This thesis describes a series of three experiments designed to estimate the nutritional cost of the immune response to the gastrointestinal nematodes Trichostrongylus colubriformis and Teladorsagia circumcincta in sheep. For each experiment, animals were allocated hierarchically by liveweight into one of four groups that were either infected (group IF), similarly infected and concurrently immuno-suppressed with weekly intramuscular injections of 1.3mg kg liveweight (LW)⁻¹ of methylprednisolone acetate (group ISIF), immunosuppressed only (group IS) or remained as controls (group C). Body composition of all animals was estimated using x-ray computer tomography prior to infection and at the conclusion of each study with bodyweight and faecal nematode egg counts (FEC; eggs gram⁻¹ of fresh faeces (epg)) measured along with blood samples taken for the determination of levels of serum proteins, phosphate and antibodies. In the first trial (Chapter 3), the nutritional cost of both the acquisition and maintenance of immunity to gastro-intestinal nematodes was investigated using immunologically naive 5-month-old lambs and immunologically competent 17-month-old ewes during infection with 2,000 and 4,000 L3 infective T. colubriformis larvae d⁻¹, respectively (80 L3 T. colubriformis larvae kgLW⁻¹ d⁻¹). Profiles of FEC and comparative worm burdens at slaughter indicated an effective immune response was maintained in IF ewes and developed in IF lambs while successfully suppressed in both ISIF lambs and ISIF ewes and was confirmed by serum antibody titres. The typical reduction in voluntary feed intake as a consequence of infection was observed in IF lambs (0.30, p<0.001) but not in IF ewes, ISIF lambs or ISIF ewes, and appeared to be associated with L3 IgA. Gross efficiency of use of metabolizable energy (ME) for net energy (NE) deposition was reduced by 0.20 in lambs during acquisition of immunity and by 0.16 in ewes maintaining an established immunity. Infection in immuno-suppressed animals reduced efficiency by 0.05 and 0.15 for lambs and ewes. These findings allowed the hypothesis that the reduction in feed intake and nutrient utilization in young parasitized sheep is caused by physiological signalling associated with the acquisition phase of the host immune response to infection, rather than simply the damage caused by the parasite per se. The second trial (Chapter 4) investigated the influence of metabolizable protein (MP) supply on the metabolic disturbances associated with the acquisition phase of the immune response during infection with 2,000 L3 T. colubriformis d⁻¹. Groups of lambs were offered either a low protein (L; 62g MP kgDM⁻¹) or high protein diet (H; 95g MP kgDM⁻¹). Patterns of total daily egg excretion indicated that an effective immune response was developed in HIF, but not LIF, HISIF nor LISF and was confirmed by comparative worm burdens. The proportionate reduction in feed intake in immunologically normal animals was reduced through the provision of additional protein, being 0.12 in HIF and 0.23 in LIF. Regardless of diet, infection did not cause a reduction in feed intake in immuno-suppressed animals (p>0.05). Infection proportionately reduced the gross efficiency of ME utilization in immunologically normal animals by 0.23 in HIF (p=0.09) and by 0.51 in LIF (p=0.01), but not in immuno-suppressed animals. Immuno-suppression did not suppress serum L3 IgA levels in seven of the eight HISIF and four of the eight LISIF animals. Furthermore, only four out of the eight immunologically normal animals from both the HIF and LIF groups displayed an L3 IgA response. Consequently, regardless of immunosuppression treatment, animals were termed as IgA responders (HR or LR) or non-responders (HN or LN). Feed intake was proportionately reduced from day 22 by 0.15 in HR (p=0.03) and by 0.32 in LR (p=0.01), but was not significantly reduced in HN or LN. Gross efficiency of ME utilization was significantly reduced for LN animals only, being proportionately 0.59 (p<0.01). These findings allowed the conclusion that additional MP reduced the consequence of immunological signalling that was displayed in reduced feed intake and in nutrient utilization, both of which appeared to be associated with an IgA response. It is hypothesized that the lessening of nutritional disturbance observed in high protein and immuno-suppressed animals could be a consequence of altered physiological signalling during the immunological cascade. The third trial (Chapter 5) utilized lambs infected with the abomasal parasite T. circumcincta to explore the possibility that the reduction in feed intake and nutrient utilization is a universal phenomenon of the acquisition phase of the immune response to nematode parasites inhabiting different organs along the gastrointestinal tract. In addition, immunological changes at the site of parasite infestation in the abomasal mucosa were measured from serial biopsy tissue samples taken from a further twelve animals that were surgically fitted with an abomasal cannula and either infected (CIF) or concurrently infected and immuno-suppressed as described previously (CISIF). The development of immunity in IF animals was accompanied by a 0.17 proportional decrease in feed intake between days 15 to 28 of infection (p<0.05) and a 0.20 proportional reduction in nutrient utilization (p=0.07), none of which were observed in ISIF animals. While FEC and worm burdens indicated successful immunosuppression in ISIF animals, both serum IgA and total antibody production were not reduced. The development of immunity in CIF was reflected in an increase in both mast cells and globule leukocytes in serial abomasal tissue biopsies, both of which were reduced in CISIF (p<0.01 for both). In serial biopsy tissue, immuno-suppression did prevent a rise in tissue IgA that was apparent in CIF animals (p<0.01) although these changes were not reflected in serum IgA levels. It appears that the alleviation of the reduction in feed intake and nutrient utilization in young lambs through the use of corticosteroid induced immuno-suppression may be a universal phenomenon for both intestinal and abomasal parasites, but the association with and/or role of IgA during infection with T. circumcincta is unclear. In summary, the reduction in feed intake and nutrient utilization in sheep during infection with both the abomasal nematode T. circumcincta and the small intestine nematode T. colubriformis appears to be associated with a component(s) of the acquisition phase of the host immune response, rather than, as conventionally assumed, the direct mechanical damage of the parasite per se. It is hypothesised that the nutritional disturbance as a consequence of infection in young lambs may be the result of pro-inflammatory cytokines involved in immunological signalling that may also be associated with the production of IgA, the effects of which can be reduced through the provision of adequate MP. These studies provide evidence that the immune response to gastrointestinal parasites is nutritionally costly to the animal and have implications for application of manipulations that are intended to promote the development of a strong immune reaction in high producing animals.

