Este estudo mostrou os efeitos da fragmentação sobre a avifauna considerando diferentes aspectos, levando em conta escalas espaciais e temporais distintas e os diferentes elementos que compõem uma paisagem fragmentada. Verificamos a influência das características intrínsecas dos fragmentos, da configuração das paisagens e da variação geográfica sobre as respostas das espécies à fragmentação. Igualmente importante está a variação temporal; modificações da estrutura da paisagem passada têm reflexos na distribuição e abundâncias das aves na paisagem atualmente e a inclusão dessa variável nas análises nos permite avaliar de maneira mais precisa o grau de ameaça que as espécies sofrem nas paisagens alteradas pela ocupação humana. O primeiro capítulo mostra que no Pontal do Paranapanema o tamanho dos fragmentos e a estrutura da vegetação são os fatores principais que afetam a avifauna na região. Isso acontece devido ao grau de isolamentos intenso dos remanescentes, que devem impedir com que a maior parte das espécies florestais se movimentem entre fragmentos, tornando as características intrínsecas dessas áreas mais importantes. Vimos também que uma grande proporção das espécies florestais (~41%) é afetada ou altamente afetada pela redução e fragmentação do habitat, e que, em geral, essas espécies fazem parte de grupos funcionais mais vulneráveis, entre eles: as espécies endêmicas da Mata Atlântica, que estão próximas do limite de distribuição, que estão fora de seus centros de abundância e que têm baixa flexibilidade no uso de florestas. No segundo capítulo, avaliamos o papel de bosques agroflorestais funcionando como trampolins ecológicos (“stepping stones"). Encontramos poucas evidências de que esses elementos influenciem a composição das comunidades de aves na escala da paisagem ou para os remanescentes grandes da região. Espécies generalistas e de áreas abertas foram as mais freqüentemente observadas nos bosques agroflorestais e nos outros elementos de conexão (fragmentos pequenos e corredores ripários), enquanto apenas poucas espécies florestais foram observadas nessas áreas. No entanto, esses elementos são mais ricos em espécies dos que a matriz de pasto, o que mostra a desvantagem desse sistema tradicional de monocultura instalada na região desde o início de sua ocupação. Propomos ainda uma hipótese geral sobre a eficiência dos trampolins ecológicos que ocorre em resistências intermediárias da matriz; quando a matriz é muito resistente, muito contrastante com o habitat natural, como para as espécies florestais em nosso estudo, os indivíduos não conseguem utilizar esses elementos por não conseguirem sair dos fragmentos; em oposição, em matrizes muito permeáveis, provavelmente a necessidade dos trampolins é diminuída porque a movimentações podem ocorrer pela própria matriz; dessa forma, apenas em condições intermediárias os trampolins ecológicos terão uma eficiência ótima, como para as espécies generalistas em nosso estudo; claramente, essa relação é espécie-dependente. Observamos ainda que a distância dos bosques aos remanescentes florestais é determinante na composição, riqueza e abundância de espécies que usam esses elementos. Os trampolins devem ser considerados como alternativa de manejo da paisagem, sobretudo em áreas em que não é possível a implantação de corredores e a matriz tenha uma permeabilidade adequada. No terceiro capítulo, verificamos que a estrutura espacial da paisagem do Pontal de 1978, explica melhor as riquezas de grupos funcionais mais sensíveis do que a estrutura mais atual (2003), evidenciando um tempo de latência da resposta das espécies de cerca de 25 anos. Contrariamente, as espécies não-sensíveis estiveram mais relacionadas com a paisagem mais recente. Esse resultado mostra que as espécies sensíveis ainda não se estabilizaram nos fragmentos e que mais perdas poderão acontecer mesmo sem novos avanços da fragmentação do habitat na região. Processos semelhantes devem acontecer em outras regiões da Mata Atlântica, com o agravante de que, em muitos casos, o tempo de isolamento dos remanescentes é maior e, portanto, esse processo de perda está mais adiantado. No quarto capítulo estimamos quanto é o débito ecológico provocado pelo processo de fragmentação, i.e. quantas espécies devem ser perdidas devido a esse processo; e qual é o tempo de latência para a perda das espécies. Verificamos que o débito é muito alto especialmente para as espécies sensíveis, as endêmicas e as que têm baixa abundância relativa. Muitas dessas espécies já foram perdidas, principalmente nos fragmentos menores e de