Os efeitos de borda e a alteração da estrutura das paisagens constituem consequências negativas da fragmentação florestal responsáveis por transformações nos processos ecológicos. Decorrentes da expansão desordenada de atividades antrópicas, estas alterações podem comprometer o futuro dos remanescentes florestais e a manutenção dos recursos naturais na superfície terrestre. Nesta dissertação a dinâmica da vegetação arbórea foi analisada em oito transectos perpendiculares às bordas de seis remanescentes florestais entre 1996 e 2012. As paisagens do entorno destes transectos foram caracterizadas a partir de imagens orbitais de 1995 e 2011 para a verificação das mudanças ocorridas no uso do solo e para a investigação da influência de seus parâmetros físicos e estruturais sobre as taxas de mortalidade e recrutamento de espécies. Os resultados indicaram que, ao longo do tempo, a influência das bordas se pronunciou em direção ao interior dos remanescentes florestais, enquanto que o contraste entre a borda e o interior se atenuou. A distância média da borda das espécies: pioneiras/iniciais, anemocóricas e de dossel foi significativamente maior em 2012 do que em 1996. A comunidade arbórea apresentou menor similaridade em sua composição ao longo do tempo a menores distâncias da borda. Apesar da dinâmica verificada no uso do solo, a proporcionalidade dos parâmetros físicos e estruturais das paisagens se manteve entre 1995 e 2011. De maneira geral, estes parâmetros apresentaram pouca influência sobre a dinâmica da comunidade arbórea. Apenas as taxas de mortalidade das espécies exóticas e as taxas de recrutamento das espécies pioneiras/inicias apresentam forte relação com o tamanho e o número dos fragmentos florestais nas paisagens. Estes resultados indicam que os efeitos de borda precisam ser atenuados e que o contexto das paisagens deve ser incorporado às estratégias conservacionistas para que estas sejam efetivas e o futuro dos remanescentes florestais não seja comprometido. / Edge effects and landscape structure alterations are among the negative consequences of forest fragmentation responsible for ecological process alterations on the earths surface. Originated from the disordered expansion of anthropogenic activities these alterations may endanger the remaining forest patches future and the maintenance of natural resources. This dissertation was pledged to analyze the vegetation dynamics at forest edges and its relations with landscape features. The vegetation dynamics was examined through eight perpendicular-to-edge transects within six forest patches and the alterations on the arboreal community distribution and composition were assessed between 1996 and 2012. The surrounding landscapes of the analyzed transects were characterized from 1995 and 2011 orbital images and its land use changes were evaluated. Landscape structure and physical parameters influence were analyzed over species recruitment and mortality. The results indicated that the distance of edge influence increased over time while its magnitude was attenuated. The average distance from the edge of pioneer/earlysuccessional species, wind-dispersed and canopy species in 2012 became significantly larger than in 1996. Over time lower similarities in species composition were found to be closer to the edges. Although the observed land use changes in the surrounding landscapes of the edge transects landscape structure and physical parameters proportionality was maintained between 1995 and 2011. Overall the arboreal community dynamics were poorly associated with landscape features. A strong relation of the variables was only found between the exotic and pioneer/early-successional species mortality and recruitment and the size and the amount of forest patches within the landscapes. These results indicate that to be effective conservation planning must tackled edge effects and incorporate the landscape context otherwise they will fail for the maintenance of the future of forest patches.

