximum protein deposition rate Pdmax had no influence on the production costs and therefore an economic value of zero. Selection emphasis is then mainly put on FIC which leads to an increase in average daily gain (ADG). If FIC is above its optimum, economic values for FIC were negative, positive for Pdmax and equal to zero for Rmin or MR. Practically no differences between the selection responses under the two growth models were observed.

The use of growth models to estimate economic values for FIC allowed to attribute a positive economic value to this trait if it is below its optimum. This is not possible when economic values are estimated with an economic production model.

Three hundred and thirty (330) records from 363 lactations and 104 cows born between 1979 and 1991 from crossbreeding Holstein-Friesian (H) bulls with H, Ngaoundere Gudali (G) and Holstein × Gudali F1 (H1G1) cows, Montbeliard (M) bulls with G and Montbeliard × Gudali Fl (M1G1) cows and M1G1 bulls with M1G1 cows and raised in the Dairy Herd Unit of the Wakwa Animal and Veterinary Research Station in the sub-humid highlands of Cameroon were analysed to investigate the effects of genotype and environment on lactation and reproductive traits. Fixed effects of genotype season and year of calving parity and age of cow at calving were studied. Lactation milk yield (LMY), lactation length (LL), annualized milk production (AMP), calving interval (C1), dry period (DP) and age at first calving (AFC) were amongst the traits analysed. F1 crosses (H1G1, M1G1), backcrosses (¾ Holstein — ¼ Gudali; ¾ Montbeliard — ¼ Gudali) (H3G1, M3G1) and F2 crosses (MGF2) were compared. Results showed that H1G1 cows were proportionately 0·49 and 0·23 better than their corresponding backcross in LMY and AMP, respectively, while M1G1 cows were proportionately 0·12 and 0·24 inferior to M3G1 in these traits. M1G1 were proportionately superior to MGF2 in LMY (0·05), AMP (-0·04) and AFC (-0·14). M3G1 were superior (0·18 to 0·19) to MGF2 in LMY and AMP. H1G1 were consistently superior to M1G1 for most traits, suggesting that H1G1 may be better than M1G1 for dairy production.

Genotype with nutrition interactions in carcass composition and meat quality traits were examined by testing pigs from four selection lines and a control line on isoenergetic diets, which differed in lysine: energy content. The selected lines resulted from seven generations of selection for high daily food intake, lean food conversion ratio (LFC) and lean growth rate on ad-libitum or restricted (LGS) feeding regimes in a Large White population. There were 128 pigs in the study, with 24 pigs per selection line and 32 pigs from a control line. During performance test, 30 to 90 kg, pigs were offered one of three isoenergetic diets, 14·0 MJ digestible energy (DE) per kg dry matter, which differed in ileal digestible lysine: digestible energy (A: 0·40 , C: 0·76 and E: 1·12 g lysine per MJ DE) on ad-libitum or restricted (0.75 g/g ad-libitum daily food intake) feeding regimes.

For the majority of performance test, carcass composition and meat quality traits there was no evidence of a genotype with diet or genotype with feeding regime interaction. The selection line with feeding regime interactions for average daily gain, daily food intake and rates of lean and subcutaneous fat deposition were primarily due to no feeding regime effect for the LFC selection line. Selection for high LFC had reduced ad-libitum daily food intake to such an extent that it was not significantly different from daily food intake on a restricted feeding regime, unlike other selection lines in the study. A selection line with feeding regime interaction was detected for muscle рH24h and muscle reflectance, which resulted from the LGS selection line. LGS pigs offered food ad libitum had higher muscle рH24h and lower muscle reflectance than LGS pigs given food at a restricted level, while there was no effect of feeding regime for the other selection lines.

The general absence of genotype with nutrition interactions for traits measured in the study indicated that the ranking of genotypes for performance test traits, carcass composition and meat quality traits will not be dependent on diet or feeding regime. Genotype specific nutritional inputs will also not be required for identification of pigs of high genetic merit, within a genotype. However, diet and feeding regime had significant effects on carcass composition and meat quality traits, such that the estimated mean value of a genotype will be dependent on the diet or feeding regime used to evaluate the genotype.

