<p>This mini-thesis is concerned with histories of flower selling in Cape Town. Since the late 19th century, images and imaginings of the flower sellers in Adderley Street and to a lesser degree in other areas of the city attained an outstanding place in visualisations and descriptions of Cape Town. The flower sellers were thereby characterised in a particularly gendered, racialised and class-specific way as predominantly female, coloured and poor. This characterisation dominated to an extent that it is possible to speak of a discursive figure of the ̕̕Cape Town flower sellers̕̕. In tourism-related media and in personal memoirs, the ̕̕Cape Town flower sellers̕̕ often came to represent both the city and the inhabitants of Cape Town. The images and imaginings of the &lsquo / Cape Town flower sellers&rsquo / can partly be traced back to representations of ̓̓̕̕flower girls̕̕ in fictional stories, paintings, photographs and film in Europe, particularly in Great Britain. In Cape Town, this European discourse about flower selling developed in a specific way within colonial, apartheid and post-apartheid contexts.</p>

Influenza virus is a RNA virus that exists as different types and subtypes. Influenza A virus strains are known to cause disease in several bird and mammalian species. Wild birds are believed to constitute the natural reservoir for influenza A virus. In humans, influenza A virus causes yearly seasonal influenza epidemics of respiratory disease resulting in high morbidity and severe economic consequences. Due to the virus’ ability to change its antigenic properties by mutation, yearly vaccination is required for protection from the disease. There are many different subtypes of influenza virus which are characterized according to two surface structures - the hemagglutinin and neuraminidase proteins - , for example; H5N1. These subtypes have the ability to recombine, and thereby creating new variant combinations. If a subtype that the living population of humans has not encountered before starts to spread among humans, it can result in a pandemic. Pandemic outbreaks have occurred at irregular intervals throughout history and have had a devastating impact on mankind. For example the Spanish influenza pandemic of 1918 is thought to have killed more than 50 million people. Influenza A virus is also an important cause of disease in poultry where virus strains of some subtypes may change into forms that are highly pathogenic. These virus strains may transmit directly to man and multiple other species. This has been the case in the ongoing outbreak that started in South East Asia in 2003. All known subtypes of influenza A virus have been isolated from wild birds living in aquatic environments, mainly dabbling ducks. These species are considered to be the reservoir for influenza A virus. The virus causes sub clinical gastrointestinal infection in ducks. High amounts of virus are excreted in the feces and spread via the fecal-oral route through water where it can persist for a prolonged time. There are still many unknowns about the ecology of influenza virus in the wild bird reservoir. This thesis includes five articles where data are presented that add new knowledge on this subject. We add proof that wild ducks are indeed the host for most influenza A virus subtypes by presenting data from a meta-analysis on all published screening data from wild birds and by presenting data from a four year screening of migratory ducks that were caught and sampled at Ottenby Bird Observatory. Our investigations have shown that the prevalence of influenza virus in the wild duck population of western Eurasia shows temporal differences in comparison to the results found in studies in North America. The prevalence in western Eurasian ducks is high during the period August to December and also rises in the spring. These findings are of importance for the understanding of how influenza virus is perpetuated in nature. During the course of the study only low pathogenic subtypes were isolated. Of concern is the high frequency of isolation of virus strains of the H5 and H7 subtypes that are prone to change into highly pathogenic variants in poultry. Many of the strains isolated in our study are similar to the ones that have caused influenza outbreaks in poultry in Europe during the last seven years. This indicates that wild bird surveillance for influenza A virus can be of major value as a sentinel system to prevent outbreaks in domestic poultry. Studies on Black-headed Gulls (Larus ridibundus) revealed a previously unknown subtype, H16. This finding widened the spectra of known influenza A