O cenário atual do estado de São Paulo reflete o cenário de degradação observado na Floresta Atlântica do país, onde os remanescentes florestais estão muito fragmentados, degradados e imersos em uma paisagem agrícola, dominada predominantemente por matrizes de cana de açúcar e pastagens. Neste contexto, objetivamos investigar se a regeneração natural da borda de fragmentos florestais tem características distintas nessas duas matrizes agrícolas dominantes. Nossa hipótese foi que de a regeneração natural em fragmentos inseridos em matrizes de cana de açúcar e pastagem são distintas, porque essas matrizes definem diferentes trajetórias de degradação sobre a flora de fragmentos naturais. Este trabalho foi desenvolvido na bacia do rio Corumbataí, onde indivíduos de espécies arbustivo e arbóreas do estrato regenerante (mínimo de 30cm de altura até 15 cm de perímetro à altura do peito) foram amostrados em 60 transectos de 30x2m (subdivididos em transectos contíguos de 10x2m) sistematicamente distribuídas a partir da borda de 12 fragmentos selecionados na bacia. Primeiramente, foi feita uma caracterização florística geral da regeneração natural das bordas de fragmentos florestais inseridos nas duas matrizes agrícolas. Posteriormente, foram utilizadas as variáveis número de indivíduos.hectare-1 (RN_HA) e número de espécies/m2 (NESP_M2) e diversidade florística (ISHANNON) para analisar as diferenças da estrutura e composição da regeneração natural nas duas matrizes. Essas mesmas variáveis foram utilizadas para analisar estatisticamente o efeito de indicadores de conservação dos fragmentos, como entrada de gado, presença de eucaliptos no dossel e abundância de gramíneas e lianas, sobre a regeneração natural. A similaridade florística entre remanescentes das paisagens foi calculada através do Índice de Similaridade de Jaccard. Em todos os transectos, foram amostrados 5886 regenerantes, divididos em 58 famílias, 220 espécies e 18 morfoespécies. O índice de Jaccard mostrou que houve similaridade florística (45%) entre os regenerantes dos fragmentos da matriz de pastagem e cana de açúcar. Nos fragmentos inseridos na matriz de pastagem, a entrada de gado foi o indicador de conservação de fragmentos que apresentou maior interferência sobre a regeneração natural, diminuindo tanto o número de indivíduos por unidade de área quanto a diversidade florística. Para a paisagem de cana, a fonte de variação que mais interferiu na estrutura da regeneração foi a presença de eucaliptos no dossel, que aumentou o número de indivíduos por unidade de área. A abundância de gramíneas e lianas e a presença de epífitas não apresentaram influência sobre a regeneração quando analisados tipo de matriz ou tamanho do fragmento. Entre os sub-transectos contíguos de 10m, não houve variação na estrutura e composição da regeneração natural, indicando que nesta profundidade de borda a vegetação regenerante é homogênea quanto à florística, diversidade e classificação sucessional. Os resultados indicam que, mesmo inseridos em paisagens agrícolas profundamente antropizadas, os remanescentes florestais ainda detém elevada diversidade florística. A regeneração natural mostrou-se um bom indicador de qualidade atual e futura dos remanescentes florestais, e demonstrou, juntamente com os indicadores de conservação, que os fragmentos são passíveis de manejo, o que poderá potencializar o papel de conservação da biodiversidade exercido pela regeneração natural. / At São Paulo state, Brazil, the scenario of the Atlantic Forest reflects the scenario of degradation of the biome in the country, where forest remnants are embedded in an agricultural landscape, dominated by sugar cane fields and pasture fields. In this context, we aimed to investigate if natural forest regeneration responds differently to each agricultural matrix. Our hypothesis was that forest regeneration composition in remnants surrounded by sugar cane fields is different of forest regeneration composition in remnants surrounded by pasture fields, because these agricultural matrices impact differently forest biota. The present research was developed in Corumbataí river basin, in the countryside of São Paulo state. Shrub and tree individuals, from regeneration layer, were sampled in 60 plots 30x2 meters long (subdivided in plots 10x2 meters long) systematically distributed on the edges of 12 fragments. Firstly, we characterized the natural regeneration of the edges of forest remnants embedded in agricultural landscapes. Then we used the variables number of individuals.ha-1 (RN_HA), number of species/m2 (NESP_M2) and floristic diversity (ISHANNON) to analyze the differences in structure and composition of forest natural regeneration between sugar cane fields and pasture fields. These variables were used to analyze the statistical effects of conservation indicators, such as cattle entrance on forest remnants, abundance of woody-vines and grasses, and presence of eucalyptus trees on forest canopy, on natural regeneration. Floristic similarity between forest remnants was assessed by Jaccard Similarity Index. 5886 saplings were sampled among the 60 plots, divided in 57 families, 214 species and 31 morphospecies. Jaccard index indicated floristic similarity (45%) among the fragments inserted in sugar cane field and pasture field. Among the fragments inserted in pasture field, cattle entrance was the conservation indicator with the strongest interference over natural regeneration, decreasing both the number of plants per area and the floristic diversity. At sugar cane landscape, presence of eucalyptus on forest canopy increased the number of individuals per area. Grass and woody-vines abundance and presence of epiphytes were not significant for any variable, when analyzed with matrix type or fragment size. Among the subplots of 10m long, there was no variation of structure and composition of natural regeneration, what indicates that, in relation to floristic, diversity and sucessional classification, the regenerant vegetation is homogeneous in the first 30 meters of the edge. The results indicate that, even embedded in anthropogenic agricultural landscapes, the remnants studied still retain high floristic diversity. Natural regeneration has shown to be a good indicator of actual and future quality of forest remnants, and, together with conservation indicators, demonstrated that the fragments studied are subject to management actions, which may improve the role of biodiversity conservancy developed by natural regeneration

No description available.

