Welfare of animals can be defined as the kind of feelings the environmental conditions bring about in the animals. These feelings depend on the needs of the animals and their degree of satisfaction. Needs of animals, and so their welfare, are partly genetically determined. Therefore, welfare can be changed by breeding. The aim of this study was to investigate how welfare of pigs under modern intensive farm conditions can be improved by genetic selection, with emphasis on the precise definition of the breeding goal and determination of the animal characteristics on which selection can be based in practice.

The existing thermoregulation model was used to develop a conceptual framework that describes welfare of growing pigs and production sows with respect to each of their needs as a curvilinear function of the respective environmental conditions. The framework assumes that welfare in terms of feelings is reflected by the physiological and behavioural mechanisms the pig has to activate in order to cope with the various environmental conditions it encounters. Based on those physiological and behavioural responses to changing conditions, five welfare zones can be distinguished for each need. Breeding goals for welfare were defined in terms of the transition points between these welfare zones, such that future pigs would better cope with unfavourable or unfamiliar farming conditions, therewith quickening the domestication process, to some extent. However, as long as genetic parameters for these transition points are not available, more common welfare-related characteristics like temperament, stress resistance and robustness can be included in the breeding goal, as an alternative.

For selection among potential breeding candidates, transition points between welfare zones can be determined in sib tests, thereby also collecting the data for estimating genetic parameters. As a cheaper alternative, breeding candidates could be tested under hard conditions and selected on their coping success. In addition, various behavioural tests and operant conditioning tests (to test a pig's motivation to change its actual environment) can be carried out. Under common conditions on the farm, problems associated with coping (like incidences of diseases, injuries, and stereotypies) and/or other relevant traits (e.g. saliva cortisol levels, longevity and even production traits) should be recorded routinely and used as selection index information. Selection for improved welfare should lead to more tolerant pigs that are better able to cope with possible unfavourable farm conditions by a more efficient use of the adaptation mechanisms they already possess. It should, however, not result in lowering husbandry standards. More research is needed to assess genetic correlations among various welfare aspects and with production traits to prevent undesired side effects in future populations of pigs.

Twenty-eight lactating ewes (mean 48 (s.e. 0·37) kg live weight) were used from days 12 to 52 after lambing to evaluate the effects on digestibility and production performance of replacing barley grain with citrus pulp in diets based on ammonia-treated barley straw. Concentrates included 0·82 to 0·83 of different barley to citrus pulp proportions: 100: 0 (T1); 66: 33 (T2); 33: 66 (T3) and 0: 100 (T4), 0·115 of soya-bean meal, and urea to make diets isonitrogenous. Ewes were adapted to a common diet for 11 days after lambing, and then were given 850 g/day of each experimental concentrate together with 850 g/day of chopped barley straw for 14 days and milk production and lamb growth were recorded. During the following 14 days ewes received the same amount of concentrate but the straw was offered ad libitum and straw intake was recorded as well as milk production and lamb growth. After completing both periods, four ewes per treatment were used for total collections of faeces and urine. Apparent digestibility of organic matter and neutral-detergent fibre increased linearly (P < 0·05) with increasing levels of citrus pulp although no differences were found in the digestible organic matter content of dry matter of diets because of differences in ash content. Urinary excretion of allantoin and purine derivatives per unit of digestible organic matter intake tended to decrease with increasing inclusion of citrus pulp (proportionately by 0·15), suggesting a decrease in microbial protein synthesis, though this effect was not significant (P > 0·05). Faecal excretion of purine bases also decreased (P < 0·05) as citrus pulp inclusion increased. No treatment effect (P > 0·05) on ewe live weight, milk composition or serum glucose and 3–OH butyrate was observed when a 1: 1 straw to concentrate ratio was given, but milk production and lamb daily gain decreased linearly (P < 0·05) with increased proportions of citrus pulp in the concentrate. Similar responses were detected when straw was given ad libitum and differences among treatments in terms of straw intake were not identified. It is concluded that a lower microbial protein flow might explain in part the reduction in milk production observed when barley was replaced with citrus pulp.

This study aimed to assess whether cows are able to cope with the range of climatic conditions they are exposed to in open stables on commercial farms in central Europe. On each of four farms, ten lactating cows were observed over a total of five weeks in winter, spring and summer. Based on continuous measurements of air temperature (–13·8 to 28·7ºC) and relative air humidity (0·26 to 0·99), a mean value of a temperature humidity index (THI) was calculated for each farm and each observation day for night and day.

