The welfare of young pigs was compared in different housing systems using a range of indicators. Groups of twenty pigs were weaned at 6.4 kg into each of four pens built within environmentally-controlled rooms: a) deep-straw; b) Straw-Flow(R) (1 kg straw daily); c) large expanded metal-floored flatdeck; d) small flatdeck. Kennelled lying areas were provided in a and b. Stocking densities were 0.229 m2/pig for a,b and c, and 0.173 m2/pig for d. After 4-5 weeks (19.6 kg liveweight), sixteen pigs from each pen were moved into Straw-Flow grower pens (0.675 m2/pig) and observed until slaughter at 90.6 kg. The remaining pigs were grouped together to examine the effects of mixing. Housing type did not affect growth and feed intake. Food conversion ratio was poorer in large flatdeck than in small flatdeck or Straw-Flow. Behavioural differences between Straw-Flow and deep-straw related to differences in available straw; there were few differences between flatdecks. Access to straw affected pig-directed, pen-directed and straw-directed behaviours. Foot injuries recovered rapidly on deep-straw and increased in flatdecks immediately after weaning (Straw-Flow intermediate); injuries were more persistent in small flatdeck. Deep-straw prevented adventitious bursitis of the hock. Antibody response was examined by challenge with keyhole limpet haemocyanin (KLH). Immunoglobulin (Ig) M response was greater in straw pens. Dust concentration was higher in kennelled lying areas than flatdecks. With removal of kennels, straw pens showed less dust than flatdecks. Thus, welfare of early-weaned pigs in deep-straw was very good, flatdecks poor with Straw-Flow closer to deep-straw than to flatdecks. Early experience had limited effects. Pigs from flatdecks performed increased rooting relative to deep-straw two weeks after moving (Straw-Flow intermediate). These differences did not persist. Pigs mixed at 20 kg showed increased aggression (not significant) and decreased Ig m response to KLH, indicating reduced welfare.

Three studies were conducted on commercial farms. Study 1 sought to compare four combinations of natural mating (NM) and AI, namely AI + AI, AI + NM, NM + AI and NM + NM (first mating on day 1 of oestrous and second on day 2), and evaluate what aspects of insemination quality are related to fertility. The respective farrowing rates (FR) of AI + AI, AI + NS, NS + AI and NS + NS were (number of sows in parentheses) 92.9% (141), 89.6% (134), 89.3% (140) and 89.0% (100) and litter size (LS) was 11.31, 10.49, 11.50 and 10.61 (total births). The number of pigs born per 100 sows mated were 1051, 940, 1027 and 944. Both FR (P < 0.05) and LS (P < 0.01) were significantly influenced by parity. Sow behaviour during insemination significantly influenced (P < 0.01) FR and LS, the mean FR of calm and agitated sows being 91.3 and 83.1 percent and the mean LS being 11.32 and 10.22 respectively. Loss of semen due to backflow during insemination significantly reduced (P < 0.05) FR in sows having one AI. Backflow losses were significantly lower (P < 0.01) in calm than in agitated sows. It was concluded that AI, either alone or in combination with NM, yields fertility results comparable to NM alone and that sow behaviour during insemination has an important effect on sow fertility. Study 2 tested the effect of the addition of 5 i.u. of oxytocin to semen immediately before AI on fertility at two farms. Overall oxytoxin significantly improved FR (P < 0.05) from 73.2 percent (429) to 79.5 percent (385) and substantially improved litter size from 10.70 to 11.18. There was a significant (P < 0.05) treatment x farm interaction, FR on farm A improving from 77.2% (control) to 78.0% (oxytocin) whereas FR on farm B improved from 69.7% (control) to 84.0% (oxytocin). It was hypothesised that differences between farms in operator skill during insemination may have mediated the differing response to oxytocin.