This study investigated the effects of the percentage of Texel or Rouge de l’Ouest (Rouge) genes in lambs (0, 50, 75 or 100%; the remainder being Greyface (Border Leicester × Scottish Blackface)) on lamb carcass characteristics and meat quality. Ewe lambs were slaughtered at 34, 40, 46 and 52 kg and ram lambs were slaughtered at 40, 46, 52 and 58 kg live weight. Dressing proportion increased (P < 0·001) by 0·53 and 0·27 g carcass weight per kg live weight for each 1% increase in Texel or Rouge genes respectively. Carcass conformation classification increased (P < 0·001) by 0·016 and 0·007 units (on a five-point scale) for each 1% increase in Texel or Rouge genes respectively. Carcass fat depth measures were reduced by Texel and Rouge genes such that at a constant fat depth end point, carcass weight could be increased by 0·029 and 0·023 kg for each 1% increase in Texel or Rouge genes. Carcass lean content increased (P < 0·001) by 0·99 and 0·27 g/kg for each 1% increase in Texel or Rouge genes. Subcutaneous fat content was reduced (P < 0·001) by 0·36 and 0·29 g/kg for each 1% increase in Texel or Rouge genes. Intermuscular fat content was reduced (P < 0·01) by 0·38 g/kg per 1% increase in Texel genes, whereas Rouge genes had no effect. Similarly, bone content was reduced (P < 0·01) by Texel genes only (0·31 g/kg per 1% increase in Texel genes). Warner-Bratzler shear force values showed a quadratic effect with increasing Texel and Rouge genes, with an initial decrease from 0 to 50% followed by an increase to 100%. Cooking loss increased (P = 0·05) as the percentage of Rouge genes in lambs increased. Increasing the percentage of Texel or Rouge genes significantly increased L* values (P < 0·01 and P < 0·05 respectively), b* values (P = 0·05 and P< 0·05 respectively) and H° values (P < 0·001).

Ages at first behavioural oestrus and at elevated plasma progesterone concentration were measured in three selection groups, after seven generations of divergent selection for lean growth rate (LGA), lean food conversion (LFC) and daily food intake (DFI) in a population of Large White pigs. First physiological oestrus was defined to have occurred when a blood plasma progesterone concentration of at least 1 μg/l was detected from weekly sampling of gilts. The study consisted of 146 gilts, which were given 0·75 , 0·81 , 0·88 , 0·94 or 1·0 g/g of daily ad-libitum food intake during performance test and then 1.9, 2.05, 2.2, 2.35 or 2.5 kg/day, respectively, until conception, to determine if there were differences between selection lines in their sensitivity to changes in nutritional inputs.

Responses in oestrus and performance test traits were dependent on selection group. First physiological oestrus was later with selection for high LFC than for low LFC (234 v. 215, s.e.d. 9.1 days) but there was no significant response within each of the LG A (224 v. 226 days) and DFI (218 v. 206 days) selection groups. The probability of exhibiting oestrous behaviour signs at first physiological oestrus was significantly lower in the high LG A line (0·62 v. 0·93 or 0·5 v. 2.5, s.e.d. 0·75 on the logit scale) than in the low line but there were no responses in the LFC and DFI groups. For animals exhibiting oestrous behaviour signs at first physiological oestrus, there were no significant responses in oestrous behaviour score for the three selection groups. Live weight at first physiological oestrus in the LFC and LG A selection groups was greater in the high lines than in the low lines (120 v. 109 and 123 v. 112, s.e.d. 4.3 kg) but not in the DFI selection group (116 v. 111 kg). Responses in ultrasonic backfat (-7.3, -8.2 and 5.0, s.e.d. 1.5 mm) and muscle depth (4.9, 6.1 and -3.5, s.e.d. 1.4 mm) at first physiological oestrus were of similar magnitude in the LGA, LFC and DFI selection groups.

Increasing the ration (amount of food offered) did not have a linear effect on performance test traits and reproductive development, such that ration had to be included in the model as a fixed effect, rather than a covariate. There was no significant effect of ration or of selection line with ration interaction for traits associated with first oestrus.