qualidade inferior, em que o tempo de latência é mais curto; e, muitas outras serão perdidas em um futuro próximo nos fragmentos maiores. O tempo de latência que estimamos é menor do que o encontrado em outro estudo com aves tropicais da África, ilustrando que a urgência para impedir o desaparecimento das espécies é maior do que se pensava. Finalmente, no último capítulo verificamos que as respostas das espécies à fragmentação dependem não somente da configuração da paisagem em uma certa região, mas também da sua localização geográfica. As respostas das espécies podem variar ao longo de sua distribuição geográfica e, em geral, elas são mais vulneráveis quando estão próximas da borda de distribuição, sobretudo as endêmicas. De forma semelhante, aquelas que estão fora dos seus centros de abundância também apresentam uma sensibilidade mais elevada. As espécies com baixa abundância são afetadas independentemente de suas localizações. Verificamos também que em paisagens mais conectadas, a área dos fragmentos tem uma importância secundária e o limiar de tamanho para a ocorrência das espécies mais sensíveis é muito menor nessas situações em comparação com paisagens onde os remanescentes estão muito isolados. / We evaluated the effect of Atlantic Forest fragmentation on bird community, functional groups and populations, in the Pontal do Paranapanema region. Forest bird survey using point counts was conducted during the period of 2002-2004, in 21 patches and seven sites located in a control reserve, the Morro do Diabo State Park (36,000 ha). We identified four degrees of species’ sensitivity to fragmentation: highly affected, affected, not affected and benefited. Four functional groups revealed to be more vulnerable: Atlantic Forest endemic species; those that are near the edge of their geographical distribution; species that have low flexibility on forest type use and species that are outside of their center of abundance. The intrinsic patch parameters, such as size and quality were the most important, affecting a higher number of species, while the degree of isolation explained little about bird parameters variation. This result must be due to the intense isolation imposed to the remnants in the region, caused by the long distance between patches and the low permeability of the matrix, what make the local variables more relevant. This study shows the need to protect and manage the forest remnants to improve habitat quality. Complementarily, landscape management should be conducted to minimize the isolation of forest remnants. The highest priority should be given to the State Park and the largest patches which aggregate the highest diversity. Keywords: fragmentation, Landscape Ecology, conservation, biodiversity, species Sensitivity. The aim of this study was to assess whether stepping stones (agroforest woodlots) can increase landscape connectivity, influencing bird distribution and diversity at landscape and local levels. We also tested which categories of species can use different types of connection elements, and whether this use is influenced by the distance from these elements to forest patches. We studied two fragmented landscapes, with and without stepping stones linking large fragments, and one forested landscape. Using a point count, a bird survey was undertaken in the fragmented landscapes in five different elements: large remnants (>400 ha), agroforest woodlots (0.4-1.1 ha), small patches (0.5 – 7 ha), riparian corridor, and pasture areas (the main matrix). The latter three elements were compared with the stepping stones, in regard to their role in bird movements. The continuous forest was surveyed as a control for large remnants in the fragmented landscapes. We observed no effect of the agroforest systems at the landscape scale or for large forest remnants. Generalist and open-area species were commonly observed in the agroforest system or other connection elements, whereas only a few forest species were present in these connections. The matrix was the poorest element of all. The distance between the woodlots and source areas (large patches) was essential to determine forest species richness and abundance in the agroforest system. We suggest that there is an optimal relationship between the permeability of the matrix and the efficiency of stepping stones, which occurs at intermediate degrees of matrix resistance, and is, clearly, species-dependent. The agroforest system appeared to be more advantageous for conservation than the traditional monoculture system in the region, as they hold higher bird diversity; for