This thesis describes a series of three experiments designed to estimate the nutritional cost of the immune response to the gastrointestinal nematodes Trichostrongylus colubriformis and Teladorsagia circumcincta in sheep. For each experiment, animals were allocated hierarchically by liveweight into one of four groups that were either infected (group IF), similarly infected and concurrently immuno-suppressed with weekly intramuscular injections of 1.3mg kg liveweight (LW)⁻¹ of methylprednisolone acetate (group ISIF), immunosuppressed only (group IS) or remained as controls (group C). Body composition of all animals was estimated using x-ray computer tomography prior to infection and at the conclusion of each study with bodyweight and faecal nematode egg counts (FEC; eggs gram⁻¹ of fresh faeces (epg)) measured along with blood samples taken for the determination of levels of serum proteins, phosphate and antibodies. In the first trial (Chapter 3), the nutritional cost of both the acquisition and maintenance of immunity to gastro-intestinal nematodes was investigated using immunologically naive 5-month-old lambs and immunologically competent 17-month-old ewes during infection with 2,000 and 4,000 L3 infective T. colubriformis larvae d⁻¹, respectively (80 L3 T. colubriformis larvae kgLW⁻¹ d⁻¹). Profiles of FEC and comparative worm burdens at slaughter indicated an effective immune response was maintained in IF ewes and developed in IF lambs while successfully suppressed in both ISIF lambs and ISIF ewes and was confirmed by serum antibody titres. The typical reduction in voluntary feed intake as a consequence of infection was observed in IF lambs (0.30, p<0.001) but not in IF ewes, ISIF lambs or ISIF ewes, and appeared to be associated with L3 IgA. Gross efficiency of use of metabolizable energy (ME) for net energy (NE) deposition was reduced by 0.20 in lambs during acquisition of immunity and by 0.16 in ewes maintaining an established immunity. Infection in immuno-suppressed animals reduced efficiency by 0.05 and 0.15 for lambs and ewes. These findings allowed the hypothesis that the reduction in feed intake and nutrient utilization in young parasitized sheep is caused by physiological signalling associated with the acquisition phase of the host immune response to infection, rather than simply the damage caused by the parasite per se. The second trial (Chapter 4) investigated the influence of metabolizable protein (MP) supply on the metabolic disturbances associated with the acquisition phase of the immune response during infection with 2,000 L3 T. colubriformis d⁻¹. Groups of lambs were offered either a low protein (L; 62g MP kgDM⁻¹) or high protein diet (H; 95g MP kgDM⁻¹). Patterns of total daily egg excretion indicated that an effective immune response was developed in HIF, but not LIF, HISIF nor LISF and was confirmed by comparative worm burdens. The proportionate reduction in feed intake in immunologically normal animals was reduced through the provision of additional protein, being 0.12 in HIF and 0.23 in LIF. Regardless of diet, infection did not cause a reduction in feed intake in immuno-suppressed animals (p>0.05). Infection proportionately reduced the gross efficiency of ME utilization in immunologically normal animals by 0.23 in HIF (p=0.09) and by 0.51 in LIF (p=0.01), but not in immuno-suppressed animals. Immuno-suppression did not suppress serum L3 IgA levels in seven of the eight HISIF and four of the eight LISIF animals. Furthermore, only four out of the eight immunologically normal animals from both the HIF and LIF groups displayed an L3 IgA response. Consequently, regardless of immunosuppression treatment, animals were termed as IgA responders (HR or LR) or non-responders (HN or LN). Feed intake was proportionately reduced from day 22 by 0.15 in HR (p=0.03) and by 0.32 in LR (p=0.01), but was not significantly reduced in HN or LN. Gross efficiency of ME utilization was significantly reduced for LN animals only, being proportionately 0.59 (p<0.01). These findings allowed the conclusion that additional MP reduced the consequence of immunological signalling that was displayed in reduced feed intake and in nutrient utilization, both of which appeared to be associated with an IgA response. It is hypothesized that the lessening of nutritional disturbance observed in high protein and immuno-suppressed animals could be a consequence of altered physiological signalling during the immunological cascade. The third trial (Chapter 5) utilized lambs infected with the abomasal parasite T. circumcincta to explore the possibility that the reduction in feed intake and nutrient utilization is a universal phenomenon of the acquisition phase of the immune response to nematode parasites inhabiting different organs along the gastrointestinal tract. In addition, immunological changes at the site of parasite infestation in the abomasal mucosa were measured from serial biopsy tissue samples taken from a further twelve animals that were surgically fitted with an abomasal cannula and either infected (CIF) or concurrently infected and immuno-suppressed as described previously (CISIF). The development of immunity in IF animals was accompanied by a 0.17 proportional decrease in feed intake between days 15 to 28 of infection (p<0.05) and a 0.20 proportional reduction in nutrient utilization (p=0.07), none of which were observed in ISIF animals. While FEC and worm burdens indicated successful immunosuppression in ISIF animals, both serum IgA and total antibody production were not reduced. The development of immunity in CIF was reflected in an increase in both mast cells and globule leukocytes in serial abomasal tissue biopsies, both of which were reduced in CISIF (p<0.01 for both). In serial biopsy tissue, immuno-suppression did prevent a rise in tissue IgA that was apparent in CIF animals (p<0.01) although these changes were not reflected in serum IgA levels. It appears that the alleviation of the reduction in feed intake and nutrient utilization in young lambs through the use of corticosteroid induced immuno-suppression may be a universal phenomenon for both intestinal and abomasal parasites, but the association with and/or role of IgA during infection with T. circumcincta is unclear. In summary, the reduction in feed intake and nutrient utilization in sheep during infection with both the abomasal nematode T. circumcincta and the small intestine nematode T. colubriformis appears to be associated with a component(s) of the acquisition phase of the host immune response, rather than, as conventionally assumed, the direct mechanical damage of the parasite per se. It is hypothesised that the nutritional disturbance as a consequence of infection in young lambs may be the result of pro-inflammatory cytokines involved in immunological signalling that may also be associated with the production of IgA, the effects of which can be reduced through the provision of adequate MP. These studies provide evidence that the immune response to gastrointestinal parasites is nutritionally costly to the animal and have implications for application of manipulations that are intended to promote the development of a strong immune reaction in high producing animals.