Selection for lean growth rate had no adverse effect on reproductive development, unlike selection for lean food conversion. Detection of first oestrus with oestrous behaviour signs combined with physiological assessment may be required in genotypes selected exclusively for lean growth rate, rather than relying only on observed behavioural signs of oestrus.

Ages at first behavioural oestrus and at elevated plasma progesterone concentration were measured in three selection groups, after seven generations of divergent selection for lean growth rate (LGA), lean food conversion (LFC) and daily food intake (DFI) in a population of Large White pigs. First physiological oestrus was defined to have occurred when a blood plasma progesterone concentration of at least 1 μg/l was detected from weekly sampling of gilts. The study consisted of 146 gilts, which were given 0·75 , 0·81 , 0·88 , 0·94 or 1·0 g/g of daily ad-libitum food intake during performance test and then 1.9, 2.05, 2.2, 2.35 or 2.5 kg/day, respectively, until conception, to determine if there were differences between selection lines in their sensitivity to changes in nutritional inputs.

Responses in oestrus and performance test traits were dependent on selection group. First physiological oestrus was later with selection for high LFC than for low LFC (234 v. 215, s.e.d. 9.1 days) but there was no significant response within each of the LG A (224 v. 226 days) and DFI (218 v. 206 days) selection groups. The probability of exhibiting oestrous behaviour signs at first physiological oestrus was significantly lower in the high LG A line (0·62 v. 0·93 or 0·5 v. 2.5, s.e.d. 0·75 on the logit scale) than in the low line but there were no responses in the LFC and DFI groups. For animals exhibiting oestrous behaviour signs at first physiological oestrus, there were no significant responses in oestrous behaviour score for the three selection groups. Live weight at first physiological oestrus in the LFC and LG A selection groups was greater in the high lines than in the low lines (120 v. 109 and 123 v. 112, s.e.d. 4.3 kg) but not in the DFI selection group (116 v. 111 kg). Responses in ultrasonic backfat (-7.3, -8.2 and 5.0, s.e.d. 1.5 mm) and muscle depth (4.9, 6.1 and -3.5, s.e.d. 1.4 mm) at first physiological oestrus were of similar magnitude in the LGA, LFC and DFI selection groups. Increasing the ration (amount of food offered) did not have a linear effect on performance test traits and reproductive development, such that ration had to be included in the model as a fixed effect, rather than a covariate. There was no significant effect of ration or of selection line with ration interaction for traits associated with first oestrus. Selection for lean growth rate had no adverse effect on reproductive development, unlike selection for lean food conversion. Detection of first oestrus with oestrous behaviour signs combined with physiological assessment may be required in genotypes selected exclusively for lean growth rate, rather than relying only on observed behavioural signs of oestrus.

Measurements of conformation and pelvic characteristics of pedigree Belgian Blue cows were compared in cattle from 11 herds. In seven herds comprising 56 cows caesarean sections were routinely used to deliver pedigree calves and in four herds comprising 52 cows they were given the opportunity to calve naturally per vaginum. Cows in the seven herds were approximately 10 months younger than those in the other herds. External pelvic measurements were linear distances between the tuber coxae, between the tuber coxarum and the tuber ischium, between the tuber ischii, the vertical distance between the underside of the tail and the tuber ischii, and heights of the tuber coxae and tuber ischii from the floor. Internal pelvic measurements were the vertical distance between the ventral body of the sacrum and the pubic symphysis and the horizontal distance between the wings of the ileum at the widest point of the pelvic inlet, both made with a Rice pelvimeter. Condition scores were assessed for all cows. There were correlations (P < 0·001) between internal pelvic height and width, external measurements, and age. For cattle that calved unassisted, pelvic height (P < 0·05) and area (P < 0·05) were significantly larger than those delivered by elective caesarean section. From internal pelvic measurements of 74 cows, those which always delivered live calves had a larger pelvic area (P < 0·01) than cows that delivered a dead foetus following any parturition.