this reason, they should be considered as a management alternative, particularly when the corridor implantation is not feasible and the matrix permeability requirement is met. Keywords: Fragmentation, Landscape Ecology, Conservation, Biodiversity, Landscape Management. Landscape dynamic is an important factor effecting biodiversity, especially in humandominated landscapes. The aim of this study was to verify how the historical processes of fragmentation in the Pontal do Paranapanema region, inside the Atlantic Forest domain, influence bird diversity pattern. Maps from the region were produced for each decade: 1957, 1965, 1978, 1984, 1993 and 2003; using different types of information such as Satellite images, topographic maps and aerial photographies. For each map we calculated landscape indices which represent patch size and isolation, and related those metrics with bird community parameters. Birds were surveyed using point count methodology in 21 patches plus seven sites in a control reserve (The Morro do Diabo State Park, with 36,000 ha). Patches varied in size: seven large (>400 ha), seven medium (100 - 200 ha) and seven small (30 – 80 ha). Species richness was calculated for forest and generalist species and for different functional groups. Model selection, based on Akaike’s Information Criterion, was performed to distinguish which landscape indices, older or recent ones, explain better species richness. For non sensitive species the degree of isolation of the most recent landscape (2003) explains better the richness variation, although with a higher uncertainty in model selection. On the contrary, landscape parameters from 1978, especially patch size, were the most important variables to explain the richness of the most sensitive groups of species, what provides an evidence for a time lag of bird response to fragmentation of about 25 years. This time also corresponds to the period of intense fragmentation process in the region, when, the forest cover was reduced to < 30%, the theoretical fragmentation threshold. This time lag is a conservation concern because it indicates that the sensitive species did not get yet into a steady state condition and extinctions might occur in a near future if no landscape management is carried out to revert the negative consequences of fragmentation. This situation must even be more critical for other landscapes in the Atlantic Forest domain where the time since isolation is longer. Keywords: fragmentation, landscape dynamics, conservation, ecological debt. The aim of this study was to estimate the time lag between the fragmentation of an Atlantic forest region and the extinction of sensitive bird species in forest patches and to verify which landscape parameters might be associated with such a delay. We analyzed 21 patches varying in size (30-2,000 ha) and seven sites inside a continuous forest (control, ~36,000 ha). The original, present and future bird richness in each site were estimated using, respectively, data from literature, bird survey using point counts and a species-area relationship. A landscape data set of the historical process of fragmentation in the region was used to estimate patch size and time since isolation and the total forest cover in the landscape before fragmentation had taken place. We estimated a very high bird loss in forest patches, mainly concerning species sensitive to disturbances, endemics for the Atlantic Forest and with low relative abundance. The observed halflife, the time necessary to loose half of the species before the relaxation process, varied, on average, between 20 and 46 years. This period was shorter than those found in another study for birds in an African tropical forest (23 to 80 years), indicating that we have less time to revert this situation than previously predicted. The patch quality and size are both important factors determining the time lag for sensitive species. We have estimated that several species had already gone extinct, particularly in small and low quality patches, and that other ones will disappear in the future, even in the largest patches. Our results allowed us to speculate about other Atlantic Forest landscapes, where longer time since isolation and smaller patches can cause even higher losses. The synchronism of extinctions occurring in many regions probably will increase the list of threatened species in a near future, including many that today seem not likely to be vulnerable. Keywords: Landscape dynamic, conservation, extintion debt, theory of Island Biogeography, species-area relationship. Generally