Rhodococcus equi is a Gram-positive soil organism that causes an aggressive bronchopneumonia in foals and opportunistic infections in immuno-compromised humans. Virulent strains possess an 80 – 90 kb plasmid that encodes an immunogenic surface-located virulence-associated protein VapA. The virulence of the organism has been largely attributed to this protein since mutants lacking vapA are attenuated for virulence in mice. VapA is an unusual lipoprotein and existing evidence suggests its biogenesis may not involve normal lipoprotein processing. In order to understand the structure of VapA and other virulence associated proteins, their genes were cloned, expressed, purified and crystallised. VapG produce high quality crystals that diffracted to 1.8 Å. The structure was resolved to be a closed β-barrel with a long unstructured N-terminus which is similar to both VapB and VapD which have also recently been characterised.

This research focuses on iris recognition, the most accurate form of biometric identification. The robustness of iris recognition comes from the unique characteristics of the human, and the permanency of the iris texture as it is stable over human life, and the environmental effects cannot easily alter its shape. In most iris recognition systems, ideal image acquisition conditions are assumed. These conditions include a near infrared (NIR) light source to reveal the clear iris texture as well as look and stare constraints and close distance from the capturing device. However, the recognition accuracy of the-state-of-the-art systems decreases significantly when these constraints are relaxed. Recent advances have proposed different methods to process iris images captured in unconstrained environments. While these methods improve the accuracy of the original iris recognition system, they still have segmentation and feature selection problems, which results in high FRR (False Rejection Rate) and FAR (False Acceptance Rate) or in recognition failure. In the first part of this thesis, a novel segmentation algorithm for detecting the limbus and pupillary boundaries of human iris images with a quality assessment process is proposed. The algorithm first searches over the HSV colour space to detect the local maxima sclera region as it is the most easily distinguishable part of the human eye. The parameters from this stage are then used for eye area detection, upper/lower eyelid isolation and for rotation angle correction. The second step is the iris image quality assessment process, as the iris images captured under unconstrained conditions have heterogeneous characteristics. In addition, the probability of getting a mis-segmented sclera portion around the outer ring of the iris is very high, especially in the presence of reflection caused by a visible wavelength light source. Therefore, quality assessment procedures are applied for the classification of images from the first step into seven different categories based on the average of their RGB colour intensity. An appropriate filter is applied based on the detected quality. In the third step, a binarization process is applied to the detected eye portion from the first step for detecting the iris outer ring based on a threshold value defined on the basis of image quality from the second step. Finally, for the pupil area segmentation, the method searches over the HSV colour space for local minima pixels, as the pupil contains the darkest pixels in the human eye. In the second part, a novel discriminating feature extraction and selection based on the Curvelet transform are introduced. Most of the state-of-the-art iris recognition systems use the textural features extracted from the iris images. While these fine tiny features are very robust when extracted from high resolution clear images captured at very close distances, they show major weaknesses when extracted from degraded images captured over long distances. The use of the Curvelet transform to extract 2D geometrical features (curves and edges) from the degraded iris images addresses the weakness of 1D texture features extracted by the classical methods based on textural analysis wavelet transform. Our experiments show significant improvements in the segmentation and recognition accuracy when compared to the-state-of-the-art results.

This thesis investigates the application of multimodal metaphors in electronic Customer Loyalty Management Systems (e-CLMS) in terms of efficiency, effectiveness, user satisfaction, and understandability of the customisation tasks and information communicated. The potential of users developing loyalty as a result of better usability and user satisfaction is also accessed via questionnaires. The first experiment investigated issues of usability and the users' views of an e-commerce platform developed for these experiments using three conditions with three independent groups. A visual group (VICLMS, n=25) that was communicated information within the platform using text with graphics, a multimodal group (MICLMS, n=25) that usedrecorded speech, earcons and auditory icons and an expressive avatars group(AICLMS, n=25) that was predominantly communicated information using avatars. The second experiment evaluated three avatar-based multimodal conditions using a dependent group (n=50). This experiment evaluated user satisfaction, perceived convenience, enjoyment, ease of use and customisation, and successful completion of user tasks. The conditions were avatars with earcons (AEICLMS), avatars with auditory icons (AAICLMS) and avatars with both earcons and auditory icons (AICLMS).The use of expressive avatars in the e-CLMS interface contributed to the positive predisposition of usersto develop loyalty. Multimodal metaphors contributed more significantly to complex customisation tasks. A set of empirically derived guidelines and a validation approach is suggested for designing multimodal E-CLMS interfaces.

