The primary cereal grain used in poultry feed worldwide is maize. An understanding of how and where broiler chickens digest native starches, this being the major energy component of maize, and the influence that exogenous xylanase has on this process, is essential when optimising nutrient utilisation of a diet. In this project, both in vitro and in vivo models were used to understand broiler digestion of starches in different maize-based diets: native maize starch, maize grit and maize-soy with or without xylanase (Econase XT 25 at 100 g/t; 18,100 BXU/kg). An in vitro model to simulate the digestive tract of broilers using a ball milling technique to mimic the chicken gizzard was designed to investigate differences in the digestion of starches between the various diets. Reduced starch crystallinity and particle size of the diet was shown when samples were ball milled in HCl-pepsin, compared to roller mixed samples, leading to a greater glucose release when these samples were incubated with pancreatin. The maize-soy diet gave the greatest glucose release with pancreatin, with native maize starch giving the lowest, indicating that the processing of diets prior to digestion as well as the surrounding food matrix can impact on the digestion of starches in feed. In all in vivo trials, starch fraction measurements and SEM imaging of digesta collected from the gizzard, upper and lower small intestine, revealed the progression of starch granule digestion through the tract. Despite birds being starved prior to feeding, broilers ingested little of the native maize starch diet. Nonetheless, predominantly intact granules were evident in the gizzard, with signs of amylase attack becoming apparent in the small intestine. Following starvation, starch granules were still observed throughout the tract of birds, suggesting the presence of resistant starches. Higher digesta pH values were observed when birds were fed a maize-soy diet which gave pH 3.56, 6.70, 7.70 for the gizzard, and upper and lower small intestine, respectively, compared to pH 2.47, 6.49, 7.34 for the maize grit diet, an effect that could potentially impact on digestive processes. In feeding trials with broilers, inclusion of xylanase into the maize-soy diet and feeding on a single occasion revealed no apparent difference in the matrices surrounding the starch granules in digesta. This suggested that xylanase may be working indirectly in vivo through some physiological change in the digestion mechanism rather than direct action on the feed. To test this hypothesis, birds were fed xylanase for different lengths of time before slaughter. Peptide YY concentration in the blood was higher during the first few weeks of supplementation, with longer periods of supplementation nulling this effect, suggesting that xylanase may be acting through a prebiotic mechanism. RT-qPCR results revealed a trend towards an increase in glucose transporter (GLUT2 and SGLT1) expression at 2 and 3 weeks of xylanase supplementation, respectively. This also correlated to an increase in glucose concentration in the blood from 3 weeks of xylanase inclusion, suggesting a greater absorption capacity of birds. These results may indicate a potential mechanism of xylanase action in maize-based diets, although further work is required to confirm these findings.

Pedigree dog welfare in the United Kingdom is a controversial subject. This PhD aimed to investigate small animal veterinary surgeons’ (VS), pedigree dog breeders’ (PDB) and owners’ perceptions surrounding canine health, their roles, information acquisition and the quality of online resources available to new dog owners. Via a mixed-methods approach VS and PDB were asked about their perceptions of pedigree dog welfare using online questionnaires. Focus groups were then used to explore these parties’ opinions on advising dog owners and buyers. Dog owners were administered an online questionnaire to explore drivers of dog acquisition and sourcing of information about canine health. A further sample of dog owners provided possible internet search terms they would use to find information on how to care for a new dog. These terms were used for a scoping review to examine the quality of available information. The VS and PDB exhibited varied perceptions relating to pedigree dog breeding, inherited disease and health screening programmes. Each group considered themselves the most knowledgeable on these topics and lacked clarity on how their roles interrelated. Focus groups showed conflicting views on advising dog owners and buyers. Dog owners varied in their drivers for ownership, from whom they obtained information and the internet search terms used when looking for information online. The scoping review identified a lot of easy to understand, poor quality information. VS and PDB need to better understand and appreciate each others’ roles, perspectives and knowledge to instigate effective collaboration and improve pedigree dog health. Owners are diverse and require a multifaceted educational approach to improve knowledge. The quality of online information for owners needs to be made clear and a coordinated approach involving all stakeholders is required to enable dissemination of practical, evidence-based information on pedigree dog health.

