Food product contamination leading to a food borne illness is real and has potentially devastating impact on supply chain operations and cost. However, it is not well understood from the quantitative perspective. This research seeks to fill this gap by providing a generic model of a multi-stage food supply chain consisting of a supplier/grower, processing center and retailer(s) and analyzing the impact of food product contamination in this model. The supplier corresponds to the farm/grower of the raw material such as fruits and vegetables, the processing center processes the raw material into a final food product and the retailer corresponds to the supermarkets and grocery stores selling the food product to a customer. A situation where a contamination occurs at the supplier or processing center potentially resulting in a food borne illness to the customer is considered. The contamination is discovered through periodic sampling tests conducted by the grower, processing center or through the outbreak of a food borne illness. The supply chain is modeled utilizing a G/G/1 queuing system at the processing center and an order- up to policy at the retailer(s). This research develops and compares multi-stage supply chain models with varying number of retailers. The negative dependence of contamination on the origin and mode of detection of the contamination is quantified. The differences in individual food product attributes which can impact the cost of contamination are analyzed. The impact of supply chain structure and properties and detection policies on the severity of potential contamination cases is studied. The most cost effective sampling strategies which companies can adopt in the event of product contamination are derived. The payoff from the implementation of a quality control process which can eradicate contamination is evaluated. A numerical study of the impact of a real-world contamination event on a tomato and lettuce supply chain is also conducted. Finally, a traceability system capable of tracking and tracing back products in the event of a food product recall is incorporated in the supply chain model. The value of traceability for different supply chain scenarios is assessed through the implementation of an ARENA based simulation model.

Understanding the ecology and epidemiology of tick-borne diseases requires detailed knowledge of the complex interactions among the tick vector, the microorganisms they carry and the vertebrate hosts used by ticks, as well as the environmental conditions experienced by all three groups of organisms in this triad. In this thesis, I addressed questions relating to the biology and vector ecology of the Rocky Mountain wood tick (Dermacentor andersoni) and the American dog tick (Dermacentor variabilis). Comparisons were made of the distribution of both tick species, the vertebrate hosts used by immature ticks, and the types and prevalence of bacteria in individual ticks from multiple localities near the northern extent of their geographic ranges in western Canada. The results revealed that the distributions of both D. andersoni and D. variabilis have expanded since the 1960s, and there is now a broad zone of sympatry in southern Saskatchewan. In this zone of sympatry, D. andersoni and D. variabilis immatures were found to use the same species of small mammals as hosts and, in some cases, the same host individuals. This provides for the possibility of cross-transmission of bacteria from one tick species to the other. Bacteria of several genera (e.g. Rickettsia, Francisella, Arsenophonus and Anaplasma) were detected in D. andersoni and/or D. variabilis, some of which represented new tick-bacteria associations. However, most bacterial species were highly host (tick)-specific, except for three examples of apparent host switching from one tick species to the other at localities where the two tick species occurred in sympatry. The findings of this thesis provide a basis for understanding microbial transmission, the structure of tick-borne microbial communities, the risk of tick-borne disease in humans and animals, and the vector potential of D. andersoni and D. variabilis in geographical areas where they have not been studied previously.

Understanding the ecology and epidemiology of tick-borne diseases requires detailed knowledge of the complex interactions among the tick vector, the microorganisms they carry and the vertebrate hosts used by ticks, as well as the environmental conditions experienced by all three groups of organisms in this triad. In this thesis, I addressed questions relating to the biology and vector ecology of the Rocky Mountain wood tick (Dermacentor andersoni) and the American dog tick (Dermacentor variabilis). Comparisons were made of the distribution of both tick species, the vertebrate hosts used by immature ticks, and the types and prevalence of bacteria in individual ticks from multiple localities near the northern extent of their geographic ranges in western Canada. The results revealed that the distributions of both D. andersoni and D. variabilis have expanded since the 1960s, and there is now a broad zone of sympatry in southern Saskatchewan. In this zone of sympatry, D. andersoni and D. variabilis immatures were found to use the same species of small mammals as hosts and, in some cases, the same host individuals. This provides for the possibility of cross-transmission of bacteria from one tick species to the other. Bacteria of several genera (e.g. Rickettsia, Francisella, Arsenophonus and Anaplasma) were detected in D. andersoni and/or D. variabilis, some of which represented new tick-bacteria associations. However, most bacterial species were highly host (tick)-specific, except for three examples of apparent host switching from one tick species to the other at localities where the two tick species occurred in sympatry. The findings of this thesis provide a basis for understanding microbial transmission, the structure of tick-borne microbial communities, the risk of tick-borne disease in humans and animals, and the vector potential of D. andersoni and D. variabilis in geographical areas where they have not been studied previously.

