A leprose dos citros, causada por Citrus leprosis vírus (CiLV), é uma doença endêmica nas regiões produtoras do estado de São Paulo que causa sérios prejuízos à produção. O vírus é transmitido exclusivamente pelo ácaro Brevipalpus phoenicis e medidas de controle da doença visam principalmente reduzir a população do vetor através de aplicações de acaricidas. Apesar do controle da doença ser determinado por meio de amostragens da população do ácaro, não há nenhum estudo de longo prazo que correlacione a população do vetor com a incidência da doença. O objetivo deste trabalho foi caracterizar os padrões de crescimento temporal e os padrões espaciais da distribuição do ácaro e da doença e estabelecer uma associação entre as duas populações. A população do ácaro e os sintomas de leprose foram monitorados durante três anos sob condições naturais de epidemia, em um talhão de laranja doce, variedade Valência enxertada sobre limão cravo, A população de B. phoenicis apresentou flutuações, mas com picos crescentes ao longo do período. As oscilações na sua densidade dependeram principalmente da fenologia das plantas e menos das condições climáticas. O crescimento da incidência da doença foi lento e seguiu comportamento logístico. Apesar de não ser sistêmica a leprose dos citros comporta-se como poliética com acúmulo de inoculo de ano para ano. A infecção em ramos é a principal causa do caráter cumulativo da doença. Através de regressão logística, a probabilidade de doença foi modelada em função de covariáveis construídas de forma a captar informação da vizinhança das plantas no tempo passado a respeito da incidência do ácaro e da própria doença. Não houve correlação entre o nível de doença numa avaliação e a quantidade de ácaros em avaliações anteriores, em períodos pregressos de até 70 dias. O padrão espacial de plantas com sintomas foi altamente agregado e não correspondeu à distribuição espacial de plantas infestadas pelo ácaro, que foi fracamente agregada. A probabilidade de infestação e de infecção nas plantas foi modelada como dependente do estado das plantas vizinhas, quanto às próprias características, através de modelos autologísticos. A probabilidade de plantas estarem infestadas pelo ácaro não depende de sua vizinhança estar ou não hospedando a praga, ao mesmo tempo. Já a probabilidade de doença em uma árvore depende de árvores vizinhas estarem doentes naquele momento. O padrão espacial da incidência de plantas com sintomas visíveis provavelmente reflete o padrão de infestação pelos ácaros contaminados por CiLV, pois existe uma subpopulação de vetores inserida na população de B. phoenicis. O controle da doença baseado na população de B. phoenicis, não é portanto, recomendável. A amostragem para a tomada de decisão de controle deve ser pautada pela presença de sintomas e ácaros e não somente pela presença de ácaros. / Citrus leprosis, caused by Citrus leprosis virus (CiLV), is a disease endemic to the producing regions of the state of São Paulo, where it has a heavy impact on production. The virus is transmitted exclusively by the Brevipalpus phoenicis mite and disease control measures aim to reduce the vector population mainly by acaricide applications. Although the disease control is based on mite population samplings, there are no long-term studies available that correlate the vector population with the disease incidence. The objective of this study was to characterize temporal and spatial growth patterns of the mite, disease distribution and the association between the two populations. The mite population and citrus leprosis symptoms were monitored for three years under natural epidemic conditions, in a sweet orange stand of the variety Valencia, grafted unto Citrus limonia Osbeck. The B. phoenicis population presented fluctuations, although with increasing peaks along the study period. The oscillation density depended more on the tree phenology than on climate conditions. The disease incidence increased slowly under a logistic model. Although citrus leprosis is not systemic, it has a polietic performance and builds up inoculum year after year. The infection of branches is the main cause of the cumulative nature of the disease. The disease probability was modeled by a covariable constructed to capture information of the trees surroundings in the past, concerning mite incidence and the disease itself. There was no correlation between the disease level in one evaluation and the mite quantity in earlier evaluations, in antecedent periods of up to 70 days. The spatial pattern of trees with symptoms was highly aggregated and did not correspond to the spatial distribution of miteinfested trees, which was weakly aggregated. The probability of tree infestation and infection was modeled as dependent of the state of the surrounding trees through autologistic models. The probability of mite infestation of a tree does not depend on whether the neighbor trees host the pest or not, but on whether the neighbor trees are diseased at that moment. A vector subpopulation is inserted in the B. phoenicis population. The spatial pattern of tree incidence with visible symptoms probably reflects the infestation pattern by CiLV-contaminated mites. Sampling for decision taking regarding disease control must be based on the presence of symptoms as well as of mites, rather than on the mite population only.