La Nouvelle-Calédonie qui présente une biodiversité à la fois exceptionnelle et très menacée, fait partie des points chauds de biodiversités définis à l'échelle globale comme zones prioritaires pour la conservation. Les incendies, d'origines anthropiques et constituant l'une des principales menaces pesant sur les écosystèmes naturels néo-calédoniens, conduisent à un recul des forêts denses humides (forêts par la suite) principalement au profit des savanes sur substrats volcano-sedimentaires. Au cours de cette thèse, les processus écologiques mis en jeu dans les dynamiques des forêts et des savanes, liés aux variations du régime d'incendie, ont été étudiés à différentes échelles spatiales et temporelles. L'analyse de la distribution spatio-temporelle des forêts et des savanes à l'échelle du paysage a mis en évidence différents facteurs dirigeant la dynamique paysagères et identifié des zones de recolonisation forestière. Différents modèles de succession secondaire et cortèges d'espèces pionnières ont été identifiés à partir de l'analyse de ces zones de recolonisation. La mesure de traits fonctionnels et l'utilisation de modèles semi-physiques de dommages causés par les incendies ont révélé une faible tolérance aux incendies de ces espèces. Enfin, une analyse multivariée de la structure, de la composition floristique et des conditions micro-climatiques des zones clés de transitions entre la savane et la forêt a permis de mieux comprendre les processus d'expansion et de contraction forestière. / New Caledonia, which presents both an exceptional and highly endangered biodiversity, is one of the worldwide biodiversity hotspots for conservation priority. Human-induced fires, which are one of the main threats to natural ecosystems in New Caledonia, lead to the expansion of savannas on volcano-sedimentary substrates at the expense of rainforests. In this thesis, the ecological processes, which are involved in the dynamics of rainforests and savannas and related to changes in fire regime, have been studied at different spatial and temporal scales. The analysis of the spatio-temporal distribution of rainforests and savannas across the landscape allowed us to both point-out the different drivers of their dynamics and to identify areas of rainforest recolonization. Different secondary succession patterns and pioneer species assemblages were identified from the analysis of these areas of recolonization. The analysis of measured functional traits and the use of semi-physical models of fire-caused damages highlighted the low tolerance of these pioneer species to fire. Finally, a multivariate analysis of the structure, the floristic composition and the micro-climatic conditions of transitional zones between savanna and rainforest, has allowed us to better understand the processes of rainforest's expansion and contraction. This thesis opens a new field of research in New Caledonia with important implications in rainforest's restoration and sustainable management.

<p>The aims of this thesis are to (1) analyze the predictability (indicators) of plant and fungal species diversity in old-growth forests, and (2) assess the history and biodiversity of woodland key habitats (WKHs) and their potential to maintain species diversity in fragmented boreal forest landscapes. </p><p>Predictability was explored in Granlandet nature reserve, an unexploited landscape composed of discrete old-growth Picea forest patches of varying size isolated by wetland, reflecting conditions of insular biota at stochastic equilibrium. Data from 46 patches (0.2-12 ha) showed that most species were rare. However, species richness and composition patterns exhibited a high degree of predictability, which strengthen the possibility to apply biodiversity indicators in old-growth forest stands. Area was a key factor. The increase in species richness starts to level out at 2-3 ha. Large patches host more Red-list species in their interiors than do small ones, i.e. stand size is an important qualitative aspect of old-growth habitat. Nestedness emerged in relation to area but also in equal-sized plots. Structural complexity and habitat quality were important for species richness and compositional patterns, and small habitats of high quality could harbor many rare species. Monitoring of wood-fungi on downed logs showed that species diversity on downed logs changed over periods of 5-10 years and that the occurrences of annual species were unpredictable. It is suggested that monitoring of species with durable fruit bodies (mainly polypores) is likely to be a feasible approach to obtain comparable data over time.