Feral swine (Sus scrofa domesticus) are spreading across North America at an alarming rate. Four Canadian provinces and 39 states within the continental United States now report standing populations of feral pigs. Estimates place the number of feral hogs in Texas >2M, accounting for more than half of the United States population. It is known that feral swine impact local ecology following establishment, with regard to shifts in local flora and fauna. The overall objective of this research was to investigate the role of feral swine in tick-borne pathogen transmission in Texas. The underpinning objectives were to establish host records for tick species parasitizing feral swine, determine the species assemblages associated with feral swine among different ecoregions of Texas, determine by immunoassay to which tick-borne bacteria feral pigs were being exposed, and detect the DNA of tick-borne bacteria by polymerase chain reaction assay in the event of poor or early immune response by the host. Feral pigs (N=432) were harvested from June 2008 to June 2010 using box and corral traps and by aerial gunning. Seven species of ticks, Amblyomma americanum, A. cajennense, A. maculatum, Dermacentor albipictus, D. halli, and D. variabilis; and Ixodes scapularis, were collected. Immature stages of A. cajennense and A. americanum were collected as well. All classes of feral pigs, gender by age, were infested with ticks. Serum was collected through a multi-organizational effort from 2006 to 2010 and tested by ELISA for previous exposure to tick-borne pathogens in the genera Rickettsia and Ehrlichia (N=888) and Borrelia (N=849). Prevalence percentages by immunoassay were 27.59 percent, 13.18 percent and 2.12 percent for Rickettsia, Ehrlichia, and Borrelia, respectively. Samples positive by ELISA for exposure to Borrelia were further screened by Western Blot for exposure to Borrelia turicatae. The results were equivocal in most cases. Blood samples (N=233) were collected from 2008 to 2010 and analyzed by polymerase chain reaction for the detection of the DNA of these same three genera of bacteria. Two of the samples were positive by PCR for the presence of Borrelia DNA. Texas feral swine are serving as hosts for at least seven species of ticks and are interacting with tick-borne pathogen transmissions cycles in Texas.

­^¤åºK­n Multiplex PCR amplification of 16S rRNA gene¡BvirA¡Btpl¡Band H1d genes was developed enabling simultaneous detection in Escherichia coli¡Aan indicator of fecal contamination and food-borne microbial pathogens¡AShigella flexneri¡BCitrobacter freundii¡BSalmonella typhi¡BVibrio cholerae¡BVibrio parahaemolyticus¡Band Staphylococcus aureus¡CEach of the nine pairs of oligonucleotide primers was found to support PCR amplifications of only its targeted gene¡CThe optimized multiplex PCR reaction utilized a primer annealing temperature of 59 ¢Jand used agarose gel electrophoresis for detection of the PCR-amplified products¡CSelection of appropriate target genes¡Boligonucleotide primers ¡BPCR reaction¡Band cycling parameters resulted in the amplification of four target genes simultaneously in a single PCR reaction with the sensitivity of detection was 102 CFU after 32 cycles¡CMultiplex PCR amplification followed by differential PCR for E. coli / Shigella¡A and Citrobacter / Salmonella¡Asequenced for the PCR-amplified products of 16S rRNA gene of the seven pathogens in this study¡Aand used restriction endonuclease AfaI to confirm the PCR-amplified products of V. cholerae¡AV. parahaemolyticus and Staphylococcus aureus¡Ahas been shown to be an sensitive¡Aspecific¡Aand rapid method to detect food-borne bacterial pathogens¡C

Thesis (MSc (Veterinary Tropical Diseases)--University of Pretoria, 2009. / Includes bibliographical references. Also available in print format.

Background: The purpose of this secondary data analysis was to determine the prevalence of sexually transmitted and blood-borne infections (STBBI) and to examine the role of survival sex in sexual risk among street-involved youth in Winnipeg, Manitoba. Methods: Between 1999 and 2003, 743 Winnipeg street youth were interviewed in three separate cohorts as part of a national multi-centred study by the Public Health Agency of Canada. Results: Among the 673 Winnipeg street youth who responded to questions relating to survival sex, 26.3% reported they had engaged in survival sex. Females and Aboriginal youth were at greatest risk for sex trade involvement. Youth engaged in survival sex were diagnosed with significantly higher rates of STBBI; experienced greater amounts of abuse; and were at greater risk for participating in high risk sexual behaviours compared to their non-sex trade counterparts. Conclusion: Gender specific, culturally appropriate, and youth oriented prevention and intervention strategies are urgently required to reduce the prevalence of STBBI and survival sex among this at-risk population.

