A dissertation submitted to the Faculty of Science, University of the Witwatersrand, in fulfilment of the requirements for the degree of Master of Science. Johannesburg, 2017. / With an increase in opportunistically-collected data, statistical methods that can accommodate unstructured designs are increasingly useful. Spatial capturerecapture (SCR) has such potential, but its applicability for species that are strongly gregarious is uncertain. It assumes that average animal locations are spatially random and independent, which is violated for gregarious species. I used a data set for African elephants (Loxodonta africana) and data simulation to assess bias and precision of SCR population density estimates given violations in location independence. I found that estimates were negatively biased and likely too precise if non-independence was ignored. Encounter heterogeneity models produced more realistic precision but density estimates were positively biased. Lowest bias was achieved by estimating density of groups, group size, and then multiplying to estimate overall population density. Such findings have important implications for the reliability of population density estimates where data are collected by unstructured means. / LG2017

As populações de veado-campeiro vêm sofrendo constantes ameaças, principalmente pela caça excessiva e destruição de seu habitat. A escassez de conhecimento sobre a biologia da espécie dificulta a elaboração de diferentes formas de manejo. Assim, há necessidade de estudos básicos para a instituição de programas para a conservação do veado-campeiro, principalmente no Brasil. O presente estudo teve como objetivos estimar o tamanho da área de vida de fêmeas de veado-campeiro, além de quantificar e comparar o uso de habitat pelas mesmas nos períodos reprodutivos e não-reprodutivos no Pantanal Sul-Matogrossense. Foram utilizadas quatro fêmeas de veado-campeiro em vida livre marcadas com colares de GPS modelo ATS® G2110, as quais tiveram suas localizações registradas a cada 13 horas por um período de aproximadamente 12 meses. Os cálculos para área de vida foram feitos através do programa ArcView® GIS versão 3 Extensão Animal Movement, usando os métodos do Mínimo Polígono Convexo (MPC) com 100% das localizações, e Kernel fixo com 95%, 70% e 50% de probabilidade de uso, sendo os dois últimos para estimativas dos centros de atividade. As fêmeas tiveram uma área de vida anual média de 549,88ha ± 124,76 quando estimado pelo MPC e 258,69ha ± 34,54 quando estimado por Kernel com 95% de probabilidade de uso. Quanto às áreas de uso mensais por status reprodutivo, não houve diferença significativa entre os meses (p > 0,05), evidenciando apenas uma pequena diminuição no tamanho das áreas de vida durante o período de aleitamento e um aumento nas épocas de acasalamento. Os habitats foram estabelecidos pelo tipo de vegetação predominante na área e foram caracterizados por imagem de satélite Landsat-5 e verificação in loco da região de estudo. Para verificar se houve uso desproporcional entre as categorias de habitats em relação à disponibilidade em suas respectivas áreas de vida foi usado o teste qui-quadrado de qualidade de ajustamento (chi-square goodness-of-fit test) e posterior estimativa do intervalo simultâneo de confiança de Bonferroni. Na análise anual, os habitats campo e baía foram selecionados, apesar deste último ser o menos abundante, enquanto floresta e caronal foram evitados pelas fêmeas de veado-campeiro. No entanto, ao analisarmos a seleção de habitat mensalmente para comparar o uso entre os diferentes status reprodutivos, o habitat campo tendeu a ser preferido, mas seu uso foi igual ao disponível em todos os períodos. Sendo assim, este estudo elucidou algumas informações sobre a ecologia do veado-campeiro, a fim de facilitar o desenvolvimento de programas de conservação para a espécie. / The pampas deer populations have been under constant threats, mainly by excessive hunting and habitat destruction. The lack of knowledge about the biology of the species hampers the development of different forms of management. Thus, there is a need for basic studies of establishment of conservation programs, mainly in Brazil. This study aimed to estimate the size of home-range, to quantify and compare the habitat use by females in reproductive and non-reproductive periods rights in Pantanal Sul-Matogrossense. Four females of pampas deer were marked with GPS collars, model ATS® G2110, which had recorded their locations every 13 hours for approximately 12 months. The calculations for home-range were made using the Arcview® GIS version 3 Extension Animal Movement, using the methods of the minimum convex polygon (MPC) with 100% of the locations, and fixed Kernel 95%, 70% and 50% of probability, being the last two for estimating the activity centers. The females had an annual home-range average of 549.88 ± 124.76 ha when estimated by MPC and 258.69 ± 34.54 ha when estimated by Kernel with 95% of probability. For the monthly home-ranges for reproductive status, no significant difference among months (p > 0.05), showing only a slight decrease in the size of home-range during the suckling period, increasing in the mating seasons. The habitats were established by the type of dominant vegetation and characterized by landsat-5 imaging satellite and in loco verification of the study area. To determine if there was disproportionate use of habitats between the categories regarding availability in their respective home-range, was used the chi-square goodness-of-fit test and subsequent estimation of Bonferroni simultaneous confidence intervals. In annual analysis, field and bay habitats were selected, despite the last being the less abundant, while forest and caronal were avoided by females of pampas deer. However, when analyzing the monthly habitat selection to compare the use between the different reproductive status, the habitat field tended to be preferred, but its use was equal to that available in all periods. Thus, this study clarified some informations about the ecology of pampas-deer, facilitating conservation programs development for the species.