</p><p>Assessments of biodiversity of WKHs were performed in two areas with contrasting histories of forest exploitation, namely in south boreal and north boreal Sweden. Analyses of the history of 15 south boreal WKHs showed that fire-suppression, selective logging until mid-20th century and abandonment by modern forestry has shaped their forest structure. These WKHs are not untouched forests, they lack key structural components and harbor few Red-list species. Artificial interventions to restore natural processes and patterns are needed to further increase their suitability for threatned species. Modeling analyses of species richness in 32 WKHs in north boreal Sweden, some of which have not been isolated by modern forestry until recently, indicated an excess of crustose lichen species, i.e. WKHs may face delayed species extinctions. By contrast, the results indicate that wood-fungi have tracked the environmental changes. Differences in substrate dynamics between epiphytes on living trees and species growing on decaying logs may explain the diffeence between species groups. The results indicate that population densities of Red-list species were low, which may result in further depletion of species diversity.</p><p>Continuing species declines and extinctions are likely if not conservation of WKHs are combined with other considerations in th managed forest landscape. Both WKHs and their surroundings must be managed and designed to maintain biodiversity over time. For a successful future conservation of boreal forest biodiversity monitoring of WKHs must be combined with monitoring of refeence areas. </p>

The aims of this thesis are to (1) analyze the predictability (indicators) of plant and fungal species diversity in old-growth forests, and (2) assess the history and biodiversity of woodland key habitats (WKHs) and their potential to maintain species diversity in fragmented boreal forest landscapes. Predictability was explored in Granlandet nature reserve, an unexploited landscape composed of discrete old-growth Picea forest patches of varying size isolated by wetland, reflecting conditions of insular biota at stochastic equilibrium. Data from 46 patches (0.2-12 ha) showed that most species were rare. However, species richness and composition patterns exhibited a high degree of predictability, which strengthen the possibility to apply biodiversity indicators in old-growth forest stands. Area was a key factor. The increase in species richness starts to level out at 2-3 ha. Large patches host more Red-list species in their interiors than do small ones, i.e. stand size is an important qualitative aspect of old-growth habitat. Nestedness emerged in relation to area but also in equal-sized plots. Structural complexity and habitat quality were important for species richness and compositional patterns, and small habitats of high quality could harbor many rare species. Monitoring of wood-fungi on downed logs showed that species diversity on downed logs changed over periods of 5-10 years and that the occurrences of annual species were unpredictable. It is suggested that monitoring of species with durable fruit bodies (mainly polypores) is likely to be a feasible approach to obtain comparable data over time. Assessments of biodiversity of WKHs were performed in two areas with contrasting histories of forest exploitation, namely in south boreal and north boreal Sweden. Analyses of the history of 15 south boreal WKHs showed that fire-suppression, selective logging until mid-20th century and abandonment by modern forestry has shaped their forest structure. These WKHs are not untouched forests, they lack key structural components and harbor few Red-list species. Artificial interventions to restore natural processes and patterns are needed to further increase their suitability for threatned species. Modeling analyses of species richness in 32 WKHs in north boreal Sweden, some of which have not been isolated by modern forestry until recently, indicated an excess of crustose lichen species, i.e. WKHs may face delayed species extinctions. By contrast, the results indicate that wood-fungi have tracked the environmental changes. Differences in substrate dynamics between epiphytes on living trees and species growing on decaying logs may explain the diffeence between species groups. The results indicate that population densities of Red-list species were low, which may result in further depletion of species diversity. Continuing species declines and extinctions are likely if not conservation of WKHs are combined with other considerations in th managed forest landscape. Both WKHs and their surroundings must be managed and designed to maintain biodiversity over time. For a successful future conservation of boreal forest biodiversity monitoring of WKHs must be combined with monitoring of refeence areas.

