MENVSC (Geography) / Department of Geography and Geo-Information Sciences / The Limpopo lowveld is at risk of floods from tropical storms from the South West Indian Ocean. The flood risk is particularly high in low altitude areas with poor infrastructure and limited resources. This study assessed flood vulnerability and adaptation at Lenyenye, Ga-Kgapane Masakaneng and Nkowankowa Section B and C within the Mopani District in the Limpopo Province of South Africa. The research objectives were to establish the determinants of flood vulnerability, assess the levels of flood vulnerability and the community‟s coping strategies. A quantitative survey approach was employed using questionnaires which were administered to affected households to identify determinants of flood vulnerability, indicators and coping strategies by communities. A qualitative survey was also undertaken to supplement the information obtained from the quantitative survey. Key informant interviews were conducted with disaster management authorities in the study area to provide information on indicators, flood experience, adaptation and mitigation measures. Field observations were undertaken to observe the physical landscape and flood impacts. Secondary data were acquired through records, maps, Census 2011 and from the South African Weather Service. Collected data were imputed into the flood vulnerability index to measure the level of flood vulnerability. The results of this study will contribute to flood disaster risk reduction in the lowveld. The results indicate that flood vulnerability in the study areas is determined by dwelling quality, poor or lack of drainage, education levels, employment status, rainfall amount and topography. The calculated flood vulnerability levels in the three case study villages indicate that Ga-Kgapane Masakaneng, Lenyenye and Nkowankowa Section B and C have a „vulnerability to floods‟ level. However, the FVI also showed that the economic aspect scored a high vulnerability to floods in Ga-Kgapane. In Nkowankowa Section B and C; physical component obtained a „high vulnerability to floods.‟ Major coping strategies in the three case villages were: making „Le-guba‟ around houses; sand-bags; making a furrow and channel around houses and on roads; temporary relocation and lastly relocating to a safer area. Key recommendations are public awareness; integrating modern mitigations with local knowledge; development of programs to ensure resilience through incorporation of (Integrated Development Planning) IDP and flood management and flood early warning system.