Animal health surveillance is an important part of animal health care, particularly in countries dependent on livestock for food production and international trade. There are two major issues related to the provision of e®ective surveillance activities. Firstly, for good information to become available, the design and conduct of data collection activ- ities should be carried out following sound statistical principles. In reality, constraints such as imperfect tests and unavoidably-biased sampling strategies hinder straightfor- ward analysis and interpretation of survey results. Risk-based surveillance is used to target high-risk sub-populations to increase e±ciency of disease detection; however, biased datasets are generated. This thesis develops methodologies to design risk-based surveillance systems and al- low statistically valid analysis of the inherently biased data they generate. The ¯rst example describes the development of a method to analyse surveillance data gathered for bovine spongiform encephalopathy (BSE). The data are collected from four dif- ferent surveillance streams of animals tested for BSE, with each stream containing unavoidable biases and limitations. In the BSurvE model, these data are combined with demographic information for each birth cohort to estimate the proportion of each birth cohort infected with BSE. The prevalence of BSE in a national herd can then be estimated using the method of moments, whereby the observed number of infected animals is equated with the expected number. The upper 95% con¯dence limit for the prevalence is estimated both for infected countries and for those where no BSE has previously been detected. A similar approach to that used in BSurvE is then applied to surveillance data for trichinellosis, for which risk-based post-mortem testing is also performed. Negative results from multiple species using di®erent, imperfect tests are combined to give an estimate of the upper 95% con¯dence limit of the national prevalence of trichinellosis in a reference population. This method is used to provide support for freedom from trichinellosis in Great Britain. A di®erent approach to risk-based surveillance is explored as the surveillance strategy for detection of exotic causes of abortion in sheep and goats in New Zealand is examined. Using a geographic information system (GIS) maps of disease risk factors were overlain to produce a risk landscape for the lower North Island. This was used to demonstrate how areas of high- and low-risk of disease occurrence can be identi¯ed and used to guide the design of a risk-based surveillance programme. Secondly, within one surveillance objective there may be many ways in which the available funds or human resources could be distributed. This thesis develops a method to assess BSE surveillance programmes, and provides tools to facilitate BSE detection on the basis of infection risk and to increase the e±ciency of surveillance strategies. A novel approach to allocation of resources is developed, where portfolio theory con- cepts from ¯nance are applied to animal health surveillance. The example of surveil- lance for exotic causes of sheep and goat abortion is expanded upon. Risk of disease occurrence is assessed for a population over di®erent time periods and geographical areas within a country, and portfolio theory used to allocate the number of tests to be carried out within each of these boundaries. This method is shown to be more likely to detect disease in a population when compared to proportional allocation of the available resources. The studies presented here show new approaches that allow better utilisation of imperfect data and more e±cient use of available resources. They allow development of surveillance programmes containing an appropriate balance of scanning and targeted surveillance activities. Application of these methods will enhance the implementation and value of surveillance in animal health.