Little has previously been described about how or why owners, veterinary nurses or veterinary surgeons make decisions about pets under their care. The Animal Welfare Act (2006) and the Royal College of Veterinary Surgeons’ oath should ensure the health and welfare of pets in the United Kingdom (UK) is the central focus of those decisions. The aim of this thesis was to characterise the nature and basis of decisions made about the treatment and welfare of osteoarthritic dogs by owners, veterinary surgeons and veterinary nurses. Three studies were performed. Firstly, two rapid reviews identified and appraised the outcome measures used in the peer-reviewed literature to assess canine quality of life and canine osteoarthritis. Secondly, thirty-two interviews were performed with owners of dogs with osteoarthritis and five focus groups were performed with veterinary surgeons and veterinary nurses who manage osteoarthritic dogs. Thematic analysis performed on transcripts of those interviews and focus groups identified key themes. Thirdly, a prospective study was performed to test a novel home monitoring outcome measure for use by owners of osteoarthritic dogs, developed using data gathered in the previous studies. The first study found outcomes assessed in the peer-reviewed literature focus predominantly on physical health. Those assessments are frequently unvalidated, may be subject to bias and neglect other aspects of the dogs’ welfare impacted by osteoarthritis and its management. The second study identified four important themes in the interviews and focus group data in relation to decision making. Most owners were highly motivated to make good decisions about their dogs’ welfare. However, many barriers to dog-focused decisions were recognised including: incorrect prior knowledge; ineffective veterinary consultations, in part due to different language used by owners and veterinary surgeons; the lack of available, relevant evidence on which to base decisions; an inability to reliably interpret canine behaviour; and risk aversion. A wide range of impacts of canine osteoarthritis on the welfare of the dogs, their owners and the veterinary professionals caring for those dogs were described. The third study identified several significant deficits in existing outcome measures designed for owners to assess their osteoarthritic dogs. Dogs with osteoarthritis may have day-to-day variations in their physical health and demeanour and owners appear to assess a complex mix of inputs to, and indicators of, their dogs’ welfare when decision making. More work is needed to develop outcome measures that are relevant to owners and more accurately reflect all aspects of canine welfare. This thesis is the first in-depth body of work using evidence synthesis and qualitative methods to characterise how decisions are made about osteoarthritic dogs under veterinary care. Most decisions about osteoarthritic dogs are made by owners using unvalidated assessments with little veterinary guidance. Valid, relevant and practical outcome measures are needed to collect information on which decisions can be based. Evidence does not exist to guide the majority of decisions made; the evidence that does exist appears to be poorly disseminated, particularly amongst owners. Relevant evidence must be created through well designed clinical trials to support those decisions then widely disseminated. Veterinary consultations are not always effective in making decisions focused on the best interests of osteoarthritic dogs, particularly in relation to their welfare; differences in language and perspectives may play a significant part in this. Future work in this field should involve collaboration between owners, veterinary professionals in general and specialist practice, animal welfare scientists and experts in dog behaviour. Methodological approaches taken and conclusions drawn from this thesis may be relevant to many other veterinary diseases.