virus subtypes in nature. Influenza A virus was also isolated in samples from Guillemots (Uria aalge) in the Baltic Sea. This was the first time influenza A virus was isolated from this species in Europe. The isolated virus strains contained a mix of genes, some of which must have been derived from influenza A virus strains present in the North American bird population. This finding proves that limited exchanges between the virus strains present on the American and the Eurasian continents exist, which is of concern for evaluating the risk of spread of highly pathogenic virus strains by wild birds to the Americas. / Influensavirus är RNA virus och indelas i olika typer och subtyper. Influensa A virus orsakar sjukdom hos ett flertal fågel- och däggdjursarter. Vilda fåglar anses utgöra den viktigaste reservoaren för influensa A virus. Hos människa orsakar influensa A virus årliga epidemier av luftvägssjukdom med hög sjuklighet och stora ekonomiska konsekvenser för samhället. Eftersom frekventa mutationer orsakar ändringar i virusets ytstrukturer krävs årlig vaccination med nytt anpassat vaccin för att ge skydd mot sjukdom. Det finns många olika subtyper av influensa A virus. Dessa karaktäriseras med två av virusets ytstrukturer; hemagglutinin och neuraminidas, vilket till exempel skrivs H5N1. Virus av olika subtyper kan rekombinera och på så sätt skapa nya varianter. Om en subtyp som tidigare ej cirkulerat bland världens befolkning orsakar ett utbrott kan detta leda till en världsomfattande epidemi, en så kallad pandemi. Pandemier har drabbat mänskligheten med viss regelbundenhet genom historien och haft förödande konsekvenser. Till exempel orsakade pandemin ”Spanska sjukan” under åren 1918-1920 mer än 50 miljoner dödsfall. Influensa A virus orsakar också förödande utbrott i fjäderfäbesättningar. Virus av vissa subtyper kan mutera till högpatogena varianter och orsaka så kallad högpatogen aviär influensa. Dessa högpatogena varianter kan även överföras till och orsaka sjukdom hos människa och andra djur vilket varit fallet under det pågående utbrott av H5N1 som startade i sydöstra Asien 2003. Alla kända subtyper av influensa A virus har isolerats i material från vilda fåglar vilka lever i vattenmiljö, framförallt från änder. Dessa arter anses därför utgöra influensavirusets reservoar i naturen. Hos änder orsakar viruset framförallt en subklinisk infektion i gastrointestinalkanalen och sprids genom faekal-oral överföring via vatten i vilket viruset kan förbli aktivt en längre tid. Det finns fortfarande många obesvarade frågor angående influensa A virus ekologi bland vilda fåglar. I denna avhandling presenteras fem artiklar som tillför ny kunskap inom detta område. I avhandlingen styrks bevisen för att vilda änder utgör virusets reservoar i naturen dels genom en metaanalys av samtliga publicerade data rörande fynd av influensa A virus hos vilda fåglar, dels med hjälp av data från fyra års provtagning från flyttande vilda änder vid Ottenby fågelstation. Resultaten påvisar temporala skillnader i influensvirusets prevalens i den västeuroasiatiska andpopulationen jämfört med den nordamerikanska. Prevalensen i den västeuroasiatiska andpopulationen är hög under perioden augusti till december och i viss mån även under våren. Dessa fynd talar för att influensavirus kontinuerligt cirkulerar i andpopulationen. Under studien av förekomsten av influensa A virus hos änder isolerades enbart olika lågpatogena subtyper. Subtyperna H5 och H7 var vanligt förekommande. Dessa subtyper är benägna att utvecklas till högpatogena varianter om de sprids till fjäderfäbesättningar med svåra konsekvenser som följd. Genom studier av virus släktskap visas att de virus vi isolerat från vilda änder är snarlika de som orsakat utbrott bland fjäderfä i Europa under de senaste sju åren. Detta styrker värdet av att övervaka förekomsten av influensavirus hos vilda fåglar för att på så sätt förhindra utbrott av sjukdom bland fjäderfä. Undersökning av prover från skrattmås (Larus ridibundus) ledde fram till upptäckten av en helt ny subtyp av influensavirus; H16. Därmed utvidgades spektret av kända subtyper i naturen. Influensa A virus isolerades från sillgrisslor (Uria aalge) i Östersjön vilket inte tidigare gjorts hos denna art i Europa. Dessa virus innehöll gener från både nordamerikanska och euroasiatiska fågelpopulationers virus. Det visar att det finns ett utbyte av virus mellan fågelpopulationerna på de skilda kontinenterna. / On the day of the defence data the status on article IV was Submitted and the title was "Multi-year surveillance of influenza virus type A in migratory waterfowl in northern Europe".