This thesis is concerned with the application of recently developed epidemiological and statistical tools to inform the optimisation of a national surveillance strategy of considerable importance to human health. The results of a series of epidemiological investigations of surveillance strategies for zoonotic Salmonella are presented. Salmonella are one of the most common and serious zoonotic foodborne pathogenic bacteria globally. These studies were motivated by the increasing focus on the cost-effectiveness of surveillance while maintaining consumer confidence in food supply. Although data from the Danish Salmonella surveillance and control programme has been used in these investigations, the techniques may be readily applied to other surveillance data of similar quality. The first study describes the spatial epidemiological features of Danish Salmonella surveillance and control programme data from 1995 to 2004, using a novel method of spatially adaptive smoothing. The conditional probability of a farm being a case was consistently high in the the south-west of Sonderjylland on the Jutland peninsula, identifying this area for further investigation and targeted surveillance. The identification of clustering of case farms led into the next study, which closely examines one year of data, 2003, for patterns of spatial dependency. K-function analyses provided evidence for aggregation of Salmonella case farms over that of all farms at distances of up to six kilometres. Visual semivariogram analyses of random farm-level effects from a Bayesian logistic regression model (adjusted for herd size) of Salmonella seropositivity, revealed spatial dependency between pairs of farms up to a distance of four kilometres apart. The strength of the spatial dependency was positively associated with slaughter pig farm density. We describe how this might inform the surveillance programme by potentially targeting herds within a four kilometre radius of those with high levels of Salmonella infection. In the third study, farm location details, routinely recorded surveillance information, and industry survey data from 1995 were combined to build a logistic seroprevalence model. This identified wet-feeding and specific pathogen free herd health status as protective factors for Salmonella seropositivity, while purchasing feed was a risk factor. Once adjusting for these covariates, we identified pockets of unexplained risk for Salmonella seropositivity and found spatial dependency at distances of up to six km (95% CI: 2–35 km) between farms. A generalised linear spatial model was fitted to the Jutland data allowing formal estimation of the range of spatial correlation and a measure of the uncertainty about it. There was a large within-farm component to the variance, suggesting that gathering more farm level information would be advantageous if this approach was to be used to target surveillance strategy. The fourth study again considers data from the whole study period, 1995 to 2004. A detailed temporal analysis of the data revealed there was no consistent seasonal pattern and correspondingly no benefit in targeting sampling to particular times of the year. Spatiotemporal analyses suggested a local epidemic of increased seroprevalence occured in west Jutland in late 2000. Lorelogram analyses showed a defined period of statistically significant temporal dependency, suggesting that there is little value in sampling more frequently than every 10 weeks on the average farm. The final study uses findings from the preceding chapters to develop a zero-inflated binomial model which predicts which farms are most at risk of Salmonella, and then preferentially samples these high-risk farms. This type of modelling allows assessment of similarities and differences between factors that affect herd infection status (introduction) and those that affect the seroprevalence in infected herds (persistence and spread). The model suggested that many of the herds where Salmonella was not detected were infected but at a low prevalence. Using cost and sensitivity, we compared the results with those under the standard sampling scheme based on herd size, and the recently introduced risk-based approach. Model based results were less sensitive, but showed significant cost savings. Further model refinements, sampling schemes, and the methods to evaluate their performance are important areas for future work, and should continue to occur in direct consultation with Danish authorities.

The objectives of this thesis were to investigate patterns of lamb pneumonia prevalence of a large sample of New Zealand flocks including an investigation of spatial patterns, to evaluate farm-level risk factors for lamb pneumonia, to determine the efficacy of a commercially available vaccine for the disease and to estimate the likely cost of lamb pneumonia and pleurisy for New Zealand sheep farmers. Data were collected by ASURE NZ Ltd. meat inspectors at processing plants in Canterbury, Manawatu and Gisborne between December 2000 and September 2001. All lambs processed at these plants were scored for pneumonia (scores: 0, <10% or ≥10% lung surface area affected) involving 1,899,556 lambs from 1,719 farms. Pneumonia prevalence was evaluated for spatial patterns at farm level and for hierarchical patterns at lamb, mob and farm levels (Chapter 3). The average pneumonia prevalence in Canterbury, Feilding and Gisborne was 34.2%, 19.1% and 21.4% respectively. Odds ratios of lambs slaughtered between March and May were vastly higher than those slaughtered in other months indicating longer growth periods due to pneumonia. Since pneumonia scores were more variable between mobs within a flock than between flocks, it was concluded that pneumonia scores were poor indicators for the flock pneumonia level due to their lack of repeatability. There was no statistically significant spatial autocorrelation in pneumonia prevalence for any region, hence lamb pneumonia appeared to be largely independent of topographical and geo-climatic factors. A questionnaire-based case-control study was conducted investigating farm-level factors from a sample of farms with either high (case) or zero (control) pneumonia prevalence at slaughter (Chapter 4). Significant risk factors for case farms were: (1) shearing lambs on the day of weaning, (2) breeding ewe replacements on-farm (3) number of lambs sold (an indicator of flock size) and (4) increased percentage of lambs sold late in the season (March to May). Significant protective factors included: (1) set stocking lambs after weaning, (2) injecting lambs with Vitamin B12 at the time of tailing, (3) injecting lambs with Vitamin B12 at weaning. In Canterbury, flocks with Romney ewes and other ewes had a higher risk of pneumonia than those with fine wool type ewes (Merinos, Corriedales or Halfbreds). In a clinical trial, 8,364 lambs from seven commercial sheep farms with a history of lamb pneumonia were vaccinated with Ovipast Plus® or placebo by systematic random allocation within mob and farm. An assessment of the extent of pneumonic lesions was conducted at slaughter and lamb growth rate was monitored through the growth period (Chapter 5). The vaccination trial showed no statistically significant effect of Ovipast® vaccination on the extent of lung lesions at slaughter or ADG of lambs from the first treatment until slaughter. No significant differences were found between isolation rates of Pasteurella spp and patho-histological classifications from pneumonic lung samples of placebo and vaccinated lambs. A spreadsheet-based stochastic model was constructed to estimate the cost of lamb pneumonia and pleurisy to New Zealand farmers. The estimate was based on data of the effect of pneumonia on lamb growth rate, distributions of pneumonia severity, prevalence of moderate to severe pneumonia (≥10% lung surface area) and pleurisy prevalence (Chapter 6). The simulated annual average cost of pneumonia was NZ $28.1 million and that of pleurisy NZ $25.1 million. The combined cost of pneumonia and pleurisy to New Zealand farmers had an average of NZ $53.2 million (95% stochastic interval = $32.4-$78.9 million), or US $31.9 million per annum. This would equate to NZ $2.32 per lamb. In comparison, animal health, shearing expenses and feed expenses cost NZ $2.37, $2.62 and $1.85 per lamb, respectively. This research has demonstrated sub-clinical pneumonia to be a widespread disease in the New Zealand sheep farming population while previous research has focussed on case studies of affected farms. The estimated costs of pneumonia and pleurisy to New Zealand farmers ($53.2 million) highlight the financial effects of these diseases and the need for further research. We also found that the commercially available vaccine could neither prevent sub-clinical effects (lamb growth rate) nor clinical manifestations (pneumonic lung lesions) of lamb pneumonia. The case-control study has revealed farm-level factors which, in the absence of effective vaccines, indicated management practices that farmers might perceive as opportunities to control lamb pneumonia. However, it is advisable to evaluate the efficiency of such management changes. Pneumonia is aetiologically complex disease involving the interplay of many environmental, host and pathogen factors. It is also a difficult disease to study in the absence of diagnostic tests in live animals. However, further research should focus on the development of management changes until effective vaccines are available. A starting point for this research would be to evaluate the impact of such management changes in reducing the incidence of lamb pneumonia. More specifically, the roles of stress during crowding of lambs for extended periods warrants further investigation. The development of efficient vaccines requires an analysis of pathogens, especially Pasteurella (Mannheimia) haemolytica and Mycoplasma species, the sources of infection, their strain diversity and transmission dynamics.