The effect of genotype on the size of adipocytes and on the lipogenic enzyme activities in different fat depots in Asturiana, Avilena, Morucha, Parda Alpina, Pirenaica, Retinta and Rubia Gallega steers was studied (six animals per breed). Steers were weaned at about 6 to 8 months of age and 220 to 260 kg live weight (LW). During the fattening period the steers were given commercial concentrates and straw, both ad libitum, until slaughter at 470 kg average LW. Adipocyte size and activity of the following lipogenic enzymes were studied: glycerol 3-phosphate dehydrogenase (G 3-PDH); fatty acid synthase (FAS); NADP-malate dehydrogenase (MD), glucose 6-phosphate dehydrogenase (G 6-PDH) and NADP-isocitrate dehydrogenase (ICDH) in the omental, perirenal, subcutaneous and intermuscular adipose depots. Also the perirenal fat depot was removed and weighed and the 10th rib was dissected. Differences between breeds were significant for the amount of fat in the different adipose depots (P < 0·001), for adipocyte size (except in intermuscular depot) and for G 3-PDH, FAS, MD, ICDH (P < 0·001) and G 6-PDH (P < 0·01). In general, steers of breeds that had more fat also had bigger adipocytes and higher lipogenic enzyme activities. In this way, Asturiana breed (lean breed) had smaller adipocytes and lower lipogenic enzyme activities than Morucha breed (fatter breed). Finally, adipocytes from the perirenal depot were the biggest and showed the highest G 3-PDH and FAS activities while adipocytes from the intermuscular depot had the smallest size and the lowest lipogenic activities.

The effect of feeding different sources of dietary fibre at varying rates of inclusion on the concentrations of skatole and indole in the backfat of finishing pigs and their subsequent carcass quality was studied. Sugar-beet pulp (SBP) and straw (S) were selected because of their assumed widely differing hind-gut fermentability characteristics. They were individually and in combination added to a diet at three rates of inclusion designed to provide concentrations of 40, 80 and 120 g non-starch polysaccharide (NSP) per kg. The digestible energy (DE) contents varied between the three levels of inclusion of NSP but were formulated to a lysine: energy ratio of 0.625 g: MJ DE with other amino acids being in the same individual proportion to lysine. The nine experimental diets were offered to 10 replicate (five entire males and five females) pigs of initial live weight 55 kg. Food intakes were adjusted between treatment to ensure that energy and nutrient intakes were common. Animals were slaughtered on reaching 85 kg live weight. Adipose tissue was analysed for skatole and indole and meat quality characteristics assessed. No significant differences in performance were detected although carcasses from animals given sugar-beet pulp were significantly fatter. There were significant correlations between the level of skatole in the backfat and the following eating quality characteristics: skatole odour intensity of the fat, abnormal odour intensity of fat, abnormal odour intensity of lean and abnormal flavour intensity of lean with correlations coefficients of 0·656 , 0·709 , 0·496 and 0·523 respectively (coefficients >0.276; P < 0·05). The range in skatole levels in the backfat were attributable substantially to very high levels found in three boars. No significant correlations were found between indole concentrations in the backfat and eating quality characteristics. However no effects of dietary NSP type or inclusion level on the eating quality characteristics of pig meat, with pigs given diets on the basis of equal energy and nutrient intakes, were detected.

The aim of this study was to rank diverse beef cattle genotypes for meat quality characteristics and to determine whether that ranking changed depending on the environment in which the animals were finished. Breed groups, ranging from 100% Bos indicus to 100% Bos taurus content, were derived from Indian zebu (Brahman), African zebu (Boran), British breeds (Hereford-Shorthorn), continental breeds (Charolais and Simmental) and Sanga (Tuli and Belmont Red). Heavy steers (>600 kg) were raised in a tropical environment and finished either on pasture or in a feedlot.

For striploins (longissimus) from feedlot finished steers, cooking loss was greatest for zebu (Z) steaks, least for British (B) steaks (P < 0·001), and intermediate for the other breeds. For striploins from pasture finished steers and eye rounds (semitendinosus) from both pasture finished and feedlot finished steers, there were no breed differences in cooking loss.