Although differences between selective pressures in hatcheries and streams have been theorized to cause genetic divergence between hatchery and wild salmonids, evidence of this is lacking. This study was initiated to document the presence or absence of genetic change in hatchery and wild stocks by characterizing genetic traits in fish of various life history stages within a single generation. Nine biochemical traits (enzyme loci) and 12 meristic traits were characterized for adult fall chinook and one or more juvenile stages of their progeny of the 1984 brood year. Study groups consisted of hatchery-reared and naturally-reared subunits of populations in two tributaries of the lower Columbia River: Abernathy Creek and the Lewis River. Parents of both groups from Abernathy Creek were primarily of hatchery origin, whereas parents of both groups from the Lewis River were primarily of wild origin. The experimental design thus included reciprocal comparisons of hatchery and wild-reared groups from each of two stocks: one that has been propagated under hatchery conditions for at least five generations and one that has evolved in a stream environment. Both biochemical and meristic traits varied among adult and juvenile stages within hatchery and wild groups. Changes in some of these traits appear to have been caused by natural selection. This was true even for Abernathy hatchery and Lewis wild groups, which have been in the same environment for many generations. The direction and/or degree of change in some biochemical and meristic traits differed between hatchery and wild groups from a given stream, suggesting that selective pressures of the hatchery and wild environments differed in those cases. However, it could not be determined from these data whether the observed divergence of traits reflects general differences in hatchery and stream environments, or if it reflects population-specific responses to site-specific environmental conditions. The extent to which patterns of genetic change within a single generation might vary among year classes or generations is likewise unknown. Evidence of temporal changes in biochemical and meristic traits of hatchery and wild fish within a single generation has important implications regarding the use of those traits to characterize stocks. Assumptions of temporal stability of biochemical or meristic traits within or between year classes should be applied with caution. Sampling strategies of studies involving these characters should account for the possibility of temporal heterogeneity. Finally, these results suggest that workers using allozymes as genetic tags should test the assumption of selective neutrality of the particular allozyme markers being used. / Graduation date: 1991

Newly emerged steelhead fry (Oncorhynchus mvkiss) of hatchery and wild origins were studied in laboratory stream channels and natural streams. Objectives of the study were to determine if and how earlier emerging hatchery fry influence the emigration, realized densities, growth, habitat use, social structure, and activity patterns of localized populations of wild steelhead fry when the hatchery fry have a competitive advantage conferred by larger size and prior residence. During 1986 and 1987, the above variables were observed daily among hatchery and wild steelhead fry in laboratory stream channels for 8 weeks following emergence in June. The habitat use and social activities for fry of both origins were observed weekly in natural stream reaches from June through August in 1987 to corroborate lab findings. In lab channels, both hatchery and wild fry received 2 treatments: living alone (allopatry) and living together (sympatry). In the lab, fry of hatchery origin emerged 7 to 10 d prior to wild fry and remained larger in size during the 8 weeks of study both years. In natural stream reaches, fry of each origin were observed only in allopatric situations. Wild fry in the field emerged from natural redds while hatchery fry were released in stream reaches as unfed, newly emerged (swim-up) fry. Hatchery and wild fry in lab sections were found to be very similar in their emigration rates, distances to nearest neighbor, growth rates, and use of habitat. Both fry types, regardless of treatment or environment (lab or field), established similar stable social structure and used the same types of aggressive acts. Among all lab groups, once a fry became dominant, it retained that social status to the end of the study period. Significant differences (P<.05 both years) among comparison tests were: 1) in allopatric lab sections, wild fry maintained larger densities than hatchery fry, 2) in sympatry, hatchery fry had a greater tendency to establish stable focal points and social hierarchies more readily, defend larger areas, have better condition, prefer pools with overhead cover more frequently, be more aggressive, and reach