Footrot is a contagious hoof disease of ruminants. It is endemic in New Zealand and throughout sheep and goat farming regions of the world. The disease results from a mixed bacterial infection, but the essential agent is Dichelobacter nodosus, a Gram-negative, anaerobic bacterium that possesses type-IV fimbriae on its surface. Genetic variation in the fimbriae of D. nodosus was investigated in this study. Using the polymerase chain reaction (PCR), the variable region of the gene encoding the fimbrial subunit (fimA) was amplified from bacterial DNA extracted from footrot lesions. Different fimA amplimers were differentiated by single-strand conformation polymorphism (SSCP) analysis. In conjunction with DNA sequencing, 15 unique sequences of D. nodosus fimA were obtained from 14 footrot samples taken from 6 farming regions throughout New Zealand. When these sequences were compared to fimA of known serogroups, it revealed that there were at least 15 D. nodosus strains, representing 8 serogroups, present on New Zealand farms. The predominant serogroup was B which contained 6 strains, followed by serogroups F, H and G. No strains from serogroups D and I were detected in this investigation. Twelve out of the 15 New Zealand D. nodosus strains had fimbriae different to those previously reported and the presence of multiple strains on a single hoof was common (86% of samples). The fimA sequences from the 12 D. nodosus strains incorporated into the footrot vaccine currently available in New Zealand were determined. A primer set targeting the relatively conserved fimA regions and based on the published sequence of serogroup M Nepalese isolates (designated M-Nep), failed to amplify fimA from the vaccine serotype M strain (designated as M-SPAHL). When the downstream primer was substituted with a primer that was specific for other serogroups of D. nodosus, the fimA gene was successfully amplified. Cloning followed by DNA sequencing, revealed that M-SPAHL fimA was different to M-Nep fimA. The predicted amino acid sequence of M-SPAHL fimA did not show homology to any known serogroups or serotypes. The most similar sequence was from serotype F1, and not M-Nep. The sequence difference between M-SPAHL and M-Nep was larger than that expected within a serogroup. The consequences of serological relatedness and sequence dissimilarity are discussed. Only eight of the 15 New Zealand field strains had fimbriae identical to those of the vaccine strains, while the remaining seven strains possessed different fimbriae. In addition, the vaccine contained two more D. nodosus strains, representing two sera groups, that were not found on the New Zealand farms investigated in this study. This may, to some extent, explain why the current footrot vaccine is at times less efficient in New Zealand. Another 17 footrot samples were screened for new or additional D. nodosus strains. Two PCR amplimers (designated X and Y) derived from footrot samples generated SSCP patterns different to those of previously identified strains. DNA sequencing revealed that these two fragments possessed novel sequences. The upstream of X (nt 1-183) was identical to serotype M1 while its downstream (nt 223-414) was identical to serotype F1; the upstream of Y (nt 1-116) was identical to serotype E1 whereas its downstream (nt 148-423) was identical to serotype F1. A 14-mer sequence consisting of two partially overlapping Chi-like sequences, 5'-GCTGGTGCTGGTGA-3', was also found in these fragments. Two primer sets with the downstream primer specific for serotype Fl and the upstream primer specific for serotype M1 or E1, produced PCR products of the expected sizes from the footrot samples from which fragments X and Y were isolated, respectively. These primer sets did not appear to amplify artificially mixed genomic DNA from serotypes M1 and F1 or E1 and F1. However, when the reactions were re-amplified, PCR recombination artefacts were observed, suggesting that PCR recombination does occur, but at a low frequency. It therefore seems more likely that fragments X and Y reflect genuine fimA sequences of D. nodosus which have resulted from in vivo DNA recombination, than from a PCR recombination artifact. The genetic capability for recombination at the fimbrial subunit locus may therefore endow D. nodosus with the ability to alter its antigenic appearance. D. nodosus strains present in footrot lesions can be genotyped using a PCR-SSCP/sequencing technique. However, this typing technique requires cloning and screening of D. nodosus fimA sequences, which is both laborious and costly. A rapid molecular typing system for D. nodosus was therefore developed in this study. A close examination of available D. nodosus fimA sequences revealed regions that appear to be specific for serogroups and serotypes. These regions were used to design a panel of sequence-specific oligonucleotide probes (SSOPs), and a rapid and accurate D. nodosus typing system using PCR and reverse dot-blot hybridisation (PCR/oligotyping) was subsequently developed. The variable region of D. nodosus fimA, amplified and labelled with digoxigenin (DIG) in a single multiplex PCR amplification, was hybridised to a panel of group- and type-specific, poly-dT tailed oligonucleotides that were immobilised on a nylon membrane strip. A mixture of positive control poly-dT tailed oligonucleotides was also included on the membrane. After hybridisation the membrane was washed to a defined specificity, and DIG-labelled fragments that had hybridised were detected. The specificity of the oligonucleotides was verified by the lack of cross-reactivity with D. nodosus fimA sequences that had a single base difference. DNA from 14 footrot samples previously genotyped by PCR-SSCP/sequencing, was assayed using the PCR/oligotyping technique. All types of D. nodosus which had been detected previously with a PCR-SSCP/sequencing method were detected by this