Lead is a toxic heavy metal associated with adverse health effects ranging from developmental neurotoxicity to reproductive effects. While lead affects people of all ages, infants and children are the most vulnerable and susceptible to the neuro-developmental effects of lead exposure. Maternal blood lead concentrations that do not produce clinical toxicity on pregnant women have been linked to adverse offspring development. Observed reproductive effects to low lead levels during pregnancy include the risk of spontaneous abortions, effects on birth weight and preterm birth. There are particular concerns with regard to reductions in IQ scores. Research evidence suggests that an incremental increase in blood lead levels of 1 µg/dL is associated with approximately 1 IQ point deficit. Of particular concern is that currently no threshold has been observed or exists for developmental neurotoxicity to the chronic low lead exposures levels. While the developed countries have built evidence for lead exposure sources, have identified the most vulnerable groups to lead exposure, and have instituted control actions for lead exposure, it is not the case in developing countries such as Botswana. Currently, there is very little knowledge about the potential sources of lead exposure among different population groups not only in Botswana but also in most developing nations. There is also an evident limited knowledge on the behaviours and practices of different population groups that could potentially expose them to lead in developing countries. This thesis explores the following questions: i) Are there specific risk behaviours and practices peculiar to pregnant women in Botswana that could potentially expose them to lead? ii) What are the environmental lead concentration levels and their potential to expose pregnant women? iii)What are the blood lead concentrations at each stage of pregnancy and after delivery in Botswana and, iv) Can we use the information from these three questions to predict lead exposure levels during pregnancy and after delivery? v) Can we use the new information to a develop a policy dissemination brief to inform policy on lead exposure sources in Botswana, develop guidelines for health professionals for assessing and screening lead exposure levels during pregnancy and after delivery, develop an awareness leaflet for lead education? To address the specific risk behaviour and practices of pregnant women, a comprehensive validated risk assessment questionnaire was administered among 142 pregnant women during the first trimester of pregnancy (defined as 8-12 weeks) in four villages of different geographical settings and nomenclature (small/rural, major and semi urban). For purposes of this work the validation process involved obtaining information (from experts in the field and communities) relevant to the purposes of the study and to confirm that the tools employed for collection of data in all trimesters were suitable in terms of both construct and content. Data was collected between September 2009 and February 2010. To address potential environmental sources of lead exposure during pregnancy soil (n=28), water (n=28) and traditional cosmetic clay - letsoku (n=3) samples were collected in November 2010, February 2011 and May 2011 from the homes and in the vicinity of the study population to determine lead concentrations. To know baseline blood lead levels at each stage of pregnancy, blood samples were collected from September 2009 to February 2011 from pregnant women between weeks 8-12 (first trimester, n=137), 20-24 (second trimester n=126) and weeks 34-36 (third trimester n=106). Blood lead levels of women who completed the entire study from trimester on until after delivery (n=63) were then used to construct blood lead prediction models using statistical models. Pregnant women in the study area ingested non-food items such as soil, match sticks, pencil, chalk and animal feed such as bone meal (86%). Women applied used and unused car oils (in particular brake fluid) and other harmful substances for “treatment of skin conditions and for beautification purposes (74%). Older women (defined as 35 years in this study) were at a significantly higher risk to ingest soils (p<0.01). Mean (±SEM) lead concentrations in water exceeded the WHO drinking water quality standards nineteen fold (0.19±0.019 ppm (n=28) Major villages, had significantly higher Pb concentrations (p<0.05) in soils and water compared to small villages. Mean blood lead levels (±SEM) for the first, second and third trimesters were 1.96(±0.14)µg/dL, 2.49(±0.17) µg/dL, 2.66(±0.19) µg/dL respectively. Blood lead levels increases from the first to third trimester ranged from 1.6-5%. Blood lead concentrations significantly differed among locations (p<0.01). The highest concentrations were observed in women from smaller villages that were poorer (p<0.02). Pica, multiple risk behaviours/practices (engaging in two or more risk behaviours/practices), trimester of pregnancy, poor food supplementation and diet were predictors of blood lead levels ≥ 2µg/dL. There was a dose response relationship between supplement intake and an increase in blood lead levels. These findings suggest that pregnant women and their unborn babies could potentially be exposed to lead because of the environment in which they live, their economic status, lifestyle, behaviors and practices. Drinking water is a potential threat for lead exposure, not only among pregnant women, but other vulnerable groups such as infants and children. This study is the first in Botswana and one of the few in Africa to investigate lead exposure sources at each stage of pregnancy and after delivery. It is also the first to identify new potential lead exposure behaviors and practices such as the application of auto oils by pregnant women for treatment of skin diseases. The findings suggest the need to train health workers and equip them with the skills and knowledge to assess and screen women who could potentially be exposed to lead. Further, pregnant women need to be sensitized on potential lead exposure sources, to prevent lead poisoning. This study has been able to use the results to develop a policy brief for disseminating the results to decision makers, guidelines for utilization by health workers to screen lead exposure levels and an awareness leaflet for pregnant women. These have been validated and pretested at community and Government levels. / Thesis (PhD)--University of Pretoria, 2013. / School of Health Systems and Public Health (SHSPH) / Unrestricted

The most commonly used exposure assessment for heat stress is based on Wet Bulb Globe Temperature (WBGT), and the limits are based on empirical relationships that demonstrate a sustainable exposure. The ISO 7933 (2017) describes Predicted Heat Strain (PHS), which is a rational model for heat stress assessment that is used to assess time-limited exposures. Investigators have examined PHS validity under a variety of time-limited conditions. The purpose of this paper is to evaluate if PHS can predict sustainable exposures. The data used for this study were from two previous heat stress studies using a progressive heat stress protocol. The studies included 29 participants wearing four different ensembles (woven clothing, particle barrier, water barrier, and vapor barrier coveralls) at three levels of metabolic rate and relative humidity. Each trial provided data for a fully compensable (sustainable) exposure and an uncompensable (time-limited) exposure. The heat stress data from each condition were used to see if PHS demonstrated a steady-state response indicating a sustainable exposure. From the analysis, the sensitivity and specificity respectively for the ensemble types were: 1.00 and 0.18 for woven clothing, 0.95 and 0.60 for particle barrier, 0.91 and 0.83 for water barrier, 0.91 and 0.80 for vapor barrier, and 0.94 and 0.65 for all ensembles. The data show that while the sensitivity of PHS (correctly identifying unsustainable conditions) is good for the different ensembles, PHS specificity (correctly identifying sustainable exposures) was weak. From an occupational health and safety perspective, using PHS to identify sustainable exposures leads to protective decisions.