Worldwide, there is an increasing demand for food, especially meat and milk. Alongside concerns around sustainability and other issues such as animal welfare, high expectations are put on livestock systems for an increased production and efficiency in order to meet such a demand. Helminth infections are ubiquitous on cattle farms and represent a growing concern for the industry around the world. In the UK, Ostertagia ostertagi and Fasciola hepatica are recognised as parasites of major importance in terms of their economic impact on cattle production and animal welfare. Rumen fluke is another strong candidate in the list of helminths which represent challenges for the sector in the UK. However, because helminth infections are mainly subclinical, their control is often very difficult. In this context, farmers generally adopt blanket treatment in young-stock to prevent or regain production losses due to these infections. This results in increasing problems of helminth resistance to available drugs, making such a practice unsustainable. Motivated by these concerns, several guidelines for best-practice on cattle helminth control have been published in the past few years. Nonetheless, farmers’ have been reluctant to adopt the recommendations put forward by these documents. Cattle helminths infections are influenced by the interplay of a wide range of factors. These include not only interactions between different species of parasites, but also climate conditions, management practices, availability of resources, and farmers’ attitudes, for which the role of comprehensive and reliable epidemiological information is key. An alternative to the use of cattle anthelmintic drugs is to avoid contamination of pasture to prevent the exposure of most susceptible cattle. The choice of diagnostic tools and the design of the studies are determinant for capturing the complexity of factors influencing helminth infections and control. However, basic epidemiological information on helminth infections in cattle in England and Wales is currently lacking, especially for O. ostertagi, F. hepatica and rumen fluke. Second, the relationship between economic losses and helminth infections remains to be clarified, particularly in the case of poly-infections and first lactation heifers. Third, previous studies informing potential alternatives (e.g. grazing management), suffer from limitations in terms of their scope and the adequacy of their recommendations. Finally, although being as relevant as epidemiological information, understanding what are the factors driving farmers’ decisions on cattle helminth control is a topic still poorly addressed in the literature. To address the issues above, this project was based on a mix-methods research (quantitative and qualitative methods) and a multidisciplinary framework that incorporates both veterinary epidemiology and sociology. The research analyses the cases of dairy and beef cattle in England and Wales by using longitudinal and cross-sectional studies, respectively. For dairy cattle, 43 farms (1,500 heifers) were studied. Data was collected and analysed in relation to the prevalence of O. ostertagi and F. hepatica; current practices in helminth control; demographics and management risk factors of young-stock helminth infections; impacts on milk production, reproduction and health performances in heifers; and farmers’ attitudes. As for beef cattle, data was collected for both single and poly-infections in 974 cattle (at slaughter), to support the analysis of the prevalence of O. ostertagi, F. hepatica and rumen fluke; demographic risk factors; and impacts on prime beef carcase performance. The main findings of this thesis are summarised below: Prevalence: the ubiquity of O. ostertagi and the significant presence of F. hepatica infections are confirmed; rumen fluke infections, most probably by C. daubneyi, can be considered as well-established in the UK; poly-infections by the three parasites are very common within the sample analysed; Risk factors: different types of grazing management practices can help with reducing dairy heifer exposure to O. ostertagi on pasture at specific times during their first years of grazing; Impact: the three parasites were significantly associated with low carcase performance in prime beef cattle; heifer exposure to O. ostertagi was significantly associated with lower milk production, reproduction and health performances; Farmers’ practices and attitudes: dairy farmers tend to overuse anthelmintic drugs on young-stock; they actively search and access information on cattle helminth infections and critically assess it in terms of management and business sustainability; farmers’ epistemology and contextual challenges should be taken into account while developing guidelines for helminth control. This thesis makes several contributions to veterinary and sociological studies of cattle helminth control. The different studies conducted shed light on a series of overlooked epidemiological and behavioural aspects that are critical for helminth control in the UK. Importantly, the thesis contributes to a better understanding of the complexity that is inherent to cattle helminth control. By considering both the epidemiology of the infections and broader societal and cultural factors, it offers a comprehensive analysis and a pioneer representation of how the system of cattle helminth control might operate in the UK. The results of this research are extremely valuable to veterinarians, farmers, experts, and policy-makers who all wish to develop and implement sustainable control of helminth infections in cattle.