the effect of fragmentation on species focus on the landscape or regional scale, removing the geographic variation effect, nonetheless the addition of this variable might help us understand how species respond to fragmentation in a general form. Based on this idea, we compared the effects of forest fragmentation on bird species sensitive to habitat reduction and segmentation, functional groups and coincident species that occur in three landscapes in two geographically distant regions: the Pontal do Paranapanema region, in the extreme west of the State of São Paulo; and Ibiúna e Caucaia, 700 km to the east of the state. We undertook bird survey, using point count methodology in 37 forest fragment and 11 sites inside two continuous forests. The numbers of sample in each landscape were: Pontal – 21 patches and 7 sites in a control area; and Ibiúna and Caucaia – 8 patches in each landscape and 4 sites in a same control area. In the landscape of Pontal, bird species suffer the most intense effects of fragmentation, although patches are larger than in the other landscapes. The most sensitive species are: endemic, on the limit of distribution, out of abundance center and with low abundance. Many of them are affected only in the Pontal region, result of its geographical location. The spatial structure of landscape explains much of the species diversity and composition variation in the sampled areas. In Pontal and Ibiúna, patch size is the most important variable and, in Caucaia, the high degree of connectivity is the most relevant, making the species less vulnerable. Furthermore, because of its high connectivity, the patch size threshold that defines species loss, is much smaller in Caucaia than in Pontal (50 ha e 300-400 ha, respectively). The comparison of the effects of fragmentation on birds in distinct landscape shows that species responses varies along its distribution, which are constrained to local landscape parameters and bird features. This results show that there is no general recipes for biodiversity conservation, but different strategies must be used according to the regional condition. Keywords: connectivity, fragmentation, Landscape Ecology, conservation, geographical Variation.

Around the world, forests are generally being fragmented and as plant populations decline in both sizeand number of individuals, the genetic variation will also decrease whitch in the end leads to totalextinction of the plant. In this paper, I wanted to investigate what specific species of trees and shrubsyou usually find in the edge and core habitats in some of the forests in the Baltic Sea region. I alsowanted to identify the edge effects and their sources and finally compare the results between thedifferent regions.The data sources used are derived from the ongoining research project Landscape structures, patterns ofbiodiversity and conservation strategies in the Baltic Sea region by Kari Lehtilä, Patrik Dinnétz ochTiina Vinter who have made a record of all herbaceous plant species in four randomly chosen forestregions situated around the Baltic Sea: Estonia, Skåne, Södermanland and Germany. I made statisticalanalyses, using the computer program R, concerning the distribution of species in the edge habitats ofthe forests as well as in the core habitats. I performed generalized linear models and then variationanalyses with the factors light and biomass in order to investigate wether edge effects might appear dueto these factors. I also used the ecological characteristics – Ellenberg indicators – light, soil pH andnitrogen performing linear regression models. I then compared the results – within the regions but alsobetween them.As the results showed great variancy both within the regions and between them no general explanationto the edge effects could be affirmed. The Ellenberg analyses showed no difference between the species,i. e. the ecological characteristics could not explain why some species were associated to edge habitatsand some were not. All four regions had in common that the greatest biodiversity were to be found inthe edge habitats, which is characteristical for a habitat in an early successional state. Due to the thefeatures of the forests, the specific species that received the label ”edge species” or ”core species” werenot always mutual in all regions. For example Quercus robur turned out to be a typical forest species inGermany while in Estonia it was strongly associated to edge habitats and in this investigation one mightdiscuss later states of successions in Germany rather than stronger edge effects in Estonia.

