The prevalence and public health implications of Salmonella brandenburg, an emerging pathogen of sheep in New Zealand : a thesis presented in partial fulfilment (50%) of the requirements for the degree of Master of Veterinary Science in Veterinary Public Health at Massey University, Palmerston North, New Zealand

Since the first case of Salmonella brandenburg abortion was recorded from a single sheep farm in the South Island of New Zealand in 1996, the outbreak had spread to more than 300 farms by the lambing season of 2000. This study, which was funded by Meat New Zealand and other stakeholders, is a pilot project to estimate the prevalence of Salmonella brandenburg and consider its potential foodborne and occupational health risk implications. Eight farms, four control farms and four affected farms were chosen from known infected areas. Control farms were those which had not experienced Salmonella brandenburg outbreaks, while the four affected properties had clinical outbreaks during the 2000 lambing season which had either been laboratory confirmed or not. At each farm faecal samples were collected from 50 lambs and 50 ewes at drafting and at slaughter. Therefore a total of 200 samples were taken from each group of animals sent for slaughter. The sampling was done in two phases to determine the effect of time interval on the prevalence of Salmonella brandenburg in sheep between abortion outbreaks and slaughter. The same sampling routine was followed for the November-December (Phase I) and February-March (Phase II) periods. The isolates from faecal cultures that had been confirmed as Salmonella spp. by slide agglutination test were sent to the Institute of Environmental Science and Research (ESR) for serotyping. All the 133 samples sent for serotyping turned out as S. brandenburg. The primary comparison of the study was done between the prevalence of Salmonella brandenburg in animals from affected farms and control farms. In addition the study was also interested in comparing S. brandenburg prevalence within class i.e. in lambs or ewes from the same farm at drafting and at slaughter. Comparison was also made between classes i.e. the prevalence of the organism in lambs and ewes from the same farm at drafting and at slaughter. During phase I of the on-farm sampling the prevalence of Salmonella brandenburg in tested lambs and ewes from affected farms was 12.0.% and 18.7% respectively. The on-farm prevalence for the control farms was 4.0% for lambs and 3.5% for ewes. During phase I of abattoir sampling the overall prevalence for the affected farms was 9.0 % for lambs and 22.0 % for ewes compared to 0.0% for lambs and 1.0% for ewes from control farms. The high prevalence of Salmonella brandenburg in animals from affected farms as compared to control farms showed that affected farms were associated with high excretion rates and therefore high levels of environmental contamination. During phase II of the on-farm sampling the prevalence of S. brandenburg in tested lambs and ewes from affected farms was 2.5% and 2.7% respectively. The phase II on-farm prevalence of S. brandenburg from control farms was 0.0% for lambs and 0.8% for ewes. During the same phase, abattoir prevalence of S. brandenburg in lambs and ewes from affected farms was 0.0% and 2.7% respectively compared to 0.5% for lambs and 0.0% for ewes from control farms. Like in phase I the overall prevalence of S. brandenburg was higher in animals from affected farms as compared to animals from control farms. The study also showed that the prevalence of the organism was very high during phase I compared to phase II irrespective of class of animal or site of sampling. This could have been due to the high number of animals still excreting the organisms closer to the outbreak period or the high level of environmental contamination. Both factors would have contributed to a higher prevalence of positive cultures. The higher prevalence of positive cultures during the November-December period as compared to the February-March period showed that the risk of infection and product contamination was greatest at commencement of the season (November-December) but was greatly reduced by February. Therefore further research is required to find the production and processing methods that might reduce the risk of infection and product contamination during the period of November-December. However the positive cultures of Salmonella brandeburg in control farms suggested a spreading disease outbreak and that the absence of clinical outbreaks of the disease did not mean an absence of infected animals on-farm. It is very important to do further investigations to find on-farm risk factors that might result in the absence or presence of clinical outbreaks. The study did not show any obvious differences in the S. brandenburg prevalence within class between on-farm and slaughter samples. There was also no obvious difference in the prevalence of the organism between ewes and lambs from the same farm, either during on-farm or abattoir sampling. Pulsed-field gel electrophoresis (PFGE) of the 24 isolates, which were a representative sample of the study, gave an identical profile. The PFGE and the serotyping suggested that the outbreak strain had become the dominant serotype in the sampled farms in the outbreak regions of the South Island of New Zealand. Therefore factors that gave rise to this dominance should be further investigated