The protein requirements and nutrition of bulls intensively reared on barley were investigated in the light of current knowledge of protein metabolism in the ruminant. Five experiments were carried out using cattle and sheep. 1. The response of barley beef bulls to RDP and UDP supplementation was examined. Although results were not significantly different, urea supplemented diets gave the highest growth rates of 1.45 and 1.47 kg/d; soya bean meal supplemented diets gave growth rates of 1.31 and 1.36 kg/d while the unsupplemented barley diet gave a growth rate of 1.23 kg/d. Feed conversion ratios ranged from 5.89 to 6.51. 2. The response of barley beef bulls to barleys of different protein content and degradability and to RDP and UDP supplementation was examined. Results were not significantly different although the high protein, low dg barley gave slightly lower liveweight gains (1.29 and 1.24 kg/d) than the low protein, high dg barley (1.32, 1.33 and 1.34 kg/d). RDP and UDP supplementation gave no additional growth response on any of the diets. Feed conversion ratios were approximately the same as the first trial ranging from 6.08 to 6.65. 3. A study of the barley-beef diets used in the first bull trial was carried out using surgically modified sheep fitted with Ash-type re-entrant cannulae in the duodenum. Five of the six animals died shortly after surgery and it was concluded that the surgery, and the type of cannulae used were a major factor causing this. 4. A study of the barley-beef diets used in the second bull trial was carried out using surgically modified sheep fitted with Hecker-type re-entrant cannulae. 24 h duodenal digesta collections with Cr2O3 as a marker were carried out followed by 7 d nitrogen balance and digestibility trials. Results obtained were nullified by incomplete marker recovery and it was concluded that the Hecker-type of re-entrant cannulae may be unsuitable for total digesta collection. The low protein, high dg barley diets were more digestible than the other barley diets and it was suggested that this may have improved rate of microbial synthesis in the bulls fed on these diets. 5. A study was carried out to examine certain factors affecting the degradability of barley as determined by the nylon bag technique. Variety, CP content and nitrogen fertiliser treatment were studied. Variety had a significant effect on dg (0.15-0.42), malting varieties being more degradable. Degradability tended to decrease as nitrogen fertiliser level increased (0.06-0.23). There also seemed to be a general trend that as CP content increased, dg decreased (0.6-0.8). The results were discussed in relation to current knowledge of ruminant protein metabolism. The new ARC protein system was shown to be acceptable as a framework and the need for more practical data was emphasised.

The genetic outcome of operating an open nucleus system (ONS) in a population of 1200 sheep divided into individual flocks, was investigatd using simulation techniques, and a comparison to alternative breeding systems was made. Selection was for best linear unbiased prediction of breeding value (EBV) for a trait measurable on individuals before selection. Migration of animals between flocks was based on the EBV. The results of stochastic simulations indicated that, compared to selection within closed flocks (CF), the annual genetic gain could be increased by 25 to 40 per cent if an ONS was operated. The variation in selection response, and the rate of inbreeding were also lower. Simulation results showed that approximately 25 per cent of nucleus sires should be selected from the base flocks. Initial genetic differences between flocks had a short term effect on genetic gain in the ONS. Screening the population to ensure that the nucleus flock was initially composed of the best ewes, also only had a short term effect. The advantage of an ONS over CF selection was effected by flock size but not intitial heritability. The use of dispersed nucleus system and a sire referencing scheme with selection of sires across flocks, was investigated. The outcome of these systems was very similar to the ONS. The higher rate of genetic gain achieved in the ONS, compared to CF selection, was attributed to higher genetic selection differentials and maintenance of genetic variation within flocks. These effects were included in a deterministic model of the ONS. The model described the simulated data well, although changes in genetic variance at the outset of the ONS were underestimated. Failure to account for finite population effects in previous studies has resulted in underestimation of the benefits of an open nucleus system in small populations.