Dans ce travail, nous nous sommes intéressés à l'étude asymptotique d'objets combinatoires aléatoires. Deux thèmes ont particulièrement retenu notre attention : les cartes planaires aléatoires et les modèles combinatoires liés à la théorie des fragmentations. La théorie mathématique des cartes planaires aléatoires est née à l'aube de notre millénaire avec les travaux pionniers de Benjamini & Schramm, Angel & Schramm et Chassaing & Schaeffer. Elle a ensuite beaucoup progressé, mais à l'heure où ces lignes sont écrites, de nombreux problèmes fondamentaux restent ouverts. Résumons en quelques mots clés nos principales contributions dans le domaine : l'introduction et l'étude du cactus brownien (avec J.F. Le Gall et G. Miermont), l'étude de la quadrangulation infinie uniforme vue de l'infini (avec L. Ménard et G. Miermont), ainsi que des travaux plus théoriques sur les graphes aléatoires stationnaires d'une part et les graphes empilables dans $\R^d$ d'autre part (avec I. Benjamini). La théorie des fragmentations est beaucoup plus ancienne et remonte à des travaux de Kolmogorov (1941) et de Filippov (1961). Elle est maintenant bien développée (voir par exemple l'excellent livre de J. Bertoin), et nous ne nous sommes pas focalisés sur cette théorie mais plutôt sur ses applications à des modèles combinatoires. Elle s'avère en effet très utile pour étudier différents modèles de triangulations récursives du disque (travail effectué avec J.F. Le Gall) et les recherches partielles dans les quadtrees (travail effectué avec A. Joseph).

Este estudo mostrou os efeitos da fragmentação sobre a avifauna considerando diferentes aspectos, levando em conta escalas espaciais e temporais distintas e os diferentes elementos que compõem uma paisagem fragmentada. Verificamos a influência das características intrínsecas dos fragmentos, da configuração das paisagens e da variação geográfica sobre as respostas das espécies à fragmentação. Igualmente importante está a variação temporal; modificações da estrutura da paisagem passada têm reflexos na distribuição e abundâncias das aves na paisagem atualmente e a inclusão dessa variável nas análises nos permite avaliar de maneira mais precisa o grau de ameaça que as espécies sofrem nas paisagens alteradas pela ocupação humana. O primeiro capítulo mostra que no Pontal do Paranapanema o tamanho dos fragmentos e a estrutura da vegetação são os fatores principais que afetam a avifauna na região. Isso acontece devido ao grau de isolamentos intenso dos remanescentes, que devem impedir com que a maior parte das espécies florestais se movimentem entre fragmentos, tornando as características intrínsecas dessas áreas mais importantes. Vimos também que uma grande proporção das espécies florestais (~41%) é afetada ou altamente afetada pela redução e fragmentação do habitat, e que, em geral, essas espécies fazem parte de grupos funcionais mais vulneráveis, entre eles: as espécies endêmicas da Mata Atlântica, que estão próximas do limite de distribuição, que estão fora de seus centros de abundância e que têm baixa flexibilidade no uso de florestas. No segundo capítulo, avaliamos o papel de bosques agroflorestais funcionando como trampolins ecológicos (“stepping stones”). Encontramos poucas evidências de que esses elementos influenciem a composição das comunidades de aves na escala da paisagem ou para os remanescentes grandes da região. Espécies generalistas e de áreas abertas foram as mais freqüentemente observadas nos bosques agroflorestais e nos outros elementos de conexão (fragmentos pequenos e corredores ripários), enquanto apenas poucas espécies florestais foram observadas nessas áreas. No entanto, esses elementos são mais ricos em espécies dos que a matriz de pasto, o que mostra a desvantagem desse sistema tradicional de monocultura instalada na região desde o início de sua ocupação. Propomos ainda uma hipótese geral sobre a eficiência dos trampolins ecológicos que ocorre em resistências intermediárias da matriz; quando a matriz é muito resistente, muito contrastante com o habitat natural, como para as espécies florestais em nosso estudo, os indivíduos não conseguem utilizar esses elementos por não conseguirem sair dos fragmentos; em oposição, em matrizes muito permeáveis, provavelmente a necessidade dos trampolins é diminuída porque a movimentações podem ocorrer pela própria matriz; dessa forma, apenas em condições intermediárias os trampolins ecológicos terão uma eficiência ótima, como para as espécies generalistas em nosso estudo; claramente, essa relação é espécie-dependente. Observamos ainda que a distância dos bosques aos remanescentes florestais é determinante na composição, riqueza e abundância de