Many MBA buildings have been studied from different points of views and approaches / however, their spatial patterns were scarcely investigated. This thesis aims to assess the usefulness of Depth Analysis, a sub-method of Space Syntax, in the understanding of spatial patterns of MBA palaces. The palaces of key MBA sites, Acemh&ouml / y&uuml / k, Tell Mardikh and K&uuml / ltepe, were selected for the analysis. In the conclusion it is suggested that Depth Analysis can be a very useful method in understanding the configuration of spaces when there is sufficient available data. An equally successful application of this method lies in the comparative analysis of different buildings.

British Columbia’s current mountain pine beetle epidemic has led to salvage and mitigation harvesting strategies intended to slow the dispersal of beetles, and recover economic value from infested timber stands. These resulting harvesting strategies will alter the spatial pattern of forest landscapes in impacted regions, often resulting in forest fragmentation. As a result, wildlife habitat, hydrologic regimes, local carbon budgets, and soil dynamics, amoung other ecological properties, are expected to be negatively impacted. Monitoring of forest fragmentation in Canada is now required for the Montreal Process, an international forest monitoring policy. Effective methods that quantify changes in forest fragmentation, the breaking up of forest land cover into smaller, and more numerous parts, are required to meet forest monitoring objectives. This research provides two new methods that build upon existing approaches widely used for quantifying the spatial patterns of landscape features (i.e., landscape pattern indices). The first approach I demonstrate aids the quantification of forest pattern change over two time periods, by accounting for the impact of composition on spatial configuration. The value of this method is demonstrated using a case study that highlights the impacts of forest harvesting, associated with insect salvage and mitigation activities. This method allows landscapes that have changed primarily in composition to be distinguished from those that have experienced large configurational change. In the second approach I use multivariate cluster analysis for regionalization (the grouping of objects in space), and identify regions within a study area where increased fragmentation is observed. Regions delineated based on forest spatial pattern can be linked to underlying processes. Ancillary information (e.g., elevation) can be used to identify areas where observed forest pattern is due to underlying physiological features. Pattern indices (e.g., patch perimeter-area ratio) can be used to distinguish between patterns arising from forest disturbance that is likely natural (e.g., fire) or anthropogenic (e.g., harvest activity) in origin. The methods presented in this thesis may be most appropriate when observed changes in landscape pattern can be attributed to substantial changes in landscape composition.

Wildfire processes in Canada are expected to change as a result of climate change. Predictive modeling of wildfire occurrence and susceptibility requires knowledge of ignition expectations and landscape conditions leading to burn. This research examines and quantifies the spatial and temporal patterns of wildfire across Canada with focus on wildfire occurrence and national scale drivers of susceptibility. Baseline ignition expectations and trends are identified and used to create unique fire ignition regimes, assess anthropogenic influence on ignitions, and determine regions with anomalously high ignitions. The aspatial and spatial characteristics of land cover were characterized for pre- and post-fire landscapes. These included land cover composition, configuration, and abiotic covariates. Temporal trends in forest pattern following ignition are examined and national scale drivers of wildfire susceptibility determined. Fire ignition regimes and anomalous ignition regions provide spatially explicit outputs for exploring ignition expectation in Canada. Wildfire was identified to burn mainly in coniferous forests with little fragmentation. Fragmentation increased after wildfire and regeneration of pre-fire forest pattern took 20 years. Additionally, anthropogenic proximity positively influenced ignition expectation, ignition trend, and wildfire susceptibility. This research provides broad scale methods to assess wildfire occurrence and susceptibility across Canada and will facilitate understanding of changing wildfire processes in the future. Additionally, this research highlights the importance of anthropogenic activity on natural fire processes.