Chapter 1: This chapter is a review of literature covering different aspects of ruminant forage nutrition. In this review, aspects of the rumen and its microbial environment and the nutrition requirements of rumen microbes, plant cell wall structure and its degradation by rumen microbes and the regulation of feed intake and digesta flow in ruminants are covered. The last part deals with supplementation of low quality forage diets, with emphasis on the use of high quality forages as supplements and their effect on food intake. Chapter 2: The first part of this chapter describes the experimental methods and analytical techniques used, and the source and analysis of forages used in this study. Natural pasture (veld) hay harvested in the dry season was used as the poor quality forage. Napier hay, harvested at 4 weeks growth during the wet season, and groundnut hay, harvested when the crop was mature, were used as the supplementary high quality forages. Veld hay had a low nitrogen and high NDF content (3.8 and 848 g/kg DM respectively), while napier hay and groundnut had high nitrogen and low NDF content (22.1, 21.0 and 700, 455 g/kg DM respectively). These differences in nutritive value indicated by chemical composition were confirmed by in sacco dry matter degradation and in vitro gas production tests. Groundnut hay, despite having a high solubility and high rate of degradation, had the least degradable cell walls due to lignification. Chapter 3: Expts 1 and 2. The objective of these experiments was to test the response in intake and digestibility, to supplementation of veld hay with different levels of napier or groundnut hays when the basal diet was not supplemented with urea. With both napier hay and groundnut hay, supplementation increased total feed intake (P < 0.05). There was a tendency for the intake of veld hay to be decreased by high levels of forage supplement.

One of the main deterrents against group-housing systems of dry sows is the aggression between animals and the expensive designs necessary to prevent this. This is related to the low feeding level accepted in practice. The low level of feeding is also linked to the occurrence of stereotypies which are generally accepted as an indication of poor welfare. An ad libitum feeding system which does not result in excessive intake and concomitant problems of obesity, might improve sow welfare and have economical advantages in reducing capital cost. Suitable diets might be high-fibre diets. High inclusion levels of several fibrous materials failed to prevent excessive intake. Only unmolassed sugarbeet pulp gave acceptable results, when included at a level of 580 to 650 g/kg. Composition of the fibre appeared to be more important in regulating voluntary intake than fibre concentration of the diet per se. A diet containing a high level of sugarbeet pulp (SBP) had a long transit time and high nutrient digestibility. Nutrient digestibility was influenced by protein level and source. A SBP diet appeared to regulate intake by physical, physiological and metabolic mechanisms. The large volume of wet sugarbeet pulp will increase gastro-intestinal fill. The rate of glucose absorption and secretion of insulin was reduced. A high pre-feeding plasma VFA concentration indicated a more constant supply of nutrients for a longer period after feeding. A SBP diet increased feeding time four-fold. Animals receiving a SBP diet were less active and engaged to a lesser extent in post-feeding oral behaviours. A marked reduction in time spent rooting suggests that the sows were less food-motivated.

The interaction effects of genotype and nutrition on the performance of sows during lactation is still poorly understood; this series of experiments explored the factors contributing to breed differences in lactational performance and the development of appropriate feeding strategies. Two extreme types of sow were used in these experiments; lean pure-bred White sows (W: Large White and Landrace) and a prolific but fatter Meishan synthetic damline (M: 50% Meishan genes). M sows had higher numbers of piglets, a different feed intake curve and produced more milk with a higher fat content compared to W breed types. Their response to dietary protein was different for the two breeds; the Meishan sows, with higher initial levels of body fat reserves, used the extra protein to increase milk production. The White sows did not partition the extra protein into milk but used it to conserve their own maternal body reserves and improved their weaning to conception interval. The breed differences in sows performance were found not to be solely due to differences in litter size or piglet genotype or maternal body fat reserves. M sows showed better maternal behaviour with lower activity and shorter suckling intervals. It can be concluded from these experiments that Meishan synthetic sows are inherently better mothers, and the way in which sows of different breed types partition feed nutrients during lactation is very different. A dietary lysine concentration of 9 g/kg (giving daily intakes of 53-61 g/day at 98-111 MJ DE/day) optimised lactational and rebreeding performance in both breed types.