espécies que usam esses elementos. Os trampolins devem ser considerados como alternativa de manejo da paisagem, sobretudo em áreas em que não é possível a implantação de corredores e a matriz tenha uma permeabilidade adequada. No terceiro capítulo, verificamos que a estrutura espacial da paisagem do Pontal de 1978, explica melhor as riquezas de grupos funcionais mais sensíveis do que a estrutura mais atual (2003), evidenciando um tempo de latência da resposta das espécies de cerca de 25 anos. Contrariamente, as espécies não-sensíveis estiveram mais relacionadas com a paisagem mais recente. Esse resultado mostra que as espécies sensíveis ainda não se estabilizaram nos fragmentos e que mais perdas poderão acontecer mesmo sem novos avanços da fragmentação do habitat na região. Processos semelhantes devem acontecer em outras regiões da Mata Atlântica, com o agravante de que, em muitos casos, o tempo de isolamento dos remanescentes é maior e, portanto, esse processo de perda está mais adiantado. No quarto capítulo estimamos quanto é o débito ecológico provocado pelo processo de fragmentação, i.e. quantas espécies devem ser perdidas devido a esse processo; e qual é o tempo de latência para a perda das espécies. Verificamos que o débito é muito alto especialmente para as espécies sensíveis, as endêmicas e as que têm baixa abundância relativa. Muitas dessas espécies já foram perdidas, principalmente nos fragmentos menores e de qualidade inferior, em que o tempo de latência é mais curto; e, muitas outras serão perdidas em um futuro próximo nos fragmentos maiores. O tempo de latência que estimamos é menor do que o encontrado em outro estudo com aves tropicais da África, ilustrando que a urgência para impedir o desaparecimento das espécies é maior do que se pensava. Finalmente, no último capítulo verificamos que as respostas das espécies à fragmentação dependem não somente da configuração da paisagem em uma certa região, mas também da sua localização geográfica. As respostas das espécies podem variar ao longo de sua distribuição geográfica e, em geral, elas são mais vulneráveis quando estão próximas da borda de distribuição, sobretudo as endêmicas. De forma semelhante, aquelas que estão fora dos seus centros de abundância também apresentam uma sensibilidade mais elevada. As espécies com baixa abundância são afetadas independentemente de suas localizações. Verificamos também que em paisagens mais conectadas, a área dos fragmentos tem uma importância secundária e o limiar de tamanho para a ocorrência das espécies mais sensíveis é muito menor nessas situações em comparação com paisagens onde os remanescentes estão muito isolados. / We evaluated the effect of Atlantic Forest fragmentation on bird community, functional groups and populations, in the Pontal do Paranapanema region. Forest bird survey using point counts was conducted during the period of 2002-2004, in 21 patches and seven sites located in a control reserve, the Morro do Diabo State Park (36,000 ha). We identified four degrees of species’ sensitivity to fragmentation: highly affected, affected, not affected and benefited. Four functional groups revealed to be more vulnerable: Atlantic Forest endemic species; those that are near the edge of their geographical distribution; species that have low flexibility on forest type use and species that are outside of their center of abundance. The intrinsic patch parameters, such as size and quality were the most important, affecting a higher number of species, while the degree of isolation explained little about bird parameters variation. This result must be due to the intense isolation imposed to the remnants in the region, caused by the long distance between patches and the low permeability of the matrix, what make the local variables more relevant. This study shows the need to protect and manage the forest remnants to improve habitat quality. Complementarily, landscape management should be conducted to minimize the isolation of forest remnants. The highest priority should be given to the State Park and the largest patches which aggregate the highest diversity. Keywords: fragmentation, Landscape Ecology, conservation, biodiversity, species Sensitivity. The aim of this study was to assess whether stepping stones (agroforest woodlots) can increase landscape connectivity, influencing bird distribution and diversity at landscape and local levels. We also tested which categories of species can use different types of connection elements, and whether this use is influenced by the distance from these elements to forest patches. We studied two fragmented landscapes, with and without stepping stones linking large fragments, and one forested