procedure. However, for three of the 14 footrot samples, PCR/oligotyping detected additional types of D. nodosus. Further PCR amplification using type-specific primers, confirmed that these types were present in the original footrot samples. These results indicate that PCR/oligotyping is a specific, accurate, and useful tool for typing footrot samples. In combination with a rapid DNA extraction protocol, D. nodosus present in a footrot sample can be accurately genotyped in less than two days. Individual animals from the same farm, or the same paddock, were often infected by different strains of D. nodosus. This suggests a host role in mediating footrot infection, or that the interaction between the pathogen and the host is important. In order to better understand the interaction between the bacterium and the host, two polymorphic ovine class II MHC genes DQA1 and DQA2, which have been previously shown to be important in footrot infection, were also investigated in this study. PCR-SSCP/sequencing analysis of the DQA1 locus revealed ten unique ovine DQA1 sequences, with five of them being newly identified. This increases the number of known ovine DQA1 alleles from 8 to 13 (including a null allele), implying a high level of polymorphism at the ovine DQA1 locus. D. nodosus present on 20 footrot infected sheep from the same flock were genotyped, together with the ovine DQA1 and DQA2 genotypes of their hosts. Preliminary results showed that sheep with the same DQA1 and DQA2 genotypes tended to be infected by similar types of D. nodosus. Different types of D. nodosus were generally found on sheep with different genotypes at either the DQA1 or the DQA2 locus. This suggests the diversity in D. nodosus infection may be associated with the heterogeneity in the host MHC. However, as only a small number of animals from the same sire were analysed, further investigation is needed to gain a better understanding of the interaction between D. nodosus and the host MHC.

Animal health surveillance is an important part of animal health care, particularly in countries dependent on livestock for food production and international trade. There are two major issues related to the provision of e®ective surveillance activities. Firstly, for good information to become available, the design and conduct of data collection activ- ities should be carried out following sound statistical principles. In reality, constraints such as imperfect tests and unavoidably-biased sampling strategies hinder straightfor- ward analysis and interpretation of survey results. Risk-based surveillance is used to target high-risk sub-populations to increase e±ciency of disease detection; however, biased datasets are generated. This thesis develops methodologies to design risk-based surveillance systems and al- low statistically valid analysis of the inherently biased data they generate. The ¯rst example describes the development of a method to analyse surveillance data gathered for bovine spongiform encephalopathy (BSE). The data are collected from four dif- ferent surveillance streams of animals tested for BSE, with each stream containing unavoidable biases and limitations. In the BSurvE model, these data are combined with demographic information for each birth cohort to estimate the proportion of each birth cohort infected with BSE. The prevalence of BSE in a national herd can then be estimated using the method of moments, whereby the observed number of infected animals is equated with the expected number. The upper 95% con¯dence limit for the prevalence is estimated both for infected countries and for those where no BSE has previously been detected. A similar approach to that used in BSurvE is then applied to surveillance data for trichinellosis, for which risk-based post-mortem testing is also performed. Negative results from multiple species using di®erent, imperfect tests are combined to give an estimate of the upper 95% con¯dence limit of the national prevalence of trichinellosis in a reference population. This method is used to provide support for freedom from trichinellosis in Great Britain. A di®erent approach to risk-based surveillance is explored as the surveillance strategy for detection of exotic causes of abortion in sheep and goats in New Zealand is examined. Using a geographic information system (GIS) maps of disease risk factors were overlain to produce a risk landscape for the lower North Island. This was used to demonstrate how areas of high- and low-risk of disease occurrence can be identi¯ed and used to guide the design of a risk-based surveillance programme. Secondly, within one surveillance objective there may be many ways in which the available funds or human resources could be distributed. This thesis develops a method to assess BSE surveillance programmes, and provides tools to facilitate BSE detection on the basis of infection risk and to increase the e±ciency of surveillance strategies. A novel approach to allocation of resources is developed, where portfolio theory con- cepts from ¯nance are applied to animal health surveillance. The example of surveil- lance for exotic causes of sheep and goat abortion is expanded upon. Risk of disease occurrence is assessed for a population over di®erent time periods and geographical areas within a country, and portfolio theory used to allocate the number of tests to be carried out within each of these boundaries. This method is shown to be more likely to detect disease in a population when compared to proportional allocation of the available resources. The studies presented here show new approaches that allow better utilisation of imperfect data and more e±cient use of available resources. They allow development of surveillance programmes containing an appropriate balance of scanning and targeted surveillance activities. Application of these methods will enhance the implementation and value of surveillance in animal health.