Satisfactory investigations of the equine foot appear to be limited by the histo-morphological complexity of internal hoof structures. Foot lameness is considered to be one of the most debilitating pathological diseases in the horse. In most species, foot lameness is traditionally linked to hoof deformity, and a set of molecular events have been defined in relation to the disease. The differential capsular growth rate is essential in order to understand the aetiology of hoof distortions. It is now well established that keratinocyte progenitor cells grow from the papillae and not from the lamellae. The importance of hoof material synthesis, and therefore the underlying cell proliferation from the papillae, in the pathology of laminitis has been underlined in different species. Despite this knowledge, it remains unclear how cell proliferation itself is involved in this pathology. So far, there is controversy regarding the incidence of foot lameness in horses, as it is unclear whether it is foot lameness that triggers hoof distortions or vice-versa. In addition, very little scientific evidence exists to prove either point; this requires a deeper understanding of the physical biology and biomechanics of the hoof given that the horse stands on its hooves. It is remarkable that very little physics has been introduced to enhance our understanding of the disease. In order to develop a better understanding of foot lameness, we aim to provide a model of foot lameness encompassing anatomy, cell biology and physics to help reveal new evidence for evaluating the incidence of pathological changes within internal structures of the animal’s foot. It is anticipated that the results found will aid in bridging this gap in knowledge. Fifty-four equine hooves were collected and categorized based upon their morphology, and tissue sections were harvested from the periople, coronary and proximal part of lamellae. Markers of progenecity (p63), cell proliferation (PCNA, Ki-67) and differentiation (K14, K10) were used to assess the hoof growth rate in coronary regions using immunohistochemistry (PCNA, Ki-67, p63) and immunofluorescence (p63 and K14). For assessing the involvement of physical forces in hoof growth, the Micro CT-scans of individual foot were analysed to highlight differences in the ultrastructure of the feet, for example differences in dorsal curvature between different foot samples. The results presented in this study confirm previous findings, add evidence that a high percentage of Ki-67 and p63 positive cells are found in the proximal part of the papillae, and seem to contribute to hoof deformities. This evidence might be attributed to physical growth forces present in the proximal part of the hoof whereby the change in cell size might contribute to force generation, tissue remodelling (i.e. cornification) and hoof deformities. As hoof distortion can result from uncontrolled epidermal basal cell proliferation from the epidermal papillae. A model of cultured keratinocytes under different stresses such as stretch, hydrostatic pressure and/or low oxygen conditions. This revealed that these stresses together with low oxygen conditions stimulate cell differentiation, and upregulate keratin levels. Consequently, this leading to changes in their morphology and cell surface area. Together these results suggests that cell culture, growth, passage and physical or biological treatments have improved our knowledge of natural cell stimulation and aid our understanding of how equine progenitor keratinocytes (EPKs) would hypothetically respond to external stimuli other than biological ones. The results presented regarding physical forces showed that there is a strong correlation between the dorsal hoof curvature and the level of Ki-67 and p63 abundance at the dermal papillae. In addition, K14 and TUNEL apoptotic cells localized along the dermal papillae in the coronary hoof region were also found in ‘dished’ hooves. Specifically, when the hoof curvature is positive (i.e. for the ‘dished’ hoof) there is an increase in the mitotic index shown by Ki-67 and p63 expression in the quarter regions compared to the dorsal one. Negative hoof curvatures (i.e. having a claw-like shape), however, did not shown any correlation with Ki-67 and p63 expression in any tissue type, except for K14 expression levels in the dorsal and quarter regions. These outcomes demonstrated that the two proteins (p63 and Ki-67) participate in hoof growth, and very probably to the differential growth of the hoof capsule as observed in chronic laminitis. This is probably through a contribution of sustained level of epidermal basal cell proliferation. In conclusion, the morphometric data of the change in cell size and the localisation of proliferation cells on the proximal part of papillae provide evidence on the way of hoof morphogenesis. These were confirmed with behaviour of cultured keratinocytes under different stresses. Additionally, the difference in the proliferation rates between the dorsal, quarter regions of the hoof wall are connected to hoof curvatures. The data gives new insights into the physical biology of the normal versus pathologically affected equine feet, and bovine and ovine feet in the future, providing new information for potential treatment(s). Consequently, this research aims to bridge gaps across scientific fields to ameliorate our understanding of the equine foot. However, further studies need to be carried out in order to understand the pathways of cellular response at the coronary region.