landscape. Using a point count, a bird survey was undertaken in the fragmented landscapes in five different elements: large remnants (>400 ha), agroforest woodlots (0.4-1.1 ha), small patches (0.5 – 7 ha), riparian corridor, and pasture areas (the main matrix). The latter three elements were compared with the stepping stones, in regard to their role in bird movements. The continuous forest was surveyed as a control for large remnants in the fragmented landscapes. We observed no effect of the agroforest systems at the landscape scale or for large forest remnants. Generalist and open-area species were commonly observed in the agroforest system or other connection elements, whereas only a few forest species were present in these connections. The matrix was the poorest element of all. The distance between the woodlots and source areas (large patches) was essential to determine forest species richness and abundance in the agroforest system. We suggest that there is an optimal relationship between the permeability of the matrix and the efficiency of stepping stones, which occurs at intermediate degrees of matrix resistance, and is, clearly, species-dependent. The agroforest system appeared to be more advantageous for conservation than the traditional monoculture system in the region, as they hold higher bird diversity; for this reason, they should be considered as a management alternative, particularly when the corridor implantation is not feasible and the matrix permeability requirement is met. Keywords: Fragmentation, Landscape Ecology, Conservation, Biodiversity, Landscape Management. Landscape dynamic is an important factor effecting biodiversity, especially in humandominated landscapes. The aim of this study was to verify how the historical processes of fragmentation in the Pontal do Paranapanema region, inside the Atlantic Forest domain, influence bird diversity pattern. Maps from the region were produced for each decade: 1957, 1965, 1978, 1984, 1993 and 2003; using different types of information such as Satellite images, topographic maps and aerial photographies. For each map we calculated landscape indices which represent patch size and isolation, and related those metrics with bird community parameters. Birds were surveyed using point count methodology in 21 patches plus seven sites in a control reserve (The Morro do Diabo State Park, with 36,000 ha). Patches varied in size: seven large (>400 ha), seven medium (100 - 200 ha) and seven small (30 – 80 ha). Species richness was calculated for forest and generalist species and for different functional groups. Model selection, based on Akaike’s Information Criterion, was performed to distinguish which landscape indices, older or recent ones, explain better species richness. For non sensitive species the degree of isolation of the most recent landscape (2003) explains better the richness variation, although with a higher uncertainty in model selection. On the contrary, landscape parameters from 1978, especially patch size, were the most important variables to explain the richness of the most sensitive groups of species, what provides an evidence for a time lag of bird response to fragmentation of about 25 years. This time also corresponds to the period of intense fragmentation process in the region, when, the forest cover was reduced to < 30%, the theoretical fragmentation threshold. This time lag is a conservation concern because it indicates that the sensitive species did not get yet into a steady state condition and extinctions might occur in a near future if no landscape management is carried out to revert the negative consequences of fragmentation. This situation must even be more critical for other landscapes in the Atlantic Forest domain where the time since isolation is longer. Keywords: fragmentation, landscape dynamics, conservation, ecological debt. The aim of this study was to estimate the time lag between the fragmentation of an Atlantic forest region and the extinction of sensitive bird species in forest patches and to verify which landscape parameters might be associated with such a delay. We analyzed 21 patches varying in size (30-2,000 ha) and seven sites inside a continuous forest (control, ~36,000 ha). The original, present and future bird richness in each site were estimated using, respectively, data from literature, bird survey using point counts and a species-area relationship. A landscape data set of the historical process of fragmentation in the region was used to estimate patch size and time since isolation and the total forest cover in the landscape before fragmentation had taken place. We estimated a very high bird loss in forest patches, mainly