Pixel based and object based vegetation community classification methods were performed using 30 meter spatial resolution Landsat satellite imagery of the Arthur R. Marshall Loxahatchee National Wildlife Refuge (Refuge), a remnant of the northern Everglades. Supervised classification procedures using maximum likelihood and parallelepiped algorithms were used to produce thematic maps with the following vegetation communities : wet prairie, sawgrass, cattail, tree island, brush, aquatic/open water. Spectral data, as well as NDVI, texture and principal component data were used to produce vegetation community classification maps. The accuracy levels of the thematic maps produced were calculated and compared to one another. The pixel based approach using the parallelepiped classification algorithm on the spectral and NDVI dataset had the highest accuracy level. A generalized form of this classification using only three vegetation communities (all wet prairie, tree island/brush and aquatic/open water) was compared to a previously published classification which used 1987 SPOT imagery in order to extract information on possible vegetation community transitions that are occurring within the Refuge. Results of the study indicate that 30 meter spatial resolution may be useful for understanding broad vegetation community trends but not species level trends. Pixel based procedures provide a more accurate classification than object based procedures for this landscape when using 30 meter imagery. Lastly, since 1987 there may be a trend of tree island/brush communities replacing wet prairie communities in the northern part of the Refuge and a transition to wet prairie communities in place of tree island/brush communities in the southern portion of the Refuge. / by Dorianne M. Barone. / Thesis (M.S.)--Florida Atlantic University, 2008. / Includes bibliography. / Electronic reproduction. Boca Raton, FL : 2008 Mode of access: World Wide Web.