Lameness is one of the most significant challenges faced by the dairy industry worldwide. The negative impacts on the health, welfare and productivity of dairy cows means that controlling lameness is important for future sustainability. The aims of this thesis were to evaluate the risk factors for lameness in dairy cows, in particular body condition score (BCS) and early life events. Claw horn disruption lesions (CHDLs) are one of the major causes of lameness in dairy cattle, however their aetiopathogenesis is still poorly understood. The primary hypotheses explored within this thesis were related to risk factors for CHDLs in particular, with the aim of progressing the current knowledge and understanding of these lesions. Data obtained over an 8 year period (2003 to 2011) from the Crichton Royal research herd held at the Scotland’s Rural College (SRUC) Dairy Research and Innovation Centre in Dumfries, Scotland were used to investigate risk factors for lameness in dairy cows. The associations between fine thresholds of body condition score (independent of body weight) and lameness were explored using mixed effects multinomial and binomial logistic regression models. Cows with BCS < 2 in the previous 3-weeks were at greatest risk of mild or severe lameness. Low BCS was found to be associated with an increased risk of repeated lifetime lameness events in all animals, as well as first lifetime lameness events in cows in their second lactation or greater, indicating that low BCS may be an important risk factor for lameness in herds managed under similar conditions to the study herd. Furthermore, a novel approach to estimating population attributable fractions (PAF) for lameness risk factors which used simulation to investigate complex changes in herd BCS and previous lameness events over time in two dairy herds, showed that a loss in BCS of 0.5 across the whole herd may contribute towards between 4 and 11% of the total number of lameness events in these herds. Despite differences in methods of lameness identification (weekly locomotion scores versus treatment of clinical cases) and the environment between the herds investigated, the estimated PAF were similar, giving some indication that findings may be to some extent generalisable. In comparison to BCS, previous lameness events appeared to have a potentially larger impact on the total number of lameness events in the herd. The median estimated PAF for all previous lameness was between 79 and 83% in the two herds i.e. approximately 80% of lameness events in these two herds may be avoidable by preventing all previous lameness events. Between 9 and 21% of total lameness events may be attributable to previous lameness events occurring > 16 weeks before a risk period. These findings indicate that repeated lameness events (i.e. accumulation of lameness) may contribute towards a large proportion of herd level lameness. Preventing or increasing the time to first lameness events in a dairy herd may be important to reduce the number of repeat cases and lessen overall burden of lameness. Interactions with environmental and/or animal-based risk factors may also be important for preventing repeat lameness events. Further research is required to identify the reasons for and methods of preventing repeated lameness events, which could have considerable impact on reducing lameness at the herd level. The development of a method for estimating PAF for lameness risk factors using simulation enabled longitudinal data to be analysed, taking into account repeated measures and changes in risk factors over time. This is an important step forward in evaluating risk factors for lameness in dairy cattle, by identifying those that will have greatest impact on controlling disease in the population. Further work is needed to estimate the PAF for other risk factors including environmental factors. The large effect of previous lameness events on future lameness highlights the potential importance of lameness in heifers, as the future of the dairy herd. Mixed effects logistic regression models and linear regression were used to investigate the impact of hoof lesions in heifers around the time of first calving on milk yield, future lameness and culling risk. Severe claw horn disruption lesions (CHDL) post-calving were associated with an increased risk of future lameness and reduced milk yield (average daily and adjusted for the duration of time that cows spent within the herd), indicating that managing heifers to reduce severe CHDL could have positive benefits on health, welfare and production for the dairy herd. Mild CHDL pre-calving and 2 to 4 months post-calving were found to be associated with a decreased risk of future lameness and decreased culling risk. This novel finding suggests that some degree of mild insult may confer some resilience, if occurring at a time when heifers are able to cope, potentially through adaptive changes in the hoof. This suggests that management of heifers could have major impacts on lifetime claw health. The impact of previous lameness on herd level lameness was also investigated through exploring how different culling rates effect lameness prevalence, with the use of simulation. As herds with a lower culling rate will have an older age profile this increases the duration of time over which repeat lameness events might occur. Herds adopting a culling rate of 40% versus 20% were found to have a relative reduction in lameness prevalence of 29.2% (38.2% versus 27.1%, respectively). In herds with lower culling rates particular attention to lameness prevention and control may be necessary to maintain the health, welfare and productivity of the herd; this would be especially true for herds with a high background risk of lameness in month 1 of each lactation. Multilevel statistical modelling and simulation-based methods, combined with a rich longitudinal dataset, have aided in the identification of important key findings that contribute towards advancing our knowledge and understanding surrounding the aetiopathogenesis of lameness, and in particular CHDL. Identifying risk factors that, if controlled, will have the greatest effect on reducing lameness at the herd level is important and this work highlights where future research can build on this so that progress can be made in controlling lameness. This is relevant and important for the dairy industry worldwide.