FundaÃÃo Cearense de Apoio ao Desenvolvimento Cientifico e TecnolÃgico / A estacionalidade e a estocasticidade pluviomÃtrica sÃo os principais reguladores da estrutura e da dinÃmica das populaÃÃes em vegetaÃÃes tropicais sazonais secas. Essa estocasticidade, no entanto, pode resultar em ambientes heterogÃneos tanto em relaÃÃo à disponibilidade hÃdrica como em relaÃÃo à disponibilidade vertical e horizontal de luz. Na vegetaÃÃo decÃdua espinhosa do semiÃrido brasileiro (caatinga), uma das espÃcies lenhosas mais abundantes à Cordia oncocalyx, Ãrvore endÃmica e explorada por extrativismo devido ao valor madeireiro, energÃtico, forrageiro, medicinal e paisagÃstico. Para entender como Ãrvores da caatinga respondem à sazonalidade e estocasticidade pluviomÃtrica foram analisados, ao longo de dois anos, a estrutura e a dinÃmica populacional de C. oncocalyx na RPPN Serra das Almas, CrateÃs-CearÃ. Foram determinados: i) estÃdios ontogenÃticos, modelo arquitetural e relaÃÃo altura-diÃmetro; ii) fenodinÃmica; iii) padrÃo espacial; iv) taxas de natalidade, mortalidade, recrutamento e crescimento biomÃtrico. ParÃmetros de estrutura e de dinÃmica foram correlacionados com precipitaÃÃo, umidade do solo, temperatura e disponibilidade vertical e horizontal de luz. A estrutura de C. oncocalyx se assemelha a das espÃcies tolerantes à sombra de florestas pluviais por concentrar a maior densidade nos estÃdios iniciais, mas difere porque forma banco de infantes ao invÃs de banco de plÃntulas. A disponibiliade vertical de luz, com apenas 4.7% atingindo o solo, e a arquitetura PrÃvost indicam a luz como fator limitante para os estÃdios iniciais. Contudo, o coeficiente alomÃtrico diferiu dos modelos de similaridade geomÃtrica, similaridade elÃstica e estresse constante. O crescimento contÃnuo em diÃmetro, mesmo apÃs os indivÃduos atingirem altura mÃxima, aponta a baixa densidade de Ãrvores e as restriÃÃes hÃdricas como as principais forÃas impulsionadoras da alometria, as quais podem limitar o crescimento em altura mas nÃo em diÃmetro. As fenofases vegetativas e reprodutivas respondem similarmente aos pulsos de chuva, com ajustes na Ãpoca, duraÃÃo e intensidade, correlacionadas com variaÃÃes na precipitaÃÃo e umidade do solo, excluÃdo o fotoperÃodo como gatilho. Menor sincronia, separaÃÃo temporal de fenofases e estoque de frutos no banco do solo, foram estratÃgias de distribuiÃÃo de risco apresentadas pela populaÃÃo. A modificaÃÃo da agregaÃÃo em semente e plÃntula para aleatoriedade em infante, e a dissociaÃÃo dos reprodutivos e sementes em relaÃÃo aos estÃdios juvenil e imaturo, indicam que a mortalidade dependente de densidade atua na estruturaÃÃo espacial da populaÃÃo. Entretanto, a retomada da agregaÃÃo nos estÃdios juvenil, virgem e reprodutivo indica que a estocasticidade pluviomÃtrica pode gerar a distribuiÃÃo agregada desses estÃdios. A natalidade e o recrutamento ocorrem na estaÃÃo chuvosa, mas a mortalidade registrada apenas em plÃntula e infante se dà ao longo do ano com pico nos meses secos. O crescimento em altura e diÃmetro mostrou decrÃscimo ou estagnaÃÃo na seca e acrÃscimo na chuva. Juvenil apresentou maior crescimento em altura em Ãreas com dossel mais aberto (rs=0.24). A taxa de incremento populacional (ʎ) foi de 1,0336 e a maior sensibilidade foi na transiÃÃo de infante para juvenil. DominÃncia local e tendÃncia de aumento populacional sÃo explicadas pela formaÃÃo de dois bancos, sementes e infantes, os quais