For both feedlot finished and pasture finished steers, striploin steaks from B steers were most tender, and Sanga (S) and zebu × continental cross (ZC) steaks were more tender than Z steaks. Warner-Bratzler initial yields suggested that the lower toughness of these breeds was due to lower myofibrillar toughness. Peak force minus initial yield suggested a smaller, opposite effect of increasing connective tissue toughness associated with increasing Bos taurus content. This breed effect on connective tissue toughness was more prominent in the eye round samples. In the feedlot finished steers, Z eye round was as tender as B eye round. In pasture finished steers, Z eye round had a peak force 1·3 kg lower than B in contrast to the results for striploin where Z was 1·2 kg higher than B. Sanga eye round followed the same trend as B relative to Z but to a lesser extent. From the feedlot, S eye round was more tender than Z eye round and from pasture, S eye round was the same as Z eye round. There was no evidence of heterosis f or any meat quality attribute.

These results are best explained in terms of a model incorporating differential effects of breed on the myofibrillar and connective tissue components of toughness. The implications of this study are that the tenderness of grilling and roasting cuts of meat from the predominantly Brahman-based beef herd of northern Australia can be improved through crossbreeding with any of the taurine breeds studied.

We have examined the effect of dietary fats containing n-3 polyunsaturated fatty acids on the conjugated linoleic acid (CLA) content of beef m. longissimus lumborum. Four groups of eight Charolais steers were given, for 120 days, grass silage plus a barley/sugar-beet feed concentrate containing one of four fat supplements: Megalac (saturated), linseed (high 18: 3),fish oil (high 20: 5 n-3, eicosapentenoic acid and 22: 6 n-3, docosahexaenoic acid) or linseed plus fish oil. The concentrates supplied 400 g/kg dry-matter (DM) intake and were designed to supply 45 g/kg of the total dietary fat calculated to be 60 g/kg of DM and to contain similar amounts of linoleic acid. Muscle from steers given the Megalac supplement contained 11·3 mg CLA per 100 g muscle and this was increased two- to three-fold in animals given the more unsaturated fat supplements. The increased deposition of CLA was similar for both linseed and fish oil supplements although the concentrations of total n-3 polyunsaturated fatty acids in the fish oil diet were much less than in the linseed diet. This suggests potent inhibition of conversion of CLA to trans vaccenic acid by fish oil fatty acids or their rumen metabolites.

Three experiments were conducted to investigate the effect of bag incubation sequence on the degradation of food components in situ in the caecum of mature, caecally fistulated Welsh-cross pony geldings (mean live weight 278 kg) offered hay ad libitum. In experiment 1 a fibre-based commercial horse concentrate was incubated in situ using a forward (3, 5, 16, 8, 24, 48 h) or reverse (48, 24, 8, 16, 5, 3 h) incubation sequence. Dry matter (DM), crude protein (CP), neutral-detergent fibre (NDF) and acid-detergent fibre (ADF) degradation coefficients and calculated effective degradability (ED) values were determined. In experiment 2 unmolassed sugar-beet pulp (USBP), hay cubes (HC), soya hulls (SH) and a 2: 1 mixture of oat hulls: naked oats (OHNO) were incubated in situ as for experiment 1. In experiment 3 unprocessed barley (UB), micronized barley (MB), extruded barley (EB) and dehydrated grass (DHG) were incubated in situ according to slightly different forward or reverse incubation sequences of (2, 4, 6, 12, 8, 24, 48 h) and (48, 24, 8, 4, 12, 6, 2 h) respectively. In experiments 2 and 3 only DM degradation parameters were studied.