Animal health surveillance is an important part of animal health care, particularly in countries dependent on livestock for food production and international trade. There are two major issues related to the provision of e®ective surveillance activities. Firstly, for good information to become available, the design and conduct of data collection activ- ities should be carried out following sound statistical principles. In reality, constraints such as imperfect tests and unavoidably-biased sampling strategies hinder straightfor- ward analysis and interpretation of survey results. Risk-based surveillance is used to target high-risk sub-populations to increase e±ciency of disease detection; however, biased datasets are generated. This thesis develops methodologies to design risk-based surveillance systems and al- low statistically valid analysis of the inherently biased data they generate. The ¯rst example describes the development of a method to analyse surveillance data gathered for bovine spongiform encephalopathy (BSE). The data are collected from four dif- ferent surveillance streams of animals tested for BSE, with each stream containing unavoidable biases and limitations. In the BSurvE model, these data are combined with demographic information for each birth cohort to estimate the proportion of each birth cohort infected with BSE. The prevalence of BSE in a national herd can then be estimated using the method of moments, whereby the observed number of infected animals is equated with the expected number. The upper 95% con¯dence limit for the prevalence is estimated both for infected countries and for those where no BSE has previously been detected. A similar approach to that used in BSurvE is then applied to surveillance data for trichinellosis, for which risk-based post-mortem testing is also performed. Negative results from multiple species using di®erent, imperfect tests are combined to give an estimate of the upper 95% con¯dence limit of the national prevalence of trichinellosis in a reference population. This method is used to provide support for freedom from trichinellosis in Great Britain. A di®erent approach to risk-based surveillance is explored as the surveillance strategy for detection of exotic causes of abortion in sheep and goats in New Zealand is examined. Using a geographic information system (GIS) maps of disease risk factors were overlain to produce a risk landscape for the lower North Island. This was used to demonstrate how areas of high- and low-risk of disease occurrence can be identi¯ed and used to guide the design of a risk-based surveillance programme. Secondly, within one surveillance objective there may be many ways in which the available funds or human resources could be distributed. This thesis develops a method to assess BSE surveillance programmes, and provides tools to facilitate BSE detection on the basis of infection risk and to increase the e±ciency of surveillance strategies. A novel approach to allocation of resources is developed, where portfolio theory con- cepts from ¯nance are applied to animal health surveillance. The example of surveil- lance for exotic causes of sheep and goat abortion is expanded upon. Risk of disease occurrence is assessed for a population over di®erent time periods and geographical areas within a country, and portfolio theory used to allocate the number of tests to be carried out within each of these boundaries. This method is shown to be more likely to detect disease in a population when compared to proportional allocation of the available resources. The studies presented here show new approaches that allow better utilisation of imperfect data and more e±cient use of available resources. They allow development of surveillance programmes containing an appropriate balance of scanning and targeted surveillance activities. Application of these methods will enhance the implementation and value of surveillance in animal health.

There is mounting evidence that parasitic infections change nutrient utilisation within the tissues, and that this is responsible for the reduction in animal performance that has been observed. Feeding forages that contain condensed tannins (CT) are thought to alleviate the impact of parasite infection on amino acid (AA) and protein metabolism by improving protein supply post-ruminally. However, there has been no quantification of how nutrients are partitioned in the lamb fed fresh forages during a parasitic infection. Therefore, the objective of this study was to quantify the partitioning of AA between the gastrointestinal tract (GIT), liver and the hind limb tissues (muscle, skin, fat) in lambs during an established parasite infection. It was hypothesised that the feeding of CT would alter the partitioning of AA between the GIT, liver and hind limbs in lambs with an established parasite burden due to the increased availability of dietary AA to the small intestine. This hypothesis was tested in two separate experiments, which had a similar experimental design. In the first experiment (Experiment One; 1999) lambs were fed fresh Lucerne (Medicago sativa; contains no CT). In Experiment Two, which was conducted in 2000, the lambs were fed fresh Sulla (Hedysarum coronarium; 2.2% CT; Experiment Two). One week prior to infection, permanent indwelling catheters were placed in the mesenteric artery, and the mesenteric, portal and hepatic veins and vena cava for blood sampling. Additional permanent catheters were placed in the mesenteric vein (upstream from the sampling catheter) and abdominal aorta for infusion of para-aminohippuric acid (PAH) and indocyanin green (ICG) respectively, to measure plasma flow across the splanchnic tissues (PAH) and the hind limbs (ICG). A permanent Teflon cannula was fitted in the abomasum for the infusion of [1-13C]-valine and [35S]-cysteine (Chapters Five, Six and Seven only) on day 48 post infection to measure valine and cysteine kinetics across the mesenteric-drained viscera (MDV), portal-drained viscera (PDV), liver, total splanchnic tissues (TSP; PDV + liver) and hind limbs. A temporary catheter was inserted into the jugular vein two days before the start of blood sampling for the infusion of deuterium oxide (D2O), and [13C]-sodium bicarbonate and [35S]-sulfate (Chapters Five, Six and Seven only) on day 45 post infection, and [3,4-3H]-valine on day 48 post infection. Lambs were dosed with 6 000 L3 T. colubriformis larvae for 6 d (n=5) or kept as parasite free controls (n=6). Faecal egg production was monitored every second day from day 22 to day 48 post infection and total intestinal worm burdens were determined at slaughter. Blood was continuously collected from the mesenteric, portal and hepatic veins, the mesenteric artery and the vena cava in 2-hour aliquots. Plasma