Les oiseaux marins sont des espèces emblématiques. Passeuses de frontières, de la mer à la terre, de l’air à l’eau, des pays du nord au pays du sud, elles créent un lien entre les différents sociaux- écosystèmes marins de notre planète. Du fait de leur ubiquité, elles sont exposées à de nombreuses menaces autour du monde. Parmi elles, les interactions avec les pêcheries représentent la part de risque la plus importante pour ces espèces lorsqu’elles sont en mer. Leur statut de conservation en est affecté, et des actions prioritaires visant à réduire ces impacts doivent être mis en place. Ces espèces bénéficient de l’intérêt, voir de la sympathie des populations et le grand public est sensible au sort de leurs populations. Mettre en place des stratégies et des outils permettant la conservation des populations d’oiseaux marins répond donc à une demande sociétale urgente.Le fou de Bassan (Morus bassanus) est une espèce emblématique de la conservation des oiseaux marins en France métropolitaine. Au sein de la Réserve Naturelle Nationale de l’archipel des Sept-Îles, la seule colonie de reproduction Française de cette espèce bénéficie d’un statut de protection fort. Malgré cela, au cours de la dernière décennie, notre étude a mis en évidence une inversion de la dynamique de la population et une baisse du succès reproducteur. La mise en place d’un suivi bio-télémétrique nous a alors permis de chercher à comprendre et expliquer ces changements. Nous avons notamment mis en évidence, que durant la saison de reproduction, les fous des Sept-Îles souffraient de la diminution de leurs proies naturelles et se rabattaient alors sur des rejets de pêche. La consommation de ces subsides anthropiques affecte les efforts de recherche alimentaire, la condition des individus et finalement leur reproduction. De plus, nous avons montré que durant la période internuptiale, ils étaient exposés à de forts risques de captures accidentelles et à une diminution globale de leurs proies préférentielles, affectant les taux de retours à la colonie et expliquant potentiellement la baisse observée de la taille de la population.Ces travaux nous amènent à conclure que la bonne conservation des fous des Sept-Îles, comme celle de toute la mégafaune marine, ne pourra se faire qu’en adoptant une approche écosystémique des pêches. Particulièrement, le partage de certaines ressources entre prédateurs supérieurs et pêcheries devra être pris en compte dans la gestion des stocks, la diminution des rejets de pêche devra être favorisée et des aires marines protégées pélagiques excluant les activités de pêche, dessinées à partir des zones d’intérêt pour les oiseaux marins, devront être mise en place. / Seabirds are flagship species, boundary objects linking air and water, oceans and continents, Northern and Southern countries, binding a great variety of socio-ecosystems across the planet. Due to their ubiquity, they are exposed to numerous global threats. Among them, interactions with fisheries might be the main risk for seabirds at sea. The conservation status of seabirds is thereby affected, and priority actions due to reduce these impacts have to be established. Indeed, seabirds catch the attention of all stakeholders and of the general public, who are sensitive to the fate of their populations. Implementing tools and strategies allowing seabird conservation is therefore an urgent societal request. The Northern gannet (Morus bassanus) is emblematic of seabird conservation in metropolitan France, with a single breeding colony under strict protection within the Réserve Naturelle Nationale de l’archipel des Sept-Îles. Despite all conservation efforts, colony size and breeding success have been declining in recent years. A decadal biotelemetry study allowed us to test hypotheses linked to this decline. Notably, we showed that, during the breeding season, gannets shifted from feeding on natural prey, to taking fisheries waste. The consumption of these anthropogenic subsides affects foraging effort, adult body condition and reproductive output. Further, we showed that, during the inter-breeding period, gannets were exposed to enhanced bycatch risk and competition with fisheries for small pelagic fish. This had a strong impact on adult inter-annual return rates to the colony, potentially explaining the recent decline of the Sept-Îles gannetry. Overall, we conclude that an integrated conservation plan for Northern gannets, as well as for the marine megafauna in general, is only possible through ecosystem-based fisheries management. Specifically, the joint use of fish stocks by marine predators and fisheries should be taken into account by management schemes, at-sea dumping of fishery wastes should be reduced, and marine protected areas including true no-take zones should be designed, also by taking into account the spatial ecology of the marine megafauna such as seabirds.