Of the three starch-based foods studied in experiment 3 (UB, MB and EB), incubation sequence did not significantly P > 0·05) affect any of the degradation parameters examined. Conversely however, of the six fibre-based foods which were examined across the three experiments, incubation sequence did significantly P < 0·05) affect in situ degradation parameters in the commercial horse concentrate in experiment 1, the SH food in experiment 2 and the DHG food in experiment 3. Depending on the food or food constituent studied (i.e. DM, CP, NDF or ADF) degradation coefficients a, b, c and a + b along with ED values calculated at fractional outflow rates of 0·05 and 0·025 could all be statistically different CP < 0·05) according to whether a forward or reverse incubation sequence was used. It is postulated that this effect is related to the basic digestive physiology of the equine caecum which is small, digesta passage rate through it is fast and digesta volumes can vary considerably. These factors may interact to create a considerable degree of non-uniformity within the caecal digesta pool in which in situ bags are incubated. Consequently, it is recommended that in future in situ experiments in the equine hindgut, animals are offered ad libitum diets in an attempt to minimize variation within the caecum. It is also recommended that in situ experimental protocols incorporate more than one incubation sequence when the degradation parameters of fibrous foods are studied in equids.

Six experimental diets with similar contents of macronutrients (450, 300 and 70 g/kg dry matter of protein, carbohydrate and fat, respectively) and, therefore, total energy, hut differing in the composition of lipid component, were each given to three replicate groups of European eel. Different fat mixtures were used to formulate diets with different levels of fatty acids considered as essential for fish: linoleic (18: 2n6), linolenic (18: ЗпЗ) and highly unsaturated fatty acid (HUFA)n3, mainly EPA (20: 5n3) and DHA (22: 6n3). One tested diet contained low levels of all the three components, three diets had a high level of each one of these, the fifth diet contained high levels of both 18C fatty acids and, finally, a sixth diet incorporated simultaneously high amounts of 18: 2n6 and (HUFA)nЗ.

Dietary lipid composition did not significantly affect food intake or protein and fat apparent digestibility. However, growth and food efficiency were the best with the diet containing approximately 5 g/kg of both linoleic and linolenic acids. Although all groups increased their body fat content over the experiment, the fatty acid composition of total muscle lipids was rather insensitive to lipid dietary composition, perhaps due to the high body fat content of the eels prior to the experiment.

Twenty-five New Zealand White × Californian lactating doe rabbits were used to study the effect of different dietary fibre concentrations on the amino acid contribution of soft faeces to the total amino acid intake. Five isoenergetic diets containing 312, 334, 360, 384, and 412 g/kg neutral-detergent fibre (DM basis) were formulated. All diets were designed to maintain the same amino acid pattern. Soft faeces production showed a trend to increase (P = 0·07) as dietary fibre increased. The content of isoleucine (P < 0·01), glutamic acid, leucine and alanine (P < 0·05) in soft faeces decreased as the dietary fibre increased. A trend to decrease with the increase of dietary fibre was also observed for aspartic acid (P = 0·06), lysine (P = 0·07), methionine and valine (P = 0·1). As a consequence, the soft faeces contribution to the total intake of individual amino acids did not vary with the type of diet. The average contribution of soft faeces to the total amino acid intake for the most limiting amino acids of lactating doe rabbits under practical conditions, methionine, threonine and lysine, were 0·17 , 0·21 and 0·18 g/g, respectively.

Thermal environment and water supply are factors which are supposed to influence the performance and well being of farm-raised mink but conclusive literature data are still very limited. The present series of experiments was conducted in order to quantify effects of thermal environment on water intake and excretion, urinary solute excretion and estimate possible stress reactions. Water intake and excretion, urinary osmolality and urinary excretion of sodium (Na), potassium (K), cortisol and catecholamines were studied in a balanced Latin-square design experiment with six adult male mink, kept at three different ambient temperatures (Ta); (5°C, 20 °C and close to 35 °C) and given three different water supplies (E: extra water in the food; N: normal ad libitum drinking water supply; R: restricted, free access to drinking water twice daily). The experiment comprised nine periods, each of 3 days. Food apparent digestibility, intake of metabolizable energy (ME), metabolic and evaporative water, oxidation of nutrients and ME requirement for maintenance (MEm) were calculated. Water intake was strongly affected by Ta, with dietary water being the major source at 5°C and 20 °C, its importance being profoundly exceeded by drinking water at 35 °C. Water excretion in urine was highest at the lowest Ta and lowest at the highest Ta. Restriction of access to drinking water resulted in lower total water intake, and excretion, mainly by decreased urinary volume, reflected by increased urinary osmolality and increased solute concentration. ME intake decreased as Ta increased but urinary and total water output per kJ ME was not significantly affected by Ta or water supply. Excretion of Na and К per kJ ME mainly monitored urinary water excretion, being highest under conditions when urine production was highest. Metabolic water made up 0·14 to 0·17 of the total water intake and evaporative loss increased from about 50 to 125 g/kg live weight (M)0·75 as Ta increased. MEm was lowest (534 kJ/kg M0·75) at 20 °C, and highest (647 kJ/kg M0·75) at 35 °C. Cortisol excretion generally tended to increase when water supply was restricted, the increase being significant at 35 °C. This, in combination with outer signs of distress in the animals, emphasized that this high temperature in combination with limited access to drinking water exposed the animals to a very stressful situation but rectal temperature remained normal, indicating intact temperature regulation mechanisms.