was harvested and AA and metabolite concentrations measured and the specific radioactivity (SRA) and isotopic enrichment (IE) of valine and cysteine were determined. After the completion of blood sampling, but while the [3, 4-3H]-valine infusate was still being administered, the sheep were euthanased by an intravenous overdose of sodium pentobarbitone. Tissue samples were rapidly collected from the sheep in the following order: skin, muscle (biceps femoris), liver, duodenum, ileum, spleen, mesenteric lymph nodes and thymus. Digesta was also sampled from the abomasum and ileum after slaughter in order for the apparent absorption of AA to be determined. The results from Experiment One (Lucerne-fed lambs) suggest that there is no re-partitioning of AA from the posterior hind limbs to the GIT and liver during an established infection- The changes that occurred within the PDV suggests that an established parasitic infection may trigger a localised alteration in AA metabolism and/or protein turnover without significantly changing the metabolism of AA and proteins in tissues peripheral to the TSP tissues and impacting negatively on the growth of the parasitised lambs. In Experiment Two (Sulla-fed lambs) a reduction in feed intake was likely to be the reason for the alterations in the first pass metabolism of AA in the TSP tissues due to the decreased apparent AA absorption by the MDV observed in the parasitised lambs. However, the results from this experiment are in agreement with those from Experiment One confirming that there is no increase in partitioning of AA from the hind limbs to the GIT or liver during an established parasite infection. Although a statistical comparison cannot be made between the data in Experiment One (Lucerne-fed) and Experiment Two (Sulla-fed), it appears that the beneficial effects of feeding CT during a parasitic infection is due to the reduction in larval establishment in the GIT of the lamb, rather than increased AA availability. In conclusion, an established infection imposes no measurable metabolic cost on the lamb, when feed intake is not reduced. When feed intake is reduced, there is no detectable mobilisation of protein from the hind limb. Therefore, localised or other sources of AA and/or energy substrates may be utilised.

This thesis is concerned with the application of recently developed epidemiological and statistical tools to inform the optimisation of a national surveillance strategy of considerable importance to human health. The results of a series of epidemiological investigations of surveillance strategies for zoonotic Salmonella are presented. Salmonella are one of the most common and serious zoonotic foodborne pathogenic bacteria globally. These studies were motivated by the increasing focus on the cost-effectiveness of surveillance while maintaining consumer confidence in food supply. Although data from the Danish Salmonella surveillance and control programme has been used in these investigations, the techniques may be readily applied to other surveillance data of similar quality. The first study describes the spatial epidemiological features of Danish Salmonella surveillance and control programme data from 1995 to 2004, using a novel method of spatially adaptive smoothing. The conditional probability of a farm being a case was consistently high in the the south-west of Sonderjylland on the Jutland peninsula, identifying this area for further investigation and targeted surveillance. The identification of clustering of case farms led into the next study, which closely examines one year of data, 2003, for patterns of spatial dependency. K-function analyses provided evidence for aggregation of Salmonella case farms over that of all farms at distances of up to six kilometres. Visual semivariogram analyses of random farm-level effects from a Bayesian logistic regression model (adjusted for herd size) of Salmonella seropositivity, revealed spatial dependency between pairs of farms up to a distance of four kilometres apart. The strength of the spatial dependency was positively associated with slaughter pig farm density. We describe how this might inform the surveillance programme by potentially targeting herds within a four kilometre radius of those with high levels of Salmonella infection. In the third study, farm location details, routinely recorded surveillance information, and industry survey data from 1995 were combined to build a logistic seroprevalence model. This identified wet-feeding and specific pathogen free herd health status as protective factors for Salmonella seropositivity, while purchasing feed was a risk factor. Once adjusting for these covariates, we identified pockets of unexplained risk for Salmonella seropositivity and found spatial dependency at distances of up to six km (95% CI: 2–35 km) between farms. A generalised linear spatial model was fitted to the Jutland data allowing formal estimation of the range of spatial correlation and a measure of the uncertainty about it. There was a large within-farm component to the variance, suggesting that gathering more farm level information would be advantageous if this approach was to be used to target surveillance strategy. The fourth study again considers data from the whole study period, 1995 to 2004. A detailed temporal analysis of the data revealed there was no consistent seasonal pattern and correspondingly no benefit in targeting sampling to particular times of the year. Spatiotemporal analyses suggested a local epidemic of increased seroprevalence occured in west Jutland in late 2000. Lorelogram analyses showed a defined period of statistically significant temporal dependency, suggesting that there is little value in sampling more frequently than every 10 weeks on the average farm. The final study uses findings from the preceding chapters to develop a zero-inflated binomial model which predicts which farms are most at risk of Salmonella, and then preferentially samples these high-risk farms. This type of modelling allows assessment of similarities and differences between factors that affect herd infection status (introduction) and those that affect the seroprevalence in infected herds (persistence and spread). The model suggested that many of the herds where Salmonella was not detected were infected but at a low prevalence. Using cost and sensitivity, we compared the results with those under the standard sampling scheme based on herd size, and the recently introduced risk-based approach. Model based results were less sensitive, but showed significant cost savings. Further model refinements, sampling schemes, and the methods to evaluate their performance are important areas for future work, and should continue to occur in direct consultation with Danish authorities.