The study examined the economy of the Highlands and Islands of Scotland generally and that of its agricultural industry in particular. Enzootic Abortion of Ewes (EAE) was found to be the principal cause of ovine abortion in Scotland and the United Kingdom. In addition, the Highlands and Islands' area had certain advantages with respect to EAE that could be exploited. The literature review concluded that the production and marketing of high health status (HHS) breeding sheep offered development opportunities. An HHS breeding sheep research programme was designed and implemented. Firstly, surveys were carried out with producers in Caithness and in the Highlands and Islands as a whole. Secondly, the views of veterinary practices throughout Scotland were sought. Thirdly, the opinions of lowground breeding sheep farmers were surveyed in Grampian and the Borders of Scotland. It was found that a potential market existed for HHS breeding sheep accredited for EAE and vaccinated against the Clostridial Diseases and Pasteurella. The Highlands and Islands' area appeared to have advantages in this respect, especially that EAE lent itself to control by sheep health schemes and that the purchase of HHS replacements was believed to be the most reliable method of controlling EAE by the farmers, crofters and veterinary practices surveyed. The Highlands and Islands' Sheep Health Association (a co-operative) was found to be the market leader in this development area. A production and marketing strategy was proposed for HISHA containing guidelines for the prevention and control of EAE, identifying the productive capability of the Highlands and Islands and customer requirements for HHS breeding sheep. Objectives were proposed for HISHA to maintain and increase its membership, specifically to develop the market for HHS breeding sheep in collaboration with farmers, crofters and veterinary practices for the benefit of HHS sheep industry.

The current trends in the intensive production of growing/finishing pigs are to devise alternative systems which help to improve welfare by the provision of a malleable substrate, such as straw, in order that the pigs can carry out natural behaviour. This project examined the welfare of growing/finishing pigs on four treatments; a new alternative system called the Straw-Flow (c) (SF), which used roughly a 1/4 of the amount of straw of a traditional deep-straw bedded system (1.9 and 8 kg/d respectively), was compared to bare-concrete (BC), fully-slatted (FS) and deep-straw (DS) treatments using a multi-disciplinary study involving behaviour, physical health, productivity and physiology. The four pen treatments were all built within the same building and they all measured 4x2.7m. There were three replicates of entire male pigs which were randomly allocated to each treatment from approximately 28 to 89 kg. During daylight hours, the pigs on the straw-based treatments (ST), ie . the SF and the DS, spent approximately 26% of their time in straw-directed behaviour. Where there was no straw (NOST), ie , the BC and the FS, there was more inactivity (45% and 59% of time on the ST and NOST respectively, p< 0.001), behaviour directed towards the pen hardware (2% and 13% of time on the ST and NOST respectively, p&60 0.001), chewing penmates; (0.05% and 0.19% of time on the ST and NOST respectively, p< 0.07) and vacuum chewing (0.2% and 1.4% of time on the ST and NOST respectively, p< 0.001). These differences were thought to be due to a lack of suitable malleable substrate on the NOST treatments which caused a redirection in the exploratory and foraging behaviour of the pigs compared to the pigs from the ST treatments. However, there was more play on the ST compared to the NOST in the form of running and scampering (0.15% and 0.02% of time on the ST and NOST respectively, p< 0.05) and shoving and pushing penmates (2.9% and 1.8% of time on the ST and NOST respectively, p= 0.14).

The major objective of the study was to evaluate the nutritive value of the locally available feed resources as sources of nutrients for dairy cattle on smallholder farms on the slopes of Mt. Kilimanjaro, Tanzania. Chapter one of the thesis is concerned with the review of literature on the farming systems approach to research. A short review of literature on roughage intake by ruminants and the evaluation methods used in determining their nutritive value is provided. In chapter two, results from a preliminary feed valuation experiment are presented. Nutritive value for 77 different locally available feeds are presented and discussed. In chapter three an experiment using twenty five yearling bulls aged 1 to 1.5 years and weighing 117 - 209 kgs was carried out in an attempt to estimate voluntary dry matter (DM) intake, digestible DM intake, DM digestibility and growth rates fed on 18 feeds. The animals were randomly allocated into five groups of five animals each. Feeds were then randomly allocated to the five groups four period of 60 days in which DM intakes were measured. A digestion trial was carried out to measure the feed's apparent digestibility. Solid outflow rates in the rumen were determined using Cr-mordanted fibre. DM degradation characteristics of the feeds were determined by incubation in nylon bags in the rumen of 3 Boran steers fed on guatemala grass. Washing loss (A) was also determined. DM intakes varied from 2.2 for banana pseudostem to 4.67 kg/day for urea treated maize stover. Digestibility ranged from 549 for untreated maize stover to 767 g/kg DM for banana pseudostem and growth rate ranged from 72 for banana leaves to 279g/day for urea treated maize stover.