THI had significant effects on skin temperature and body surface temperature (infra-red thermography) both during night and day. Rectal temperature, duration of lying and cortisol concentration in the milk was significantly affected by THI during the day but not during the night. Heart rate and frequency of lying did not significantly covary with THI. Differences between farms and interactions between THI and farm were significant for most parameters. These results suggest that the climatic conditions during the day induced stronger thermoregulatory responses than the conditions during the night. Within the measured range of climatic conditions the cows were hardly exposed to severe cold or heat stress.

In an experiment lasting from 1.5 years before first calving until third parturition, 24 purebred indigenous Bos indicus (Boran) cows and 24 Boran crosses with Holstein (proportionately 0.50 and 0.75 Holstein blood level) were given a diet composed of Bermuda grass hay and wheat bran (0.65: 0.35) offered at low, medium and high level. This level was adapted in amount to actual body weight every 2nd week corresponding to assumed 1.0, 1.2 and 1.4 times maintenance energy requirements. Cows were subjected to working exercise before and after first parturition. Body weight differences (lower with low and medium feeding level) developed mostly before calving for the first time and showed the typical decline at the start of lactation and increase in the dry period. Additionally, Boran cows given the high level increased their body weight from the first to the second lactation cycle. Body condition scores were higher in the Boran cows and, in both genotypes, with high feeding level. Independent of feeding level and genotype, calving intervals in cycles 1 and 2 were 530 and 421 days, respectively. Lactation length was considerably shorter in the Boran cows than in the crossbred cows. Milk yield, calculated over the first 13 weeks of lactation and over the whole lactation in both cycles, was 2.06 and 3.06 times higher in the crossbreds than in the Boran. Milk of Boran cows had 1.30, 1.15 and 1.20 times higher contents of fat, total solids and protein. In the first 13 weeks of lactation, milk yield of crossbreds with high feeding level (8.7 kg/day) was higher (P < 0.05) than that of the crossbreds with medium (6.5 kg/day) and low feeding level (5.4 kg/day), respectively. Boran cows never showed a significant response in milk yield to feeding level. Accordingly, the amount of organic matter intake required per kg milk increased with feeding level in the Boran cows while it remained unaffected in the crossbreds. Organic matter digestibility, as measured in lactation cycle 1, was higher (P < 0.01) in Boran. Response to high feeding level and estimated maintenance requirements of crossbred cows corresponded with current assumptions, but crossbreds subjected to medium feeding level expressed unexpectedly little difference from those receiving the low level. It seems that current recommendations are not generally applicable to indigenous tropical breeds as these responded differently from crossbreds to feeding level.

Twenty-nine Holstein-Friesian and 20 Norwegian dairy herd replacements were used to investigate the effect of habituating heifers to the milking parlour environment prior to calving on subsequent lactational performance. The heifers commenced the study at 3 weeks prior to calving when they were allocated on the basis of breed, genetic merit and live weight to either a habituation or no habituation (control) treatment. Heifers were housed together in cubicle accommodation with grass silage offered ad libitum along with 1 kg of concentrates per head per day offered in the feeding passage (control treatment) or in the food managers in the milking parlour (habituation treatment). Animals in the habituation treatment were exposed to the full milking parlour routine once daily (afternoon milking) in a 20-point rotary herringbone parlour. Holstein-Friesian heifers averaged 2.6 kg/day more milk than the Norwegian replacements. Holstein-Friesian heifers also had higher fat and protein yields than Norwegian heifers, however they also lost more body condition during the first 3 months of lactation. Habituated heifers yielded on average 1.3 kg/day more milk (P < 0.001) than the control group of heifers over the first 100 days of lactation (26.7 v. 25.4 (s.e.d. 0.38) kg/day), with the difference being greatest in the first 2 to 3 weeks of lactation. In early lactation, animals on the habituation treatment lost more live weight (0.16 v. 0.02 (s.e.d. 0.061) kg/day) (P < 0.05) and body condition than those on the control treatment. Duration of milking was longer (P < 0.001) (378.4 v. 340.5 (s.e.d. 6.53) s) and milk flow rate slower (P < 0.001) (2.20 v. 2.46 (s.e.d. 0.041) kg/min) respectively for the habituation compared with control group. Somatic cell counts (SCC) were lower (P < 0.001) for habituation group (1.66 v. 1.79 (s.e.d. 0.037) log10 SCC per ml), but there was no significant treatment effect on locomotion scores. Reproductive performance was lower for habituated heifers, with increased intervals to conception (P < 0.05) (102 v. 83 (s.e.d. 9.22) days). Habituating heifers to the milking parlour environment prior to calving increased milk production but appeared to have some detrimental effects on reproductive performance.