stable densities more quickly than the wild fry, 3) fewer hatchery fry in sympatry maintained nomadic positions than wild fry in both treatments, 4) in sympatry, hatchery fry directed more acts of overt aggression toward wild fry than other hatchery fry, 5) wild fry in sympatry usually used defensive or less offensive acts of aggression when interacting with other fry, 6) fry of both origins in natural stream reaches maintained greater distances to their nearest neighbor than fry in allopatric lab sections, 7) dominant hatchery fry in both treatments maintained larger focal areas than subdominant fry, 8) hatchery fry maintained longer lengths than wild fry through the duration of the study, and 9) hatchery fry were more aggressive in sympatry than in allopatry. Potential differences (P<.05 in one year and P<.1 in the other year) were: 1) wild fry in sympatry had lower realized densities, maintained smaller focal areas, had greater proportions of nomadic individuals, and established stable social hierarchies slower than wild fry in allopatric lab sections, 2) wild fry in sympatry had poorer condition than all other fry groups in lab sections, 3) in sympatry, wild fry were the recipients of the majority of aggressive acts perpetrated by hatchery fry and other wild fry and usually assumed the subordinate positions within the social hierarchy, 4) all fry in the lab showed a high preference for pools with overhead cover and low preference for gravel and fines and run areas, and 5) wild fry in allopatric lab sections were more socially active than hatchery fry while the reverse was observed in the natural streams. Any influences that could be attributed to inherent differences between stock origins were probably masked by size differences between fry types. The study would have been more complete had I included sympatric lab sections where wild fry emerged first and where fry types emerged simultaneously, and sympatric reaches in natural streams. Results were further confounded by the limited number of wild adults used for broodstock in the lab segment of this study. Progeny produced from so few adults (5 adults of each sex each year) would have very limited genotypic variation compared to what occurs in natural streams. This may partially explain why some findings from lab sections and natural stream reaches differed. Likewise, genotypic expression among wild fry in lab sections may have varied greatly between years. This could explain differences found between years in behavior of wild fry in similar lab treatments. Although this study does not simulate all possible scenarios, results support suspicions that introductions of hatchery fry of larger size and earlier emergence into streams containing wild stocks could disrupt the social structure and negatively influence the realized densities, spatial distribution, growth, and behavior of wild juveniles in recipient streams. / Graduation date: 1991

Anthracnose, caused by <i>Colletotrichum truncatum</i>, is a major disease of lentil (<i>Lens culinaris</i> Medik.) on the Canadian prairies. Resistance to the more virulent race Ct0 of the pathogen is extremely rare within the <i>L. culinaris</i> gene pool thus resistance is being introgressed from <i>Lens ervoides</i>. The overall hypothesis of this project was that resistance derived from <i>Lens ervoides</i> accession L-01-827A would provide uniquely low levels of resistance to <i>Colletotrichum truncatum</i>. Individual studies were performed to: i) test field-resistance of interspecific lines; ii) develop a cutting-method to generate replicated phenotypic data on single plants; iii) study the deterioration of resistance between the juvenile phase (JP) and adult phases (AP) in the line CDC Redberry; and iv) examine genetic control of resistance in a susceptible and partially resistant background. Field evaluations of <i>L. culinaris x L. ervoides</i> lines indicated resistance genes from the wild species will provide unprecedented levels of disease control and heritability of the trait was estimated to be moderate. Single plant phenotyping of resistance is unreliable, thus the feasibility of using clonal propagation of individual plants to generate replicated ratings was evaluated. Results showed consistency in segregation ratios between cutting and seedling-derived plants of the same population, thus the method was utilized for testing of genetic control of segregating populations. CDC Redberry showed varying disease levels depending on the age of the plants with resistance acquisition in the JP that decreased as the plants proceeded through the AP. The F₁, F₂ and F_2:3 generations of two introgression populations were tested for resistance. Resistance from <i>L. ervoides</i> to both races of <i>C. truncatum</i> appeared to be due to the same gene(s) or from the same linkage block in both populations. Models for genetic control