This thesis investigates empirically multimodal socially interactive e-commerce interfaces. The overall hypothesis is that multimodal social interaction will improve the usability of e-commerce interfaces and increase the user‘s feeling of social presence, decision making and product understanding when compared to an equivalent non-multimodal socially interactive interface. The investigation consisted eight conditions in three experimental phases. The first experimental phase investigated non-socially interactive, static-socially interactive, and interactive-socially interactive interfaces (three conditions) using an e-commerce platform with a dependent sample of users (n=36). The second experimental phase continued with the comparative evaluation of a further two conditions based on the results of the first phase. An audio and an avatar-based socially interactive conditions were evaluated with two independent groups of users (n=18 for each group). The third experimental phase investigated three socially interactive conditions. These were text with graphics, auditory stimuli, and avatars. The results demonstrate that socially interactive metaphors in e-commerce interfaces improved the ability of users to use presented information effectively, make decisions in comparison to non-social or static social interactive interfaces. An avatar-based socially interactive e-commerce interface improved the user‘s social presence. A set of empirically derived guidelines for the design and use of these metaphors to communicate information in a socially interactive atmosphere is also introduced and discussed.

Coastal ecosystems are undergoing unprecedented rates of environmental change. Many of these changes are anthropogenically-driven and linked to long-term, climate-related phenomena. This thesis focusses on ecological change in soft sediment intertidal habitats. One of the largest harbours in Europe, Poole Harbour, is used as a case study. It contains a variety of important habitats including intertidal mudflat and non-tidal saline lagoon. The two main themes of the thesis are 1) assessing the physical and ecological factors that determine benthic invertebrate abundance, distribution and community structure, which is examined at the scale of the whole harbour, and at the scale of individual habitats: an intertidal mudflat and a saline lagoon; and 2) predicting the response of an overwintering shorebird population, the pied avocet (Recurvirostra avosetta), to future environmental changes, such as sea-level rise and habitat loss. This is achieved by development of an individual-based model (IBM) and consideration of the species’ unique foraging behaviour. This study contributes to the understanding of the factors structuring soft sediment benthic communities, including the use of data from fine-scale hydrodynamic models. It offers a unique comparison of the spatial and temporal variables driving community structure of a saline lagoon and an intertidal mudflat. It also provides insight into the foraging ecology of the pied avocet at a level of detail that has not previously been considered, including a comparison of foraging behaviour in a tidal and non-tidal habitat, the importance of social foraging, and the novel application of an IBM to this species.

1. The functional response, i.e. the feeding rate as a function of prey density, links resource availability to population level demographic rates. The functional response is often central to behavioural-based population models that predict the effect of environmental changes on populations based on the assumption that each individual tries to maximise its fitness at all times. 2. The functional response is a well studied phenomenon for some groups of foragers and many mechanistic models of the functional response have been derived to account for variations in foraging behaviour. However, only a few functional response models have been applied to granivorous farmland birds and these models often have simplifying assumptions that are likely to be inappropriate for these foragers. Consequently, the functional response models that often lie at the core of behavioural-based population models are untested and quite possibly provide a poor prediction of the functional response. 3. The objectives of this thesis are to: (i) test the ability of functional response models to predict the functional response of granivorous birds; (ii) identify incorrect assumptions in these models; and (iii) derive new functional response models that better describe the observed behaviour of these foragers. 4. Here I have found that many of the underlying assumptions of simple functional response models are inappropriate for granivorous birds and several new mechanistic models have been derived that attempt to describe the behaviour more accurately. Vigilance was shown to frequently interact with incompatible behavioural processes (i.e. searching) and affect the feeding rate at low prey densities. Handling time was divided into two components (i.e. attack and process) and a simple mechanistic model was derived to predict the attack component of prey handling. The searching mode of granivorous birds was shown to be akin to a pause-travel forager at low prey densities and, lastly, habitat structure was shown to reduce the feeding rate by restricting the searching area. 5. These finding will help in the development of more biologically 'realistic' functional response models and provide a more accurate prediction of the functional response when used to predict the effect of environmental change on the intake rate of foragers. z