garantem um estoque para recomposiÃÃo populacional mesmo em anos de seca. / The rainfall seasonality and stochasticity are the main regulators of structure and population dynamics in seasonally dry tropical vegetations. This stochasticity, however, can result in heterogeneous environments both in relation to water availability as in the availability of vertical and horizontal light. In deciduous thorny woodland vegetation of semiarid northeastern Brazil (caatinga), one of the most abundant woody species is Cordia oncocalyx, endemic tree and explored through extractivism due to its timber, energetic, foraging, medicinal, and scenic value. To understand how the caatinga trees respond to rainfall seasonality and stochasticity were analyzed, over two years, the structure and population dynamics of C. oncocalyx in Serra das Almas RPPN, CrateÃs-CearÃ. Were determined: i) ontogenetic stages, architectural model and height-diameter relationship, ii) phenodynamics, iii) spatial pattern, iv) birth, mortality, recruitment and growth biometric rates. Parameters of structure and dynamics were correlated with rainfall, soil humidity, temperature and vertical and horizontal light availability. The structure of C. oncocalyx resembles that of shade-tolerant species of tropical rainforest to concentrate the highest density in the early stages, but differs because it forms the bank of infant rather than a seedling bank. The availability of vertical light, with only 4.7% reaching the ground, and PrÃvost architecture indicate light as a limiting factor in the early stages. Nevertheless, the allometric coefficient differ from geometric similarity, elastic similarity and constant stress models. The uninterrupted growth in diameter, even after individuals reach the maximum height, indicates the low density of trees and water restrictions as the main driving strenght of allometry, which may limit the growth in height, but not in diameter. Both the vegetative and reproductive phenophases respond similarly to variations in rain pulses, with adjustments in time, duration, and intensity, which were correlated with variations in rainfall and soil humidity, excluding photoperiod as a trigger. Lower synchrony, temporal separation of phenophases, and storage of fruits on the ground were risk-spreading strategies used by the population in the dry year. The change of aggregation in seeds and seedlings for randomness in the infant, and the decoupling of reproductive and seeds in relation to juvenile and immature stages, may indicate that the density-dependent mortality operates in the spatial structure of the population. However, the resumption of aggregation in juvenile, virginile and reproductive stages indicates that rainfall stochasticity creates the aggregate distribution of these stages. The dynamics of birth and recruitment occur in the rainy season, but recorded mortality only in seedling and infant occurs throughout the year with a peak in the dry months. The height and diameter growth showed decline or stagnation in the drought and increase in the rainfall. Juveniles showed greater growth in height in areas with more open canopy (rs = 0.24). The rate of population increase (ʎ) was 1.0336 and the highest sensitivity was in the transition from infant to juvenile. Local dominance and population growth tendency are explained by the formation of two banks, seeds and infants, which ensure a stock for recomposition even in drought years.