were consistent between F₂ and F_2:3 generations, however were different between the populations depending on whether the interspecific line was crossed into a susceptible or partially resistant <i>L. culinaris</i> background. However duplicate recessive epistasis seemed to control susceptibility in the susceptible background and when JP resistance was not a factor. There were significant differences between the JP and AP on more than a third of the F₂s tested supporting different resistance gene action based on growth phase. Resistance in the JP seemed to be due to dominant and recessive epistasis. It was postulated that the the populations may have had segregation distortion commonly found in interspecific populations, thus the previously described genotypes for cotyledon colour and albino plants were used to test the hypothesis of segregation distortion. Segregation of these traits were found to be similarly distorted to previously reported interspecific <i>L. culinaris x L. ervoides</i> populations supporting the hypothesis that continued segregation distortion was found in the introgression populations. Overall, it was found that resistance derived from <i>L. ervoides</i> accession L-01-827A is a highly effective source for the lentil breeding program, however the result suggest more than one backcross to <i>L. culinaris</i> be necessary to properly integrate the resistance genes to eventually obtain fully fertile, adapted lentil cultivars.

Anthracnose, caused by <i>Colletotrichum truncatum</i>, is a major disease of lentil (<i>Lens culinaris</i> Medik.) on the Canadian prairies. Resistance to the more virulent race Ct0 of the pathogen is extremely rare within the <i>L. culinaris</i> gene pool thus resistance is being introgressed from <i>Lens ervoides</i>. The overall hypothesis of this project was that resistance derived from <i>Lens ervoides</i> accession L-01-827A would provide uniquely low levels of resistance to <i>Colletotrichum truncatum</i>. Individual studies were performed to: i) test field-resistance of interspecific lines; ii) develop a cutting-method to generate replicated phenotypic data on single plants; iii) study the deterioration of resistance between the juvenile phase (JP) and adult phases (AP) in the line CDC Redberry; and iv) examine genetic control of resistance in a susceptible and partially resistant background. Field evaluations of <i>L. culinaris x L. ervoides</i> lines indicated resistance genes from the wild species will provide unprecedented levels of disease control and heritability of the trait was estimated to be moderate. Single plant phenotyping of resistance is unreliable, thus the feasibility of using clonal propagation of individual plants to generate replicated ratings was evaluated. Results showed consistency in segregation ratios between cutting and seedling-derived plants of the same population, thus the method was utilized for testing of genetic control of segregating populations. CDC Redberry showed varying disease levels depending on the age of the plants with resistance acquisition in the JP that decreased as the plants proceeded through the AP. The F₁, F₂ and F_2:3 generations of two introgression populations were tested for resistance. Resistance from <i>L. ervoides</i> to both races of <i>C. truncatum</i> appeared to be due to the same gene(s) or from the same linkage block in both populations. Models for genetic control were consistent between F₂ and F_2:3 generations, however were different between the populations depending on whether the interspecific line was crossed into a susceptible or partially resistant <i>L. culinaris</i> background. However duplicate recessive epistasis seemed to control susceptibility in the susceptible background and when JP resistance was not a factor. There were significant differences between the JP and AP on more than a third of the F₂s tested supporting different resistance gene action based on growth phase. Resistance in the JP seemed to be due to dominant and recessive epistasis. It was postulated that the the populations may have had segregation distortion commonly found in interspecific populations, thus the previously described genotypes for cotyledon colour and albino plants were used to test the hypothesis of segregation distortion. Segregation of these traits were found to be similarly distorted to previously reported interspecific <i>L. culinaris x L. ervoides</i> populations supporting the hypothesis that continued segregation distortion was found in the introgression populations. Overall, it was found that resistance derived from <i>L. ervoides</i> accession L-01-827A is a highly effective source for the lentil breeding program, however the result suggest more than one backcross to <i>L. culinaris</i> be necessary to properly integrate the resistance genes to eventually obtain fully fertile, adapted lentil cultivars.