Sixty twin- and 60 single-bearing ewes were used in a 2 ✕ 2 ✕ 2 factorial design incorporating nutritional and shearing-cold exposure treatment (no. = 15 per subgroup) to study the effects on periparturient immunity against nematode infection. From week –8 to week –4 relative to parturition, animals grazed ryegrass-white clover pastures at allowances designed to provide either 0·8 (low allowance) or 1·2 (high allowance) of estimated metabolizable energy requirement for a ewe carrying 1·5 lambs. During week –4, half the sheep from each reproductive effort and nutritional group were shorn and exposed to artificial wind and rain for 4 h on each of 4 days consecutively. During the same week, all animals were challenged twice with 1·5 ✕ 104 Teladorsagia circumcincta and 1·5 ✕ 104 Trichostrongylus colubriformis infective third stage larvae to supplement the natural infection from pasture. From week –3 to week + 5 relative to parturition the sheep were run together and offered pasture at rates estimated to enable them to meet their nutrient requirement.

Shearing and cold stress reduced ewe body weight (by up to 4·66 kg per sheep), and temporarily reduced the levels of IgA against T. circumcincta (by proportionately 0·24) and T. colubriformis (0·34), and raised faecal egg count (1·6 to 4·4 fold) but did not affect the parasite-specific total antibody levels in the serum. The low allowance during pregnancy was associated with a 6·09-kg decrease in ewe body weight, decrease in the levels of total antibody and of IgA against both T. circumcincta and T. colubriformis, and increased faecal egg count (by 1·5 to 10·6 fold) during pregnancy and lactation. Faecal egg count was consistently higher (up to 8·8 fold) in twin-bearing and rearing ewes than in single-bearing and rearing ewes except during the first 4 weeks of the experiment. A significant effect of litter size on antibody levels was relatively small when compared with short-term changes (reductions) in antibody levels immediately around parturition. There were significant negative correlations between faecal egg count during late pregnancy and ewe body weight in lactation and significant negative correlations between the levels of total antibody against T. colubriformis and faecal egg counts near the end of the experiment. We conclude that litter size is likely to have greater influence on the resistance of ewes to nematode infections in the periparturient period than either pasture allowance or cold stress within the range of parameters used in this work.

The origin of duodenal purine bases (PB) was studied in a digestion experiment with four heifers, cannulated in the rumen and duodenum, which received a basal concentrate (152 g crude protein (CP) per kg dry matter (DM)) together with barley straw (85: 15 fresh weight basis) or the same concentrate supplemented with soya-bean meal, carbohydrate-treated soya-bean meal, maize gluten meal or fish meal to increase its protein content to 192 g/kg DM. Tr eatments were assigned to the four animals in five experimental periods according to an incomplete Latin-square design. Each 30-day period included 20 days of change-over adaptation and 10 days of experimental measurements. The flow of digesta entering the duodenum was estimated using Yb and acid-detergent insoluble ash as indigestible markers according to a double-marker system and microbial nitrogen (N) and PB were labelled with 15N infused into the rumen. The proportion of duodenal PB of microbial origin estimated from 15N enrichment of PB-N averaged 0·66 (s.e. 0·029) and did not differ between treatments nor when protozoa or bacteria associated with liquid (LAB) and solid (SAB) fractions were used as a reference sample. On average microbial contribution to duodenal non-ammonia N was higher when estimated from the PB/N ratio than from 15N (0·67 v. 0·55 (s.e. 0·015)) although differences were small and not significant when LAB was the reference sample (0·58 v. 0·52 (s.e. 0·018)) reflecting the higher PB/N ratio of this fraction compared with SAB and protozoa (2·04 v. 1·65 and 1·60 (s.e. 0·04) mmol/g). Considering only the duodenal PB of microbial origin resulted in estimates of microbial N synthesis from the PB/N ratio of SAB similar to those derived from 15N enrichment of both bacterial fractions (12·9 v. 13·5 and 13·3 (s.e. 0·83) g/kg of organic matter apparently digested in the rumen OMADR)) but underestimated the values derived from LAB (9·9 g/kg OMADR). Regardless of the estimation method, neither the duodenal flow of microbial N nor the efficiency of microbial synthesis differed between treatments. These results suggest that a significant proportion of duodenal PB have a non-microbial origin which may lead to overestimation of microbial yield when PB are used as a marker. Differences in PB/N ratio between microbial fractions is another important factor to be considered.