To predict the impact of future disturbances such a climate change and introduction of non-native species on ecosystems, it is important to understand how disturbances may affect community composition. This is inherently diffcult since species may be expected to respond differently to disturbances such as elevated temperatures or the introduction of a new species. Furthermore, since the species in an ecosystem are interlinked by energy, nutrient and information transfers, disturbances may be amplified or absorbed, depending on the nature of the disturbance and the resilience of the ecosystem. Some species have a disproportionate effect on ecosystem function and are often referred to as keystone species. By definition the loss of a keystone species causes a catastrophic change in community composition. Therefore, the identification of keystone species could help to target conservation efforts more effciently. A dynamical food web model, representative for a chalk stream (the River Frome, Dorset) was developed and manipulated. Changes in community composition and biodiversity were assessed. For the identification of keystone species each species node was removed in turn. Although impacts were found, particularly after the removal of important prey nodes and top predators, no catastrophic shift was observed and, consequently, no keystone species were identified. Impacts of species introductions were assessed by adding representative model species to the food web. The largest impact was observed after the addition of a small competitor at intermediate trophic level. The addition of a top predator had moderate impact, whereas no negative impact was found after the addition of a larger bodied species at intermediate trophic level. Possible impacts of climate change, specifically elevated temperatures, were assessed by increasing the metabolic rates of the species nodes. No impacts were found, when energy inputs were raised accordingly, but severe impacts, were observed when energy inputs were restricted. In general, the ecosystem was considered fairly resilient to most of the tested disturbances, possibly owing to the high natural variability of the community. The findings of current study suggest that rather than focusing conservation efforts on single species, the focus should be on 'keystone structures' that maintain high ecosystem resilience.

Plants are of fundamental importance to the structure, functioning and service provision of many ecosystems. However, herbivores can have negative ecological and socioeconomic effects on plant communities through consumption, trampling and alteration of nutrient cycles. In this thesis I address a particular herbivore-plant interaction: the grazing of plants in chalk river catchments, principally the submerged macrophyte water crowfoot (Ranunculus penicillatus ssp. pseudofluitans (Syne) S.D. Webster) and terrestrial pasture grass species, by flocks of non-breeding mute swans (Cygnus olor Gmelin, 1789). This research was carried out over two years in the River Frome catchment (Dorset, UK). Based on a meta-analysis of previous waterfowl grazing studies I show that waterfowl biomass density (kg ha-1) rather than individual density (ind. ha-1) is a better predictor of reductions in plant standing crop. Most studies to date have analysed such reductions using only individual densities, despite large between-taxa variation in waterfowl body mass, diet and intake rates. I quantified the abundance, species richness, evenness, flowering and dominance of the chalk river aquatic plant community in relation to biotic and abiotic factors during the growth-, peak-, and recession-phases of the growth cycle. The relative importance of herbivory, riparian shading, water temperature and distance downstream varied between different phases of the plant growth cycle, highlighting the importance of seasonal patterns in regulation of plant community structure. The River Frome swan population varied seasonally, being highest in the winter. The population was dominated by non-breeding adults and juveniles that lived in flocks. These flocks exhibited strong seasonal habitat switches between terrestrial pasture in winter and spring, and river in summer and autumn. I provided evidence that this switch was linked to the seasonal decrease in water velocity between spring and summer, which reduced the metabolic costs of river feeding and increased the relative profitability of aquatic food resources. I used a mathematical population model and an individual-based behavioural model respectively to explore two management options for the alleviation of the swan grazing conflict in chalk rivers: population control and habitat alterations. Population control measures, such as clutch manipulations, fertility control, culling or translocations, were predicted to be unsuccessful except at impractically high levels of management effort, due to the effects of immigration and high survival rates in offsetting removed eggs or individuals. Habitat alterations, in particular the narrowing of river channels to cause a local increase in water velocity and thus swan foraging costs, are more promising management options as they require lower management effort, are less ethically controversial, and address the fundamental reason why swans select their food resources, the rate of net energy gain (‘profitability’).