Conselho Nacional de Desenvolvimento CientÃfico e TecnolÃgico / Atualmente, a doenÃa conhecida como resinose, causada pelo fungo Lasiodiplodia theobromae, vem merecendo destaque no semi-Ãrido brasileiro devido ao ataque em galhos e ramos lenhosos da planta, provocando significativas perdas na produÃÃo de castanha. O desenvolvimento de estudos epidemiolÃgicos à de fundamental importÃncia no entendimento dos aspectos ecolÃgicos envolvidos na ocorrÃncia e progresso da doenÃa. Este trabalho objetivou realizar estudos sobre a dinÃmica espaÃo-temporal do patÃgeno em trÃs diferentes clones de cajueiro. Foram estimadas a dispersÃo desta doenÃa atravÃs de seu padrÃo espacial e temporal na Ãrea em estudo, estudando-se os componentes das anÃlises como os modelos de curva de progresso da doenÃa, intensidade mÃxima da doenÃa, inÃcio de aparecimento dos sintomas e Ãrea abaixo da curva de progresso da doenÃa, de forma a comparar epidemiologicamente o desenvolvimento desta doenÃa em clones com diferentes reaÃÃes. As Ãreas de estudo estÃo localizadas em uma propriedade situada no municÃpio de Pio IX, PI. Os dados de incidÃncia e severidade foram coletados entre os anos de 2002 e 2007, e utilizados nas anÃlises espaciais e temporais para os clones BRS 226, Embrapa 51 e Faga 11. Mesmo apresentando diferenÃas quanto à incidÃncia e severidade, foi observado que a doenÃa assume um modelo aleatÃrio de dispersÃo no inÃcio da epidemia, e que o aumento da incidÃncia se dà pelo surgimento de novos focos unitÃrios e crescimento dos focos antes existentes. Posteriormente, registrou-se uma agregaÃÃo de plantas doentes. Os clones apresentaram ajustes a diferentes modelos epidemiolÃgicos de acordo com o grau de susceptibilidade. As comparaÃÃes entre as epidemias baseadas nos seus componentes apresentaram diferenÃas significativas entre os clones de cajueiro, sugerindo o uso potencial da resistÃncia genÃtica do hospedeiro como mÃtodo de controle da resinose do cajueiro. / Presently, a cashew disease known as gummosis, caused by the fungus Lasiodiplodia theobromae, deserves special attention due to its infection to branches and trunks of woody plants imposing significantly reduction in nut yield. As the importance of disease increases, it becomes of fundamental importance the development of epidemiological studies to understand the ecological aspects involved on disease establishment and progress. This work aimed to study gummosis dynamics in time and space in three different clones of cashew. Patterns of dispersion in space and disease progress on time, as well as the components involved on these models such as disease progress curve, maximum disease intensity, area under the disease progress curve and the point of initial disease were studied analyzed in order to compare epidemics in the three clones with different disease reactions. The study was conducted in a commercial cashew farm in Pio IX district in Piaui state, Brazil. Gummosis incidence and severity data of collected from 2002 to 2007 were used for both time and space analyses for BRS 226, Embrapa 51 and Faga 11 clones. Disregarding differences in disease occurrence and severity among clones, a random pattern of dispersion was observed at the beginning of the epidemic followed by the development of new foci and expansion of original foci. Later, a clustered pattern was observed. Clones showed different fitness to epidemic models accordingly with their degree of susceptibility. Comparison of epidemics based on their components have shown significant differences among clones, providing evidence for the potential use of host resistance as a means of gummosis control.