No description available.

Sustainable management of wildlife populations relies on accurate estimates of population size as harvest recommendations are dependent on estimates of sustainable surplus. Techniques for surveying wild turkey populations in Texas are constrained by land access issues, requiring that new methods be developed for population monitoring. I evaluated a combined approach using patch-occupancy modeling at broad spatial scales and intensive double observer roost surveys at local scales to estimate Rio Grande wild turkey (Meleagris gallapavo intermedia) distribution and abundance. I flew replicated aerial surveys during 2007 and 2008 to evaluate distribution of Rio Grande wild turkeys in the south Texas Coastal Sand Plains. I used a double observer approach to estimate local scale abundance. I used a single observer approach to estimate temporal variation in roost use. Detection probabilities from aerial surveys ranged between 0.24 (SE = 0.031) and 0.30 (SE = 0.083). Spatial parameters that influenced distribution of wild turkeys included size of suitable roosting habitat patches and distance to the nearest suitable roosting habitat. I conducted 100 inter-patch double observer roost counts, with counts ranging between 0 to 183 individuals. Average detection probabilities for observers were ~0.90. Roost level occupancy was ~0.84 with detection probabilities between 0.69 (SE = 0.107) and 0.79 (SE = 0.091). Based on my results, aerial surveys combined with local abundance estimation may be one viable alternative to monitor turkey populations over large spatial scales, by reducing overall survey effort without loss of estimated precision.

Rio Grande wild turkey (RGWT) abundance in portions of the Edwards Plateau has declined steadily since the late 1970s as compared to other areas of the Edwards Plateau where populations have exhibited no trend. The reasons for this decline remain unclear. Possible factors include changes in habitat, and increased human population. The overall objective of this study was to identify landscape changes and habitat characteristics that affect RGWT populations using spatial analysis and modeling at multiple spatial scales. Specific objectives for this study included the quantification of flood-induced landscape changes between 1972 and 1995 along the Medina River bottomlands and their impact on RGWT habitat, the quantification of landscape characteristics of stable and declining study sites in the Edwards Plateau, and the development and evaluation of a GIS-based habitat-suitability model for female RGWTs during the breeding season that will allow the assessment of the spatial distribution of adequate habitat in the Edwards Plateau.The analysis of the landscape characteristics along the North Prong Medina River due to flooding in 1978 had a negative impact on RGWT habitat. Changes in the spatial distribution of woody cover in the bottomlands and the removal of woody cover along riparian zones most likely limited habitat use and dispersal of RGWT along the North Prong Medina River. The analysis of landscape characteristics in sites with stable and declining of RGWTs populations showed that disturbance and a high proportion of woody cover were important factors influencing RGWT populations in areas where turkey numbers had declined. Landscape attributes were used as habitat variables to develop a habitat-suitability model for female RGWTs during the breeding season. The model performed well in characterizing high-suitability habitat for adult female RGWT during the breeding season in the study areas. The use of two scales relevant to RGWT provided important information about the high-suitability areas for female RGWT in stable and declining sites in the Edwards Plateau.

Thesis (M.S.)--University of Missouri-Columbia, 2003. / Typescript. Includes bibliographical references (leaves 101-107). Also available on the Internet.