Wong, Lik Chi. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2011. / Includes bibliographical references (leaves 96-108). / Abstracts in English and Chinese. / Abstract (in English) --- p.i / Abstract (in Chinese) --- p.iii / Acknowledgements --- p.V / Table of Contents --- p.vi / List of Tables --- p.viii / List of Figures --- p.ix / Chapter Chapter 1. --- General introduction --- p.1 / Chapter Chapter 2. --- "Seasonal and spatial variations of phytoplankton size composition in Tolo Harbour and Tolo Channel, Hong Kong" --- p.3 / Chapter 2.1 --- Literature Review --- p.3 / Chapter 2.2 --- Introduction --- p.6 / Chapter 2.3 --- Materials and methods --- p.8 / Chapter 2.4 --- Results --- p.11 / Chapter 2.5 --- Discussion --- p.32 / Chapter 2.6 --- Conclusion --- p.37 / Chapter Chapter 3. --- "Seasonal and spatial variations of primary production and copepod production in Tolo Harbour and Tolo Channel, Hong Kong" --- p.38 / Chapter 3.1 --- Literature review --- p.38 / Chapter 3.1.1 --- Factors affecting primary production --- p.38 / Chapter 3.1.2 --- Factors affecting copepod production --- p.40 / Chapter 3.1.3 --- Marine food chain --- p.42 / Chapter 3.2 --- Introduction --- p.45 / Chapter 3.3 --- Materials and methods --- p.48 / Chapter 3.3.1 --- Primary production --- p.48 / Chapter 3.3.2 --- Copepod production --- p.50 / Chapter 3.3.3 --- Density of Noctiluca scintillans --- p.53 / Chapter 3.4 --- Results --- p.54 / Chapter 3.5 --- Discussion --- p.78 / Chapter 3.5.1 --- Primary production in Tolo Harbour --- p.78 / Chapter 3.5.2 --- Copepod production in Tolo Harbour --- p.82 / Chapter 3.5.3 --- Transfer efficiency in Tolo Harbour --- p.89 / Chapter 3.6 --- Conclusion --- p.92 / Chapter Chapter 4. --- General conclusion --- p.93 / References --- p.96

Habitat selection has long been viewed as a multi-scale process. Observed species responses to resource gradients are influenced by variation at the scale of the individual, population, metapopulation, and geographic range. Understanding how species interact with habitat at multiple levels presents a complete picture of an organism and is necessary for conservation of endangered species. The main goal of this dissertation is to evaluate distribution, relative abundance, and habitat selection of a rare species, the pygmy rabbit Brachylagus idahoensis, at multiple scales in order to improve management and conservation for this species. At the broadest scale, pygmy rabbit occurrence and relative abundance were modeled in the Duck Creek allotment of northern Utah using a hierarchical spatial model. Pygmy rabbits are not easily observable, and the model used two levels of indirect detection to make statistically rigorous spatial predictions. We found that the model predicted the general pattern of rabbit occurrence and abundance within the study area, and that there was spatial heterogeneity in the probability of pygmy rabbit occurrence within a study domain that was known to be occupied. The resulting model framework could be used to develop a long-term monitoring program for pygmy rabbits and other species for which hierarchically nested levels of indirect observation are collected. The mid-scale analysis evaluated pygmy rabbit home range placement and movement with respect to sagebrush removal treatments using null models based on an optimal central place foraging behavior. While placement of home-range centers did not appear to be affected by the treatments, within-home range movements were farther from treatments than expected by the null models for two rabbits (of eight), and rabbits that approached treatment edges were less likely to enter treatments than expected by chance. Rabbits are not extirpated from sites that have been treated, but the observed reluctance to enter treated patches calls for caution when conducting sagebrush removal treatments near occupied pygmy rabbit burrows. At the finest level of resolution, the spatial ecology of pygmy rabbit use of burrows was evaluated. Both the placement of burrows in general and pygmy rabbit use of burrows were clustered. While the habitat gradients experienced by each of the rabbits evaluated affected the modeled habitat selection responses, some generalities were observed. Selection of high cover suggests that pygmy rabbit use of burrows may be linked to predator avoidance behavior. Additionally, pygmy rabbit use of clustered burrows affects management actions including: habitat modeling, monitoring, and species introduction. Explicit attention to resource distribution will improve efforts to predict species responses to management actions.