A leprose dos citros, causada por Citrus leprosis vírus (CiLV), é uma doença endêmica nas regiões produtoras do estado de São Paulo que causa sérios prejuízos à produção. O vírus é transmitido exclusivamente pelo ácaro Brevipalpus phoenicis e medidas de controle da doença visam principalmente reduzir a população do vetor através de aplicações de acaricidas. Apesar do controle da doença ser determinado por meio de amostragens da população do ácaro, não há nenhum estudo de longo prazo que correlacione a população do vetor com a incidência da doença. O objetivo deste trabalho foi caracterizar os padrões de crescimento temporal e os padrões espaciais da distribuição do ácaro e da doença e estabelecer uma associação entre as duas populações. A população do ácaro e os sintomas de leprose foram monitorados durante três anos sob condições naturais de epidemia, em um talhão de laranja doce, variedade Valência enxertada sobre limão cravo, A população de B. phoenicis apresentou flutuações, mas com picos crescentes ao longo do período. As oscilações na sua densidade dependeram principalmente da fenologia das plantas e menos das condições climáticas. O crescimento da incidência da doença foi lento e seguiu comportamento logístico. Apesar de não ser sistêmica a leprose dos citros comporta-se como poliética com acúmulo de inoculo de ano para ano. A infecção em ramos é a principal causa do caráter cumulativo da doença. Através de regressão logística, a probabilidade de doença foi modelada em função de covariáveis construídas de forma a captar informação da vizinhança das plantas no tempo passado a respeito da incidência do ácaro e da própria doença. Não houve correlação entre o nível de doença numa avaliação e a quantidade de ácaros em avaliações anteriores, em períodos pregressos de até 70 dias. O padrão espacial de plantas com sintomas foi altamente agregado e não correspondeu à distribuição espacial de plantas infestadas pelo ácaro, que foi fracamente agregada. A probabilidade de infestação e de infecção nas plantas foi modelada como dependente do estado das plantas vizinhas, quanto às próprias características, através de modelos autologísticos. A probabilidade de plantas estarem infestadas pelo ácaro não depende de sua vizinhança estar ou não hospedando a praga, ao mesmo tempo. Já a probabilidade de doença em uma árvore depende de árvores vizinhas estarem doentes naquele momento. O padrão espacial da incidência de plantas com sintomas visíveis provavelmente reflete o padrão de infestação pelos ácaros contaminados por CiLV, pois existe uma subpopulação de vetores inserida na população de B. phoenicis. O controle da doença baseado na população de B. phoenicis, não é portanto, recomendável. A amostragem para a tomada de decisão de controle deve ser pautada pela presença de sintomas e ácaros e não somente pela presença de ácaros. / Citrus leprosis, caused by Citrus leprosis virus (CiLV), is a disease endemic to the producing regions of the state of São Paulo, where it has a heavy impact on production. The virus is transmitted exclusively by the Brevipalpus phoenicis mite and disease control measures aim to reduce the vector population mainly by acaricide applications. Although the disease control is based on mite population samplings, there are no long-term studies available that correlate the vector population with the disease incidence. The objective of this study was to characterize temporal and spatial growth patterns of the mite, disease distribution and the association between the two populations. The mite population and citrus leprosis symptoms were monitored for three years under natural epidemic conditions, in a sweet orange stand of the variety Valencia, grafted unto Citrus limonia Osbeck. The B. phoenicis population presented fluctuations, although with increasing peaks along the study period. The oscillation density depended more on the tree phenology than on climate conditions. The disease incidence increased slowly under a logistic model. Although citrus leprosis is not systemic, it has a polietic performance and builds up inoculum year after year. The infection of branches is the main cause of the cumulative nature of the disease. The disease probability was modeled by a covariable constructed to capture information of the trees surroundings in the past, concerning mite incidence and the disease itself. There was no correlation between the disease level in one evaluation and the mite quantity in earlier evaluations, in antecedent periods of up to 70 days. The spatial pattern of trees with symptoms was highly aggregated and did not correspond to the spatial distribution of miteinfested trees, which was weakly aggregated. The probability of tree infestation and infection was modeled as dependent of the state of the surrounding trees through autologistic models. The probability of mite infestation of a tree does not depend on whether the neighbor trees host the pest or not, but on whether the neighbor trees are diseased at that moment. A vector subpopulation is inserted in the B. phoenicis population. The spatial pattern of tree incidence with visible symptoms probably reflects the infestation pattern by CiLV-contaminated mites. Sampling for decision taking regarding disease control must be based on the presence of symptoms as well as of mites, rather than on the mite population only.