Master thesis touches upon the interesting spheres of risk classification, measurement and management of financial institutions. Modern banks have numerous credit risk measurement models at their disposal. However, agreement about performance of those models is not that unanimous and to some point the models are blamed for breaking out of 2007 financial crisis. In the theoretical part of the thesis we provide survey of risk measurement practices in banks. We investigate the main types of risk of banks in their day-to-day activities. Special focus is paid on the credit risk and on the models and techniques of its measurement; Practical part of thesis then contains construction and accuracy estimation of particular credit-risk-model (Altman Z-score). In it we construct and compute Altman Z-score for sample of firms from two chosen sectors in United Kingdom. Main goals of the work are a) testing accuracy of the model by comparing its outputs to real development, and b) econometric testing of the specification of the model itself.

Prenatal alcohol exposure (PAE) is associated with a range of physical, growth, and neurobehavioral deficits characteristic of individuals with fetal alcohol spectrum disorders (FASD). Although declarative memory impairment is a key feature of the neurocognitive profile of FASD, the mechanisms underlying this deficit require further clarification. The aim of this cross-sectional research was to examine, both directly and indirectly (via bottom-up and top-down processes), a critical cognitive mechanism that supports successful declarative memory functioning (viz., memory encoding), in children with FASD. Data were collected from a sample (N = 88) of South African children with and without PAE. In Study I, I used a blocked design functional magnetic resonance imaging (fMRI) paradigm to investigate neural activation during visual perception, a lower-order cognitive process essential to memory encoding. The task elicited bilateral category-specific activation during the visual perception of objects and scenes in all participants. The absence of between-group differences suggests that functional recruitment of brain regions during basic visual perception is less susceptible to the effects of PAE than during higher-order processes supporting memory encoding. In Study II, I used an event-related fMRI paradigm to investigate neural activation during memory encoding itself. All participants demonstrated similar memory performance accuracy and recruited extensive bilateral networks during memory encoding. However, participants with a diagnosis of fetal alcohol syndrome (FAS) or partial FAS (PFAS) activated additional regions associated with attentional function. Within the FAS/PFAS group, higher exposure levels were associated with smaller activation increases in the parahippocampal gyri and greater activation increases in the right hippocampal formation during encoding. Data from this study therefore suggest that children with FAS/PFAS recruited more extensive neural resources to support successful memory encoding during this task. In Study III, I used a behavioral source memory paradigm to investigate higher-order executive processes essential for memory encoding. Despite similar recognition accuracy across all diagnostic groups, participants in the FAS/PFAS group showed impaired memory for source details. This pattern of impairment was only partially mediated by working memory performance. These three studies provide novel clarification of the neural and cognitive mechanisms underlying declarative memory impairments in children with FASD.

Work in demanding postures is a known risk factor for work-related musculoskeletal disorders (MSDs), specifically work with elevated arms may cause neck/shoulder disorders. Such a disorder is a tragedy for the individual, and costly for society. Technical measurements are more precise in estimating the work exposure, than observation and self-reports, and there is a need for uncomplicated methods for risk assessments. The aim of this project was to develop and validate an iOS application for measuring arm elevation angle. Such an application was developed, based on the built-in accelerometer and gyroscope of the iPhone/iPod Touch. The application was designed to be self-exploratory. Directly after a measurement, 10th, 50th and 90th percentiles of angular distribution and median angular velocity, and percentage of time above 30°, 60°, and 90° are presented. The focused user group, ergonomists, was consulted during the user interface design phase. Complete angular datasets may be exported via email as text files for further analyses. The application was validated by comparison to the output of an optical motion capture system for four subjects. The two methods correlated above 0.99, with absolute error below 4.8° in arm flexion and abduction positions. During arm swing movements, the average root-mean-square differences (RMSDs) were 3.7°, 4.6° and 6.5° for slow (0.1 Hz), medium (0.4 Hz) and fast (0.8 Hz) arm swings, respectively. For simulated painting, the mean RMSDs was 5.5°. Since the accuracy was similar to other tested field research methods, this convenient and “low-cost” application should be useful for ergonomists, for risk assessments or educational use. The plan is to publish this iOS application on Apple Store (Apple Inc.) for free. New user feedback may further improve the user interface.