Nous nous intéressons dans ce travail à la dynamique des opérateurs de Schrödinger unidimensionnels, discrets, associés à un potentiel sturmien quasi-périodique. Le résultat principal de cette thèse est une borne supérieure pour les exposants de transport qui mesurent la vitesse de propagation du système. Cette borne, valide pour presque tous les potentiels sturmiens, est sous balistique pour une force de couplage suffisante. La validité de la borne est couplée à une condition diophantienne liée au nombre irrationnel qui définit le potentiel. Cette condition est vraie presque sûrement. Nous exhibons par ailleurs un exemple d'irrationnel pour lequel une borne supérieure sous balistique est impossible indépendamment de la force de couplage. Nous faisons l'étude de la dimension fractale du spectre de l'opérateur qui minore sous certaines conditions les exposants de transport. Nous obtenons une nouvelle borne inférieure pour la dimension de boîte du spectre grâce aux propriétés connues sur la forme du pseudo spectre. Les restrictions pour obtenir une borne dynamique à partir de notre résultat sont d'avoir une condition initiale cyclique standard et que le potentiel soit associé à un irrationnel à densité bornée. Enfin dans la dernière partie de ce travail, nous démontrons que le spectre de l'opérateur associé au nombre d'argent ß = [2, 2, . . . ] possède une structure hyperbolique. L'expression du pseudo spectre peut être vu comme un système dynamique. Nous conjuguons ledit système à une dynamique symbolique abstraite selon la méthode dite des partitions de Markov. Le système se comporte comme un fer à cheval de Smale. Nous dérivons de l'hyperbolicité des propriétés pour les dimensions fractales du spectre. Dimensions dont l'attrait dynamique a été rappelé dans la partie précédente. Nous déduisons notamment l'égalité des dimensions de Hausdorff et de boîte pour cet opérateur.

In 1994 and 1995, seven cases of Human monocytic ehrlichiosis were reported in Indiana; six cf these were from southern counties. To test whether Ehrlichia cha ffeensis is present in native ticks, adult ticks representing two species, Dermacentor variabilis and Amblyomma americanum were collected in southern Indiana during the first week of May, 1995 and tested by PCR. A total of 510 D. variabilis ticks were collected and placed into 102 pools (5 ticks per pool). A total of 430 adult A. americanum ticks were collected and placed into 88 pools of 5 ticks or less.D. variabilis ticks were used to optimize the isolation of PCR amplifiable DNA and determine the minimum number of bacterial cells detected. A modified CTAB-phenol method permitted the detection of as few as 100 bacterial cells. The optimal amount of isolated DNA for a PCR amplification was determined to be 2.7 pg of total nucleic acid.The 88 pools of adult A. americanum were subjected to DNA isolation, PCR amplification, and Southern analysis to determine the presence of E. chaffeensis bacteria. Using the 16S rRNA gene from E. chaffeensis, with species specific primers Hut and HE3. Of the 88 pools (430 ticks; 21 were determined to be positive for the presence E. chaffeensis bacteria. Thus, minimal field infection rate for adult A. americanum ticks in Southern Indiana was 4.88%. This calculation is based upon the assumption that at least one tick was positive in each positive pool. / Department of Biology

Assessment of Hazard Analysis and Critical Control Points Principles in Primary School Feeding Schemes in the Western Region of Gauteng Objective: The objective of this study is to investigate the need for standard hygiene and safety practices for the National School Nutrition Program (NSNP). Food safety procedures and practices can be implemented as part of the Hazard Analysis of Critical Control Points (HACCP) program. Background: The National School Nutrition Program (NSNP) forms part of the Integrated Nutrition Program (INP) of South Africa and was implemented as a poverty alleviation strategy introduced in 1994 by government as part of the newly founded democratic Republic of South Africa. The NSNP is implemented in primary schools. Design: A questionnaire was designed to investigate, the hygiene and safety practices in primary schools in the Western Region of Gauteng. Both rural and urban schools using the NSNP were targeted. Research Project: The NSNP staff from 26 primary schools completed the research questionnaire focused on hygiene and safety practices. Results: Data collected indicated the need to implement a very basic HACCP program that include identifying the HACCP team, training for food handlers, standardized equipment and the implementation of hygiene and safety procedures. Conclusion: To implement HACCP in the NSNP, the role players from the managerial structures and food handlers at schools, the training needs of the target group and the equipment needed to comply with HACCP procedures, should to be identified. Documentation must be in place for monitoring, evaluation and verification procedures to implement the HACCP program. / Thesis (M.Sc. (Dietetics))--North-West University, Potchefstroom Campus, 2009.