Direct numerical simulations of the receptivity and instability of boundary layers on flat and curved surfaces are herein reported. Various flow models are considered with the aim to capture aspects of flows over straight and swept wings such as wall curvature, pressure variations, leading-edge effects, streamline curvature and crossflow. The first model problem presented, the flow over a swept flat plate, features a crossflow inside the boundary layer. The layer is unstable to steady and traveling crossflow vortices which are nearly aligned with the free stream. Wall roughness and free-stream vortical modes efficiently excite these crossflow modes, and the associated receptivity mechanisms are linear in an environment of low-amplitude perturbations. Receptivity coefficients for roughness elements with various length scales and for free-stream vortical modes with different wavenumbers and frequencies are reported. Key to the receptivity to free-stream vorticity is the upstream excitation of streamwise streaks evolving into crossflow modes. This mechanism is also active in the presence of free-stream turbulence. The second flow model is that of a Görtler boundary layer. This flow type forms on surfaces with concave curvature, e.g. the lower side of a turbine blade. The dominant instability, driven by a vertically varying centrifugal force, appears as pairs of steady, streamwise counter-rotating vortical rolls and streamwise streaks. The Görtler boundary layer is in particular receptive to free-stream vortical modes with zero and low frequencies. The associated mechanism builds on the excitation of upstream disturbance streaks from which the Görtler modes emerge, similar to the mechanism in swept-plate flows. The receptivity to free-stream vorticity can both be linear and nonlinear. In the presence of free-stream turbulence, nonlinear receptivity is more likely to trigger steady Görtler vortices than linear receptivity unless the frequencies of the free-stream fluctuations are very low. The third set of simulations considers the boundary layer on a flat plate with an elliptic leading edge. This study aims to identify the effect of the leading edge on the boundary-layer receptivity to impinging free-stream vortical modes. Three types of modes with streamwise, vertical and spanwise vorticity are considered. The two former types trigger streamwise disturbance streaks while the latter type excites Tollmien-Schlichting wave packets in the shear layer. Simulations with two leading edges of different bluntness demonstrate that the leading-edge shape hardly influences the receptivity to streamwise vortices, whereas it significantly enhances the receptivity to vertical and spanwise vortices. It is shown that the receptivity mechanism to vertical free-stream vorticity involves vortex stretching and tilting - physical processes which are clearly enhanced by blunt leading edges. The last flow configuration studied models an infinite wing at 45 degrees sweep. This model is the least idealized with respect to applications in aerospace engineering. The set-up mimics the wind-tunnel experiments carried out by Saric and coworkers at the Arizona State University in the 1990s. The numerical method is verified by simulating the excitation of steady crossflow vortices through micron-sized roughness as realized in the experiments. Moreover, the receptivity to free-stream vortical disturbances is investigated and it is shown that the boundary layer is most receptive, if the free-stream modes are closely aligned with the most unstable crossflow mode / QC 20101025

Many natural areas and systems within urban landscapes are small or narrow. Landscape ecology studies within forested and agricultural landscapes have found that small natural areas that are protected from development or resource extraction through land use planning are significantly affected by adjacent land use changes. Some eventually lose the values for which they were protected. Studies also indicate that natural area boundary structures and functions are important determinants of the extent to which external threats affect adjacent natural areas. Few studies have empirically tested whether small or narrow urban natural areas that are protected from development through municipal land use planning are significantly affected by adjacent land use changes. However, municipal planners and forest managers are concerned that activities of residents living adjacent to the forest edge, commonly referred to as residential encroachment, may be degrading the social values, and ecological forms and functions of their woodlands. Studies have recorded evidence of human impacts within suburban forest edges, indicating that both recreation and yard-related activities are occurring and that these activities occur at significantly higher frequencies in the forest edge than in the interiors of these forests. However, no study has differentiated residential encroachment activities from those of other recreationists. In addition, although a number of municipalities have developed policies to address these activities, little is known about these policies, the extent to which they are implemented, or their effectiveness in protecting their small or narrow forested natural areas from residential encroachment activities. The principal research questions answered in this research are: 1) Do municipalities within Southern Ontario have policies for protecting natural areas from the activities of residents living adjacent to suburban forest edges? 2) To what extent are they implementing these policies? 