To investigate the effects of (1→3)-β-D-glucan after inhalation, animals were exposed to different forms of glucan and the number of lung lavage cells was determined 24 h after exposure. None of the different forms assayed caused any increase in cell numbers. In animals exposed to endotoxin, all types of cells were increased after 24 h. A simultaneous exposure to curdlan reduced this increase in a dose-related fashion. The results suggest that (1→3)-β-D-glucan-related acute injury to the lung is induced by mechanisms other than those induced by inflammagenic agents such as endotoxin.

Motivations for entertainment consumption are complex, involving moods, wants, anticipations, and several other factors. Typically, attempts at a theoretical understanding of motivations for media selection have centered on enjoyment, or hedonic, pleasure-seeking motivations. Recently eudaimonic, or truth-seeking, motivations have also been given much attention. Most investigations into hedonic and eudaimonic motivations for media consumption have conceptualized these motivations as traits, rather than as states. Recent research has used survey methods to determine that those with hedonic motivations tend to seek out entertainment that is hedonic in nature and feel more fun affect while watching, while those with eudaimonic motivations for consuming entertainment tend to seek out media that is eudaimonic and nature and experience more meaningful affect during viewing. This experiment successfully manipulated hedonic and eudaimonic states using clips with either hedonic or eudemonic tone from three different television shows, with hedonic or eudaimonic tone in clips significantly affecting participants\' reports of hedonic and eudaimonic states as well as meaningful and fun affect. The effects of clips with hedonic or eudaimonic tone on these responses were not moderated by typical hedonic and eudaimonic trait preferences. The experiment also examined the effect that clips with hedonic or eudaimonic tone might have on subsequent program selection by allowing participants to rank hedonic or eudaimonic clips for subsequent viewing. The hedonic or eudaimonic tone of the clips only predicted participants\' subsequent  preference for eudaimonic or hedonic clips in the case of one of the three shows in the study; this effect on subsequent clip preference was found with the show that generally elicted the strongest responses from viewers in other study measures. Implications for our understanding of television consumption motivations are discussed. / Master of Arts

Aflatoxin is a poisonous mold that has spread around the world, posing a threat to food security and the agricultural economy. A total of 4,5 million people are in danger of being exposed to aflatoxins on a long-term basis around the world. Acute toxicity is caused by consuming significant amounts of toxins in a short period of time, which can lead to death in the worst cases, whereas chronic toxicity is caused by consuming small quantities over a longer period of time, which can lead to low birth weight, immunosuppression, restricted growth in children, and liver cancer in the worst cases. The occurrence of aflatoxins in West Africa was recognized, studied, and investigated in this paper based on a literature review. The findings demonstrate that large levels of aflatoxins have been found in West African raw materials and food and the human body. Children under the age of five, pregnant women, and breastfeeding mothers are the most vulnerable to aflatoxins. Furthermore, the findings reveal that the majority of the population is unaware of aflatoxins and its health implications. Inadequate governmental systems, low societal development, lack of access to health care, a low educated population, climate variability and climate change, high levels of illiteracy, and a lack of laboratories are only a few of the many obstacles that the region has in limiting aflatoxins concentrations. Sorting procedures, the use of tarpaulins, and seminars have all helped raise awareness and knowledge and reduce contamination and consumption of aflatoxin-contaminated foods. This study argues that a multi-sectoral strategy is needed to promote food security and local education. Increased limitations and regulations and higher standards may be able to help limit aflatoxin contamination and exposure.