The emerging interests in spatial pattern mining leads to the demand for a flexible spatial pattern mining language, on which easy to use and understand visual pattern language could be built. It is worthwhile to define a pattern mining language called LCSPM to allow users to specify complex spatial patterns. I describe a proposed pattern mining language in this paper. A visual interface which allows users to specify the patterns visually is developed. Visual pattern queries are translated into the LCSPM language by a parser and data mining process can be triggered afterwards. The visual language is based on and goes beyond the visual language proposed in literature. I implemented a prototype system based on the open source JUMP framework.

Unplanned hospital readmission after a recent hospitalization is an indication of poor healthcare quality and a waste of healthcare resources. The Centers for Medicare and Medicaid Services (CMS) initiated the Hospital Readmission Reduction Program (HRRP) to improve healthcare quality and reduce costs; however, studies found the risk adjustment method used in calculating the standardized readmission rate was less accurate without hospital region or community factors. Accordingly, this cross-sectional quantitative study was designed to examine spatial patterns in hospital readmission rates following Andersen's behavioral model of health service utilization. This study was the first geospatial analysis on risk standardized hospital readmissions (RSRR) based on hospital geographic locations. Secondary data from the CMS was used in assessing the global and local geospatial cluster patterns using Global Moran's Index, Anselin local Moran's Index, and graphical analysis tool to identify cluster groups. The study found hospital-wide RSRR was significantly clustered across the country or at the local level. A total of 15 optimal cluster groups were identified with wide variability in cluster size. The hospital-wide and other seven CMS published RSRRs were significantly different among all clusters. The geographically bounded hospital RSRRs provided evidence in support of adding community or regional layer to risk adjustment of RSRR. The specific cluster groups with extremely high or low readmission rates can assist national and local policymakers and hospital administrators to identify specific targets to take actions. This research has social change implications for reducing hospital readmission rates and saving healthcare costs.

Whitebark pine is an important keystone and foundation species in western North American mountain ranges, and facilitates tree island development in Rocky Mountain treelines. The manifestation of white pine blister rust in the cold and dry treelines of the Rockies, and the subsequent infection and mortality of whitebark pines raises questions as to how these extreme environments harbor the invasive disease, and what the consequences may be for treeline dynamics. This dissertation research comprises three studies that investigate abiotic factors influential for blister rust infection in treeline whitebark pines, how disease coupled with changing climate may affect whitebark pine treeline dynamics, and the connection between treeline spatial patterns and disease. The first study examined the spatial variation of blister rust infection in two whitebark pine treeline communities, and potential topographic correlates. Using geospatial and field approaches to generate high resolution terrain models of treeline landscapes, microtopography associated with solar radiation and moisture were found most influential to blister rust infection in treeline whitebark pines. Using field-based observations from sampled treeline communities, the second study developed an agent-based model to examine the effects of disease and climate on treeline pattern and process. Treeline dynamics were simulated under five hypothetical scenarios to assess changes in tree spatial patterns and populations. Blister rust-induced loss of whitebark pines resulted in a decline of facilitative processes, and an overall negative treeline response to disease—despite the beneficial effects of climate amelioration. The objective of the third study was to explore whether spatial patterns of tree proximity, size, and growth infer disease patterns. Comparing spatial patterns of tree characteristics between diseased and undiseased treeline communities, I found that trees growing near trees with larger stem diameters, and larger tree islands, tended to have more blister rust cankers, and displayed clustered spatial patterns. Undiseased treeline patterns revealed near neighbors smaller in stem diameter and tree island size, and were randomly dispersed. Blister rust diseased whitebark pines reveal spatial autocorrelation, despite the complex blister rust disease life cycle. Overall, findings from this dissertation reveal the implications of invasive disease on sensitive treeline ecotones dependent on a keystone species. / Ph. D.