3) What encroachment activities, if any, are occurring in Southern Ontario municipal forest edges? and 4) Are municipal boundary-related policies effective in limiting edge-resident encroachment activities? Using a mixed method approach, the research incorporates qualitative and quantitative data collection to answer these questions. The content analysis of official and secondary plans and social surveys of key informants within six Southern Ontario municipalities identify boundary-related policies for protecting municipal natural areas from residential encroachment activities. They also determine the extent to which the study municipalities implement these policies. Field studies in 40 forests within these municipalities used unobtrusive measurements of encroachment behaviour to describe encroachment activities under two implemented municipal boundary demarcation policies, and other boundary treatments The three research methods, together with a literature review, were used to determine whether Ontario municipal policies are effective in limiting edge-resident encroachment activities within municipal forest edges. The content analysis and interviews indicated that, in general, municipal policies were insufficient to address the edge-resident encroachment issue. Policies had been established, but not at a sufficiently authoritative policy level (i.e. the official plan level) to support their implementation by staff. In addition, policies were missing explicit goals, objectives and strategies to direct their implementation, and the municipalities had not integrated their disparate policy components into an integrated course of action through time and space. The municipalities were successful in implementing policies to prevent edge resident encroachment within natural areas adjacent to newly developing subdivisions. However, they had infrequently implemented their policies for preventing encroachment within natural areas adjacent to established subdivisions. Furthermore, all the municipalities were not frequently implementing their policies to remediate existing encroachments within natural areas adjacent to newly developing or established subdivisions. The unobtrusive measurement of encroachment behaviour confirmed that residential encroachment activities generated a housing effect zone of impact within municipal forest edges. The distribution of the evidence of encroachment was significantly biased to the forest border. Encroachment traces were highly prevalent within study forests, occurring in over 94% of sites and covering 26 to 50% of the sampled area. Encroachment traces were particularly intense in the first 8 metres from the forest border; but extended a mean maximum extent of 16 metres from the forest border, with 95% of the evidence of encroachment lying within 34 metres. Boundary type significantly affected the mean frequency, intensity and maximum extent of encroachment. Mean frequencies, intensities and extents of all encroachment, and of most encroachment categories, were generally higher in sites with boundary types that allowed edge residents ready access to the forest edge. Conversely, sites with boundary treatments that had barriers to entry, such as fences or grass strips, tended to have lower encroachment levels. Sites with multiple barriers, such as those with fences, grass strips and paths, tended to have the lowest mean frequencies, intensities and mean maximum extents of encroachment. While sites with implemented municipal post and fence policies had significantly lower mean frequencies, intensities and, in the case of fences extents of encroachment, they were not significantly different from those of sites under some of the boundary types not subject to municipal policies. They were also significantly higher than those of sites with fences and grass strips (with or without pathways). Sites with municipal posts had significantly lower mean intensities of encroachment than sites with other boundaries that enabled residents to enter the forest edge, and had significantly lower mean frequencies of waste disposal traces than fenced sites. Sites with fences also had significantly lower mean intensities of encroachment than sites with no boundary demarcation, or sites with fences and gates, and were particularly effective in reducing the incidence of yard extension encroachments, and mean maximum extents of encroachment. Despite the effectiveness of these boundary demarcation policies, and that of some of the other boundary treatments evaluated, none of the boundary treatments was effective in eliminating encroachment traces. A buffer of between 10 and 20 metres in width would be required to segregate the mean maximum extent of encroachment activities from sensitive forest edges, depending on the boundary demarcation policy, or type. The research concludes that current municipal policies are insufficient to meet the complexity and scope of the encroachment activities occurring. Some preventative policies have been developed and are regularly implemented within natural areas adjacent to new subdivisions. However, implemented boundary demarcation policies are insufficient to eliminate, or minimize residential encroachment. Wider more complex boundary policies that limit different types of encroachment and include elements that reduce access, spatially separate, and encourage informal residential surveillance (such as fences, grass strips and pathways) can further reduce encroachment levels. Few municipalities have established boundary demarcation policies to prevent encroachment within natural areas adjacent to established subdivisions, and study municipalities infrequently implement policies and bylaws to mitigate existing encroachments within these areas. Yet interviewees, and the results of the unobtrusive measurement of encroachment in study forest edges, indicate that encroachment activities are highly prevalent within these municipal forests. Policies at all levels, and particularly at the official plan level, are required to protect natural areas from edge resident encroachment, and other forms of post development impacts on natural areas. These policies are required to support the more rigorous enforcement of encroachment bylaws, and the negotiation, and implementation of effective buffers and boundary demarcation treatments. In consideration of these results and conclusions, the dissertation describes the implications for municipal planning policy and urban and regional planning theory, and provides recommendations for future research.