Quantitative water intake and excretion, urinary osmolality and urinary excretion of Na, K and cortisol were studied in an experiment with nine adult male mink. The experiment comprised four periods each of 3 days, and ambient temperature (Ta) was increased from 5°C to 20 °C (periods 1 to 2) and, after an interval of 1 day, continued at 20°C and then decreased to 5°C (periods 3 to 4). Three different water supplies were given, namely E: extra water in the food, N: normal ad libitum water supply or R: restricted, free access to drinking water twice daily. Intake of metabolizable energy (ME) was calculated, and the water intake and excretion data were related to ME intake. The effect of Ta on quantitative water intake and excretion was evident, with the greatest response recorded when Ta was decreased. Intake of drinking water was highest at 20 °C, as well as the total water balance. The water supply had some, but not a profound, influence on water intake and excretion, with tendencies for reduced drinking water intake and water excretion in urine for R mink. Intake of ME decreased during period 3 at 20 °C and then increased when Ta was decreased to 5°C, which was reflected by an increased 24-h excretion of Na and K as well as an increased osmolality of the urine. Therefore, it was concluded that water intake and excretion were affected by combined effects of Ta and ME intake. Generally, it was shown that the mink responded rapidly to changes in Ta, and within 24 h significant effects of the change could be recorded for several of the traits studied.

Relationships between photoperiod and cycles of voluntary food intake (VFI) and maintenance energy requirements (MER) were determined in Soay rams, subjected to a 6-month photoperiodic regime. Food was offered ad libitum (no. = 5) or at a predicted maintenance level (no. = 5). All rams demonstrated 6-month cycles of VFI, growth and reproductive status. Metabolizable energy intake (MEI) was greatest in rams given food ad libitum (666 (s.e. 21.7) kJ/kg metabolic live weight (M0·75) and food-restricted (528 (s.e. 12.2) kJ/kg M0·75) rams during sexual quiescence. Conversely, MEI was minimal (ad libitum, 289 (s.e. 8.4) kJ/kg M0·75; restricted, 428 (s.e. 8.1) kJ/kg M0·75) during the rut. Distinct cycles of heat production (HP) accompanied changes in MEL Changes in HP were similar (P > 0·05) for both groups (ad libitum, 520 (s.e. 22.1) to 394 (s.e. 9.2) kJ/kg M0·75; restricted 503 (s.e. 14.0) to 407 (s.e. 17.5) kJ/kg M0·75) and therefore energy retention varied more (P < 0·015) when rams were given food ad libitum (ad libitum, 131 (s.e. 43-1) to -106 (s.e. 38.2) kJ/kg M0·75; restricted, 78·0 (s.e. 27.1) to -53.0 (s.e. 38.2) kJ/ kg M0·75). Apparent digestibility of dietary energy varied inversely with MEI (P < 0·01). MERs ranged from 524 (s.e. 35.0) kJ/kg M0·75 to 401 (s.e. 27.3) kJ/kg M0·75, a proportional fluctuation of ±0·13. Changes in metabolic rate preceded those in appetite, suggesting a causal relationship.