Among British birds, a number of woodland specialists have undergone a serious population decline in recent decades, for reasons that are poorly understood. The Marsh Tit is one such species, experiencing a 71% decline in abundance between 1967 and 2009, and a 17% range contraction between 1968 and 1991. The factors driving this decline are uncertain, but hypotheses include a reduction in breeding success and annual survival, increased inter-specific competition, and deteriorating habitat quality. Despite recent work investigating some of these elements, knowledge of the Marsh Tit’s behaviour, landscape ecology and habitat selection remains incomplete, limiting the understanding of the species’ decline. This thesis provides additional key information on the ecology of the Marsh Tit with which to test and review leading hypotheses for the species’ decline. Using novel analytical methods, comprehensive high-resolution models of woodland habitat derived from airborne remote sensing were combined with extensive datasets of Marsh Tit territory and nest-site locations to describe habitat selection in unprecedented detail. Further fieldwork established the causes and frequency of breeding failure at the local population scale, and dispersal distances and success were quantified. Information from these studies was used to inform national-scale spatial analyses of habitat distribution in relation to the pattern of range contraction for the Marsh Tit and two other woodland bird species. The combined results indicate that Marsh Tits require extensive areas of mature woodland in order to accommodate large territories and short dispersal distances, with greatest selection for a woodland structure encompassing a tall, near-closed tree canopy and extensive understorey. The evidence suggests that nest-site competition, nest predation or deteriorating habitat quality have not driven the population decline. However, reduced connectivity between woodlands in the landscape, possibly due to hedgerow loss, may have interacted with increased mortality to precipitate population declines and local extinctions where habitat fragmentation was relatively high. The potential causes of increased mortality are discussed, along with priority areas for future research and the wider possible applications of remote sensing techniques in the field of woodland bird research.

Exposure to toxic elements is a significant threat to public and individual health worldwide. Toxic elements such as heavy metals are associated with increased mortality and morbidity in both men and women and are a substantial contributor to neurological deficits and developmental delay in children. Analysis of skeletal material yields important information regarding exposure to toxic elements in a given population. This project has investigated toxic element exposure in 215 adults living in urban South Africa who died between 1960 and 1999. Exposure to toxic elements, particularly exposure to lead, has significant impacts on human health, even at very low levels. To date, little research has been conducted on human exposure to toxic elements in adult urban South Africans and a clear gap exists regarding toxic element exposure rates during the latter half of the 20th century. Among the primary aims of this research is to address this gap in knowledge and to quantify human exposure to these elements during the apartheid era. Bone element concentration was analysed by ICP-MS to determine the concentration of six elements that are toxic to humans: lead, cadmium, manganese, arsenic, antimony and vanadium. The results of this research demonstrate clear racial divisions in toxic element exposure in all but one element investigated. In the case of lead and cadmium, white males in the sampled population show significantly higher bone element concentrations than either black males or black females. It is surmised that apartheid-era separation of racial groups in regards to residence, occupation and movement within the urban landscape are partly, if not significantly, responsible for these differences in toxic element exposure. Lead exposure is strongly associated with exposure to traffic in urban Pretoria and Johannesburg, which is evident in both the limited environmental data available and the present study. Designated residential areas for white individuals were situated in and adjacent to the central business districts of both cities and are the areas associated with high traffic. Black residential areas were located on the urban periphery, often near industrial areas and mine dumps. The result is a lead exposure pattern by which white individuals in the sampled population yield double the bone lead concentration of black individuals. The wide divide in socioeconomic strata between the black and white population also factors significantly and is an additional result of apartheid policy. For arsenic and antimony, black individuals, particularly females, show significantly higher bone element concentration than white individuals. These elements are strongly associated with acid mine drainage, a form of pollution which results from mining activity. The close proximity of black residential areas to mining activities and the heavy reliance on ii contaminated surface water is likely responsible for higher exposure rates to these elements in the black population. This research has established that rates of exposure to toxic elements in urban Transvaal were moderate considering the level of industrial and mining activity in the region and the notably lax environmental regulations in place during the latter half of the 20th century. Despite this, bone element levels, particularly that of lead, cadmium and manganese are within ranges documented to cause negative impacts on human health. It is highly likely, given the bone element concentrations reported here, that these elements caused significant and negative health effects in the sampled population and were a clear threat to overall public health in urban South Africans.