Great apes housed in zoological collections have an important role to play in conservation. A sound understanding about their health and welfare forms a critical part of their custodianship. Chapter 1 of this thesis outlines a systematic review of 189 published articles relating to the topic of great ape morbidity and mortality (Strong et al. 2016). It concluded that there was a critical need for an up-to-date review of zoo-housed great ape mortality, especially among the European population, to be carried out. Such a review of data relating to 681 great ape deaths was therefore performed and is outlined in Chapter 2 of this thesis. This mortality review identified the main causes of death within each taxa and age group, and allowed for a series of recommendations about future disease investigation and monitoring to be generated. Diseases of the cardiovascular system specifically, were identified as being associated with significant proportional mortality. Despite this, however, understanding about the epidemiology, pathogenesis and diagnosis of cardiovascular disorders among great apes remains poor. The remainder of the thesis therefore outlines a series of further projects and studies designed to confront this lack of knowledge and understanding: Chapter 3 focuses on cardiovascular disease epidemiology and identifies similarities and differences in disease risk between the taxa, highlighting age and male sex as potential risk factors. Chapter 4 is dedicated to the development of two protocols designed to standardise both the ante- and post-mortem investigation of cardiovascular disease in great apes. Chapter 5 addresses the controversial topic of carrying out cardiovascular disease screening in immobilised animals by comparing the effects of two anaesthetic protocols. Finally, Chapter 6 outlines a detailed study of great ape cardiovascular pathology and specifically idiopathic myocardial fibrosis in chimpanzees. The findings of each of the studies outlined in this thesis are informative, not only for the day-to-day management of zoo-housed great apes, but also for future research into their health, disease and therefore welfare.

Ovine footrot is the major cause of lameness in sheep within the UK and an economic and welfare concern for sheep farmers globally. It is characterised by under-running of the hoof-horn and preceded by interdigital dermatitis (ID). Under-running footrot is attributed to the host immune response, which is provoked primarily by Dichelobacter nodosus in addition to other bacteria pathogens. However, the precise role of these other bacteria is yet unknown. Therefore, we hypothesise that bacterial invasion and colonisation of the ovine interdigital skin contributes to a strong host inflammatory response leading to the characteristic histopathology observed. In this context, this study examined host inflammatory response [inflammatory cell infiltration, pro-inflammatory cytokines (IL-1β)], histopathological lesions and virulent D. nodosus abundance in healthy, ID and footrot conditions in an attempt to gain further insights into the pathogenesis of this important disease. To investigate this hypothesis, two studies were designed: (i) to grade histological lesions in different clinical conditions and (ii) to determine bacterial localisation in post-slaughtered interdigital skin biopsies from the abattoir. Standardised histology lesion grading systems were developed and applied using histochemical techniques (haematoxylin and eosin (H&E), periodic acid Schiff PAS). Bacterial localisation was determined in serial horizontal sections across skin depths combining histology (transverse cryosections + H&E) and qPCR technique for the quantification of bacterial DNA. Furthermore, parallel data of IL-1β expression and virulent D. nodosus load obtained from a different study were compared to histology lesions. Key findings were as follows: (i) histological lesions (cell ballooning, parakeratosis, epidermal micro-abscesses and inflammatory cell infiltration) were similar in all clinical conditions, (ii) increased inflammatory cell infiltration score corresponded significantly with high levels of IL-1β expression (p < 0.05) in footrot, and virulent D. nodosus load (p < 0.001) across all clinical conditions, (iii) across different skin depths, eubacteria localisation was consistent, D. nodosus localisation was highly variable while F. necrophorum was localised in deeper sections of healthy feet. In addition, eubacteria load was significantly higher (p=0.0002) in the epidermis near the skin surface (≤200μm) of footrot disease samples when compared to healthy samples. Eubacteria components may play contributory roles in footrot pathogenesis based on their localisation in interdigital skin. In conclusion, contrary to previous notion that the severity of disease condition was dictated by progressive pathology, data in this study showed no appreciable difference in the levels of histological lesions and inflammatory response between healthy and diseased (ID, footrot) conditions. Histological lesions and the bacterial components of the skin including the virulent D. nodosus contribute to the local inflammatory response which probably drives the progression of footrot disease.