Many natural areas and systems within urban landscapes are small or narrow. Landscape ecology studies within forested and agricultural landscapes have found that small natural areas that are protected from development or resource extraction through land use planning are significantly affected by adjacent land use changes. Some eventually lose the values for which they were protected. Studies also indicate that natural area boundary structures and functions are important determinants of the extent to which external threats affect adjacent natural areas. Few studies have empirically tested whether small or narrow urban natural areas that are protected from development through municipal land use planning are significantly affected by adjacent land use changes. However, municipal planners and forest managers are concerned that activities of residents living adjacent to the forest edge, commonly referred to as residential encroachment, may be degrading the social values, and ecological forms and functions of their woodlands. Studies have recorded evidence of human impacts within suburban forest edges, indicating that both recreation and yard-related activities are occurring and that these activities occur at significantly higher frequencies in the forest edge than in the interiors of these forests. However, no study has differentiated residential encroachment activities from those of other recreationists. In addition, although a number of municipalities have developed policies to address these activities, little is known about these policies, the extent to which they are implemented, or their effectiveness in protecting their small or narrow forested natural areas from residential encroachment activities. The principal research questions answered in this research are: 1) Do municipalities within Southern Ontario have policies for protecting natural areas from the activities of residents living adjacent to suburban forest edges? 2) To what extent are they implementing these policies? 3) What encroachment activities, if any, are occurring in Southern Ontario municipal forest edges? and 4) Are municipal boundary-related policies effective in limiting edge-resident encroachment activities? Using a mixed method approach, the research incorporates qualitative and quantitative data collection to answer these questions. The content analysis of official and secondary plans and social surveys of key informants within six Southern Ontario municipalities identify boundary-related policies for protecting municipal natural areas from residential encroachment activities. They also determine the extent to which the study municipalities implement these policies. Field studies in 40 forests within these municipalities used unobtrusive measurements of encroachment behaviour to describe encroachment activities under two implemented municipal boundary demarcation policies, and other boundary treatments The three research methods, together with a literature review, were used to determine whether Ontario municipal policies are effective in limiting edge-resident encroachment activities within municipal forest edges. The content analysis and interviews indicated that, in general, municipal policies were insufficient to address the edge-resident encroachment issue. Policies had been established, but not at a sufficiently authoritative policy level (i.e. the official plan level) to support their implementation by staff. In addition, policies were missing explicit goals, objectives and strategies to direct their implementation, and the municipalities had not integrated their disparate policy components into an integrated course of action through time and space. The municipalities were successful in implementing policies to prevent edge resident encroachment within natural areas adjacent to newly developing subdivisions. However, they had infrequently implemented their policies for preventing encroachment within natural areas adjacent to established subdivisions. Furthermore, all the municipalities were not frequently implementing their policies to remediate existing encroachments within natural areas adjacent to newly developing or established subdivisions. The unobtrusive measurement of encroachment behaviour confirmed that residential encroachment activities generated a housing effect zone of impact within municipal forest edges. The distribution of the evidence of encroachment was significantly biased to the forest border. Encroachment traces were highly prevalent within study forests, occurring in over 94% of sites and covering 26 to 50% of the sampled area. Encroachment traces were particularly intense in the first 8 metres from the forest border; but extended a mean maximum extent of 16 metres from the forest border, with 95% of the evidence of encroachment lying within 34 metres. Boundary type significantly affected the mean frequency, intensity and maximum extent of encroachment. Mean frequencies, intensities and extents of all encroachment, and of most encroachment categories, were generally higher in sites with boundary types that allowed edge residents ready access to the forest edge. Conversely, sites with boundary treatments that had barriers to entry, such as fences or grass strips, tended to have lower encroachment levels. Sites with multiple barriers, such as those with fences, grass strips and paths, tended to have the lowest mean frequencies, intensities and mean maximum extents of encroachment. While sites with implemented municipal post and fence policies had significantly lower mean frequencies, intensities and, in the case of fences extents of encroachment, they were not significantly different from those of sites under some of the boundary types not subject to municipal policies. They were also significantly higher than those of sites with fences and grass strips (with or without pathways). Sites with municipal posts had significantly lower mean intensities of encroachment than sites with other boundaries that enabled residents to enter the forest edge, and had significantly lower mean frequencies of waste disposal traces than fenced sites. Sites with fences also had significantly lower mean intensities of encroachment than sites with no boundary demarcation, or sites with fences and gates, and were particularly effective in reducing the incidence of yard extension encroachments, and mean maximum extents of encroachment. Despite the effectiveness of these boundary demarcation policies, and that of some of the other boundary treatments evaluated, none of the boundary treatments was effective in eliminating encroachment traces. A buffer of between 10 and 20 metres in width would be required to segregate the mean maximum extent of encroachment activities from sensitive forest edges, depending on the boundary demarcation policy, or type. The research concludes that current municipal policies are insufficient to meet the complexity and scope of the encroachment activities occurring. Some preventative policies have been developed and are regularly implemented within natural areas adjacent to new subdivisions. However, implemented boundary demarcation policies are insufficient to eliminate, or minimize residential encroachment. Wider more complex boundary policies that limit different types of encroachment and include elements that reduce access, spatially separate, and encourage informal residential surveillance (such as fences, grass strips and pathways) can further reduce encroachment levels. Few municipalities have established boundary demarcation policies to prevent encroachment within natural areas adjacent to established subdivisions, and study municipalities infrequently implement policies and bylaws to mitigate existing encroachments within these areas. Yet interviewees, and the results of the unobtrusive measurement of encroachment in study forest edges, indicate that encroachment activities are highly prevalent within these municipal forests. Policies at all levels, and particularly at the official plan level, are required to protect natural areas from edge resident encroachment, and other forms of post development impacts on natural areas. These policies are required to support the more rigorous enforcement of encroachment bylaws, and the negotiation, and implementation of effective buffers and boundary demarcation treatments. In consideration of these results and conclusions, the dissertation describes the implications for municipal planning policy and urban and regional planning theory, and provides recommendations for future research.