Bibliography: leaves 132-153. / viii, 153 leaves : ill. ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Spatially explicit models and computer intensive analysis were employed to explore how processes acting at the individual level scale up to population dynamics when processes are variable in space as well as the consequences of sampling spatially complex variability for drawing conclusions from limited ecological data. Dispersal and variation in marsupial mortality and development in relation to habitat selection and quality were studied, while evaluating spatially explicit models. The study of dynamics models of tick populations on sleepy lizards considered the effect of spatial and temporal variability, and demonstrated that counting ticks is a poor indicator of tick abundance. The "virtual ecologist" model is a useful method for linking the output of spatially explicit population models to reality, and will be a valuable approach for improving the design of ecological field research on spatially complex landscapes. / Thesis (Ph.D.)--University of Adelaide, Dept. of Environmental Science and Management, 1999?

Bibliography: leaves 132-153. Spatially explicit models and computer intensive analysis were employed to explore how processes acting at the individual level scale up to population dynamics when processes are variable in space as well as the consequences of sampling spatially complex variability for drawing conclusions from limited ecological data. Dispersal and variation in marsupial mortality and development in relation to habitat selection and quality were studied, while evaluating spatially explicit models. The study of dynamics models of tick populations on sleepy lizards considered the effect of spatial and temporal variability, and demonstrated that counting ticks is a poor indicator of tick abundance. The "virtual ecologist" model is a useful method for linking the output of spatially explicit population models to reality, and will be a valuable approach for improving the design of ecological field research on spatially complex landscapes.

À l’échelle macroscopique, à savoir celle du paysage, la couverture végétale des milieux arides apparaît comme fortement hétérogène. Cette organisation spatiale des peuplements est habituellement attribuée aux processus de facilitation et de compétition qui opèrent à l’échelle microscopique des individus végétaux. Le qualificatif d’auto-organisées fut donc attribué à ces végétations qui se structurent en l’absence d’hétérogénéité préexistante du milieu physique. L’auto-organisation de la végétation fut particulièrement bien étudiée dans le cas des structures périodiques connues dès les années '50 sous le nom de brousses tigrées. Depuis les années '90, un pas en avant dans la compréhension de ce phénomène fut accompli grâce au développement de modèles mécanistes de la dynamique de la phytomasse et des ressources, émanant du cadre théorique de l'auto-organisation des structures dissipatives. Ces modèles se rejoignent sur un ensemble de prédictions robustes et vérifiables concernant la formation, le maintien et la modulation par l'environnement des structures macroscopiques. Durant le même laps de temps, notre niveau d’analyse a connu une expansion sans précédent, à la fois dans le temps et dans l’espace, grâce au développement de l’imagerie satellitaire et des outils d’analyse spatiale. Nous nous trouvons dès lors à un moment charnière pour la validation macroscopique des théories d’auto-organisation des végétations en milieu aride. Le présent travail s'articule en quatre études, chacune traitant d'une prédiction différente. Nous avons mis en évidence les principales variables responsables de la formation des structures et de leur modulation en termes d’échelle et de géométrie. Enfin avons démontré la mobilité des structures sous l’effet d’une pente de terrain.

Bibliography: leaves 132-153. / viii, 153 leaves : ill. ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Spatially explicit models and computer intensive analysis were employed to explore how processes acting at the individual level scale up to population dynamics when processes are variable in space as well as the consequences of sampling spatially complex variability for drawing conclusions from limited ecological data. Dispersal and variation in marsupial mortality and development in relation to habitat selection and quality were studied, while evaluating spatially explicit models. The study of dynamics models of tick populations on sleepy lizards considered the effect of spatial and temporal variability, and demonstrated that counting ticks is a poor indicator of tick abundance. The "virtual ecologist" model is a useful method for linking the output of spatially explicit population models to reality, and will be a valuable approach for improving the design of ecological field research on spatially complex landscapes. / Thesis (Ph.D.)--University of Adelaide, Dept. of Environmental Science and Management, 1999?