Studies across different species have demonstrated that maternal stress during pregnancy can affect offspring development and welfare, both at the physiological and behavioural level. Throughout pregnancy, domestic livestock may experience various stressors that can inhibit health and productivity, and consequently impact on the development of the offspring. Several of these relate to the gestational social environment, a subject that has not received much attention from a prenatal stress perspective. In the UK, approximately half of ewes are housed prior to lambing. As a consequence, factors such as social instability, high stocking density, dog exposure and unpredictable feeding schedule have been identified as potential stressors for housed sheep. Some of these impose a degree of competition for resources while others expose the pregnant ewes to an uncertain and unpredictable environment. This project aimed to investigate the potential effect of these aspects of ewe husbandry when applied during pregnancy on lamb behaviour and welfare. The first experiment investigated the effects of the legally maximum stocking density and social mixing during pregnancy on the offspring. The second experiment looked at the effects of three different levels of maternal stress on the offspring, which included a higher stress environment where ewes were exposed to a dog, the feeding schedule was unpredictable and stocking density was the maximum allowed legally; a medium stress environment where there was no dog exposure and feeding schedule was predictable but stocking density was kept the same; and a lower stress environment with lower stocking density, ad libitum feed and no dog exposure. Lambs were followed from birth to slaughter and the effects of these potential maternal stressors on lamb behaviour and factors related to survival, immunity, growth, and epigenetic regulation of neuroendocrine stress axis function were investigated. The maternal stressors had very subtle impact on lamb behaviour and growth but birth weight and immunity were not affected. However, effects of prenatal treatment were found on brain morphology and factors regulating stress axis activity. Therefore there seems to be no major implications for lamb welfare of these maternal stressors. Hence, from a farm management perspective this study suggests that, in general, current ewe management practices in the UK are not detrimental for lamb welfare.

Foot-and-mouth disease (FMD) is a highly infectious disease affecting cloven-hoofed ruminants. FMD is endemic across Asia, Africa and South America and, as demonstrated by the 2001 outbreak in the UK, can cause devastating epidemics in FMD free countries. A comprehensive dataset comprising the births, deaths and farm-to-farm movements of almost fifty million cattle was made available to us by the Turkish authorities. In chapter 2 we discuss the processing, cleaning and analysis of this dataset. Cattle movement networks were built using farms as nodes and the farm-to-farm movements as edges before performing a network analysis. The richness and completeness of the Turkish data set is not generally available to modellers. In chapter 3 we implement various sampling strategies to determine the quantity of network data required to give accurate epidemiological predictions, using the British cattle trade network as a case study. In chapter 4 we introduce community detection as a method for analysing the cattle movement networks of both the UK and Turkey. This analysis goes beyond the calculation of the farm level statistics calculated in chapter 2 and seeks to find higher order structures in these networks. Chapter 5 develops a within farm model of FMD by drawing on outbreak data from two farms of different sizes in Turkey. This model is parametrised by running a Bayesian inference scheme on the smaller of the two farms. Simulations were then performed on both farms using the inferred posterior parameter distributions.