Alterations caused by the edge effects in a fragment have direct influence over the fauna and can determine the composition and distribution of the small non-flying mammals assemblage in a landscape. This study had as aim of analyzing the edge effect over the structure of a small mammal assemblege and the circumjacent environment in a decidual forest fragment located in the southern limit of the Atlantic Forest, in the center state of Rio Grande do Sul. Five field phases were carried out, with an effort of 6360 trap-night, and 1060 pitfall-night, totalizing 78 captured individuals (99 captures), belonging to four species of rodents and two species of marsupials. The success of capture of live traps was 1.2% and of pitfall traps was of 2.3%. Rodents were responsible for 97.4% of all captures and marsupials, 2.6%. In pitfall traps, it was observed high abundance of rodents in the interior in relation to the edge. The Analysis of Multivariada Variance by Randomization showed that there is no significant difference in the composition of species in relation to distances of the edge, despite the Analysis of Principal Components has evidenced some trends, grouping the species in accordance to the distances where they had occurred more frequently. Thus, the assemblege of small mammals was dominated by a species of rodent, Oligoryzomys nigripes, which represented 73% of all captures, appearing in all distances of the edge, but mostly in the forest interior. On the other hand, marsupials appeared only in the interior of the fragment, being A. montensis registered from 130 m from the edge, seeming to prefer habitats with denser forest cover. By using MANOVA, and on the basis of the biotic and abiotic factors analyzed along the edge gradient, it was possible to establish different distances of penetration of the edge that had varied from 10 to 160 meters. The Congruence Analysis showed that environmental variables which best explained the distribution of small mammas in the gradient was the feeding resources, lianas, trees and canopy cover. Thus, species of small mammals seem to use the modified environment in accordance to their adaptations to the available micro-habitats in the gradient edge-interior, having a noticeable edge effect in the study area. / Alterações causadas pelos efeitos de borda em um fragmento têm influência direta sobre a fauna e podem determinar a composição e distribuição da assembléia de pequenos mamíferos em uma paisagem. Este estudo teve como objetivo analisar o efeito de borda sobre a riqueza e abundância de pequenos mamíferos e o ambiente circunjacente em um fragmento de Floresta Estacional Decidual localizado no limite sul da Floresta Atlântica, no centro do Rio Grande do Sul. Foram realizadas cinco fases de campo, com esforço amostral de 6360 armadilhas-noite, e 1060 armadilhas de quedanoite, totalizando 78 indivíduos capturados (99 capturas), pertencentes a quatro espécies de roedores e duas de marsupiais. O sucesso de captura para as armadilhas de metal foi de 1,2% e para as armadilhas de queda foi de 2,3%. Os roedores representaram 97,4% das capturas e os marsupiais 2,6%. Nas armadilhas de queda, observou-se maior riqueza e abundância de roedores no interior em relação à borda. A Análise de Variância Multivariada via Aleatorização mostrou que não houve diferença significativa na composição de espécies em relação às distancias da borda ainda que a Análise de Componentes Principais tenha evidenciado algumas tendências agrupando as espécies de acordo com as distâncias onde elas ocorreram com maior freqüência. Assim, a assembléia de pequenos mamíferos esteve dominada por uma espécie de roedor, Oligoryzomys nigripes, que representou 73% do total de capturas, aparecendo em todas distâncias da borda, embora com uma tendência a ser mais florestal. Por outro lado, os marsupiais apareceram apenas no interior do fragmento, ao passo que A. montensis foi registrado a partir de 130 m do início da borda parecendo preferir ambientes com cobertura florestal mais densa. Com base nos fatores bióticos e abióticos que foram usados para estimar o término da borda e início do interior do fragmento foi possível estabelecer, através da MANOVA, diferentes distâncias de penetração da borda que variaram de 10 a 160 metros. A análise de congruência mostrou que as variáveis ambientais que melhor explicaram a distribuição dos animais no gradiente foram os recursos alimentares, lianas, árvores e cobertura do dossel. Assim, as espécies de pequenos mamíferos parecem utilizar o ambiente alterado de acordo com suas adaptações aos micro-hábitats disponíveis no gradiente bordainterior.