A comparative study of grazing behaviour, herbage intake and milk production was conducted using three strains of Holstein-Friesian (HF) heifer : 33 high production North American (HP), 33 high durability North American (HD) and 33 New Zealand (NZ) animals. Heifers were assigned, within strain, to one of three grass-based feeding systems : (1) the Moorepark (control) system (MP), (2) a high concentrate system (HC), (3) a high stocking rate system (HS). Strain of HF had no significant effect on grazing time or number of grazing bouts. The NZ strain had longer grazing bouts (P < 0.01) and spent a lower proportion of time ruminating (P < 0.05) than both the HP and HD strains. There was a significant strain ✕ feeding system interaction for biting rate. The biting rate of the NZ strain was reduced in the HC system. Biting rates in the HS feeding system were significantly higher (P < 0.001) than in the MP system. Heifers on HC had shorter grazing time (P < 0.01) with grazing bouts of shorter duration (P < 0.01). Increasing stocking rate (HS) decreased the proportion of time ruminating (P < 0.001) and tended to shorten grazing bouts (P = 0.06). The HP strain had higher (P < 0.05) herbage and total dry matter (DM) intakes than the NZ strain, while the HD strain was intermediate. Concentrate supplementation reduced (P < 0.001) herbage DM intake but increased (P < 0.001) total DM intake. The reduction of herbage DM intake per kg of concentrate DM intake (substitution rate) was greater for the NZ than the HP strain. The HP produced significantly higher milk, fat, protein and lactose yields than the NZ, while the HD strain was intermediate. The milk fat content of the NZ was higher than both the HP and HD strains, while the protein content was higher than the HP strain. Concentrate supplementation (HC v . MP) significantly increased yields of milk and milk components. Milk production responses to the HC system were much greater with the HP than the NZ strain. Increasing stocking rate (MP v . HS) significantly decreased milk protein yield. The results indicate that the choice of strain of HF may depend on the feeding system.

A naturally occurring zeolite (Anzymite™) was added to a diet containing 350 g maize silage per kg (which was treated with 10 g urea per kg (fresh weight)), 375 g barley and 275 g alfalfa hay per kg. Effect of inclusion of zeolite (0, 30 and 60 g/kg diet) on diet digestibility, ruminal fluid acidity and ammonia concentration, blood urea nitrogen level, and feedlot performance was studied in Mehraban ram lambs. The diets were given ad libitum. Digestibility coefficients of dietary dry matter and crude protein were significantly increased by zeolite (P < 0·05). The diet containing 30 g zeolite per kg had higher neutral-detergent fibre digestibility compared with the control (P < 0·05). Over all sampling times, the ruminal fluid of the sheep given the 30-g/kg zeolite diet had the highest and those given the 60-g/kg zeolite diet had the lowest pH values (P = 0·03). Before feeding, ruminal ammonia concentration was low for all treatments (4 to 8 mg/dl). At 4 h after feeding, the control diet had the lowest ruminal ammonia concentration (5·5 mg/dl) which was significantly lower than the values for zeolite diets (35 to 39 mg/dl). Blood urea nitrogen (BUN) level increased post feeding for all diets. At 4 h post feeding, the level for the control diet was significantly higher than for the zeolite diets, but at 6 h post feeding, the level of BUN was higher with 60 g zeolite per kg diet. Daily dry-matter intake for the 60 g zeolite per kg diet was significantly higher than for the control. Backfat depth in the 60 g zeolite group was significantly less than the control group. Pelvic and pericardial fats were significantly higher for the zeolite groups. Fat-tail weight in the 60 g/kg of zeolite group was significantly higher than in the 30 g zeolite and control groups. The findings indicated that, in spite of some improvements in digestibility and rumen fermentation pattern, addition of zeolite to the diet of feedlot Mehraban lambs, under the conditions of this experiment, was not advantageous.