Understanding the neuronal basis of learning and memory is a fundamental problem in neuroscience. A leading theory, the origins of which date back to the beginning of the twentieth century, is that the neural basis for memory resides in engrams (also called memory traces), ensembles of cells that are activated during learning and reactivated during memory retrieval. Recent genetic tools have allowed researchers to visualize and manipulate memory traces in small brain regions; however, the ultimate goal is to analyze memory traces across the entire brain in order to better understand how memories are stored in neural networks and how multiple memories may coexist. In order to do so, methods and technologies need to be developed that allow labeling of engram cells throughout the brain, visualization of these cells, and automated quantification of cells in an anatomically precise manner. The first of these challenges has been addressed through the development over the past several years of different mouse models that permit the labeling of active cells throughout the brain at multiple time points. One of the most powerful models, the ArcCreERT2 mouse line, uses drug-induced genetic recombination to indelibly label cells throughout the brain in an activity-dependent manner. In this thesis, I present our work utilizing this model to solve the second and third challenges: imaging of brain-wide memory traces and automated quantification of labeled cells, as well as the application of these novel methods to understanding the engram network changes following fear extinction. Intact tissue clearing and imaging is a new and rapidly growing area of focus that holds great promise for enabling the brain-wide visualization of memory traces. We utilized the leading protocols for whole-brain clearing and applied them to the ArcCreERT2 mice. We found that CLARITY and passive clarity technique (PACT) greatly distorted the tissue, and immunolabeling-enabled three-dimensional imaging of solvent-cleared organs (iDISCO) quenched enhanced yellow fluorescent protein (eYFP) fluorescence and hindered immunolabeling. Alternative clearing solutions, such as tert-Butanol, circumvented these harmful effects, but still did not permit whole-brain immunolabeling. Clear unobstructed brain imaging cocktails and computational analysis (CUBIC) and CUBIC with Reagent 1A produced improved antibody penetration and preserved eYFP fluorescence, but also did not allow for whole-brain memory trace visualization. We developed CUBIC with Reagent-1A*, a modified CUBIC protocol that resulted in eYFP fluorescence preservation and immunolabeling of the immediate early gene (IEG) Arc in deep brain areas; however, optimized memory trace labeling still required tissue slicing into mm-thick tissue sections. Nonetheless, our data show that CUBIC with Reagent-1A* is the ideal method for reproducible clearing and immunolabeling for the visualization of memory traces in mm-thick tissue sections from ArcCreERT2 brains. Recent developments in brain-wide engram tagging strategies, primarily through the use of transgenic mouse models such as the ArcCreERT2 line, and whole brain imaging strategies, such as CLARITY, CUBIC, and iDISCO, have created the circumstances to, for the first time, be able to visualize throughout the brain neuronal activity that is directly linked to behavior. However, as noted above, quantifying and analyzing these brain-wide memory traces presents its own challenge, and widely applicable, readily accessible solutions to this problem have thus far been limited. Although a handful of freely available programs and suites do exist, such as CellProfiler and ClearMap, these are generally tailored to specific approaches, and in particular, no currently available solution exists for quantifying multi-labeled engram cells imaged in three dimensions along the coronal plane, a relatively common scenario that is sure to become even more prominent as greater adoption of the underlying technologies progresses. Using ImageJ and R, we developed an image analysis pipeline to solve precisely this problem. Our strategy allows for the segmentation of both the encoding and retrieval populations, including identification of the reactivated cells, and registration of segmented cells to an anatomical atlas in order to analyze cell activity in a region- and layer-specific manner. Post-traumatic stress disorder (PTSD) can develop following a traumatic event and results in heightened, inappropriate fear and anxiety. Approximately 8% of the US population suffers from PTSD, the main treatment for which is repeated exposure to triggering stimuli under controlled conditions. A better understanding of the neural circuits modified during this process would help advance therapeutic treatment for PTSD. We sought to determine the brain-wide neuronal activity changes underlying fear extinction, the best laboratory model of exposure therapy, by using the ArcCreERT2 x eYFP mice and our newly developed brain-wide segmentation and registration pipeline. ArcCreERT2 x eYFP mice were administered a 4-shock contextual fear conditioning (CFC) paradigm followed by either a 10-day extinction protocol or re-exposure to the aversive context without extinction. Following the final exposure session, mice were euthanized, and active cells were quantified throughout the brain using the pipeline. We found that fear learning leads to increased functional connectivity of amygdalar and hypothalamic regions, and extinction leads to a decentralization of the fear memory network and disengages the thalamus and striatal amygdala. Additionally, coordinated reactivation of the basomedial amygdala and secondary somatosensory cortex with frontal association regions are differentially modulated following extinction, and we identified the temporal association area and medial habenula as novel brain regions involved in modulating freezing behavior. In summary, in this thesis, we have developed a novel engram analysis pipeline and shown its potential for quantifying brain-wide memory traces. This is the first study to analyze brain-wide functional connectivity following fear learning and extinction of a recent fear memory, as well as the first study to analyze fear memory trace reactivation patterns across the brain and relate all three measures to behavioral output. This work both greatly enhances our understanding of the neural underpinnings of fear extinction and provides a toolset for readily exploring the neural underpinnings of other behaviors and types of associative memory.