concerning species sensitive to disturbances, endemics for the Atlantic Forest and with low relative abundance. The observed halflife, the time necessary to loose half of the species before the relaxation process, varied, on average, between 20 and 46 years. This period was shorter than those found in another study for birds in an African tropical forest (23 to 80 years), indicating that we have less time to revert this situation than previously predicted. The patch quality and size are both important factors determining the time lag for sensitive species. We have estimated that several species had already gone extinct, particularly in small and low quality patches, and that other ones will disappear in the future, even in the largest patches. Our results allowed us to speculate about other Atlantic Forest landscapes, where longer time since isolation and smaller patches can cause even higher losses. The synchronism of extinctions occurring in many regions probably will increase the list of threatened species in a near future, including many that today seem not likely to be vulnerable. Keywords: Landscape dynamic, conservation, extintion debt, theory of Island Biogeography, species-area relationship. Generally the effect of fragmentation on species focus on the landscape or regional scale, removing the geographic variation effect, nonetheless the addition of this variable might help us understand how species respond to fragmentation in a general form. Based on this idea, we compared the effects of forest fragmentation on bird species sensitive to habitat reduction and segmentation, functional groups and coincident species that occur in three landscapes in two geographically distant regions: the Pontal do Paranapanema region, in the extreme west of the State of São Paulo; and Ibiúna e Caucaia, 700 km to the east of the state. We undertook bird survey, using point count methodology in 37 forest fragment and 11 sites inside two continuous forests. The numbers of sample in each landscape were: Pontal – 21 patches and 7 sites in a control area; and Ibiúna and Caucaia – 8 patches in each landscape and 4 sites in a same control area. In the landscape of Pontal, bird species suffer the most intense effects of fragmentation, although patches are larger than in the other landscapes. The most sensitive species are: endemic, on the limit of distribution, out of abundance center and with low abundance. Many of them are affected only in the Pontal region, result of its geographical location. The spatial structure of landscape explains much of the species diversity and composition variation in the sampled areas. In Pontal and Ibiúna, patch size is the most important variable and, in Caucaia, the high degree of connectivity is the most relevant, making the species less vulnerable. Furthermore, because of its high connectivity, the patch size threshold that defines species loss, is much smaller in Caucaia than in Pontal (50 ha e 300-400 ha, respectively). The comparison of the effects of fragmentation on birds in distinct landscape shows that species responses varies along its distribution, which are constrained to local landscape parameters and bird features. This results show that there is no general recipes for biodiversity conservation, but different strategies must be used according to the regional condition. Keywords: connectivity, fragmentation, Landscape Ecology, conservation, geographical Variation.

Dans ce travail, nous nous sommes intéressés à l'étude asymptotique d'objets combinatoires aléatoires. Deux thèmes ont particulièrement retenu notre attention : les cartes planaires aléatoires et les modèles combinatoires liés à la théorie des fragmentations. La théorie mathématique des cartes planaires aléatoires est née à l'aube de notre millénaire avec les travaux pionniers de Benjamini & Schramm, Angel & Schramm et Chassaing & Schaeffer. Elle a ensuite beaucoup progressé, mais à l'heure où ces lignes sont écrites, de nombreux problèmes fondamentaux restent ouverts. Résumons en quelques mots clés nos principales contributions dans le domaine : l'introduction et l'étude du cactus brownien (avec J.F. Le Gall et G. Miermont), l'étude de la quadrangulation infinie uniforme vue de l'infini (avec L. Ménard et G. Miermont), ainsi que des travaux plus théoriques sur les graphes aléatoires stationnaires d'une part et les graphes empilables dans $\R^d$ d'autre part (avec I. Benjamini). La théorie des fragmentations est beaucoup plus ancienne et remonte à des travaux de Kolmogorov (1941) et de Filippov (1961). Elle est maintenant bien développée (voir par exemple l'excellent livre de J. Bertoin), et nous ne nous sommes pas focalisés sur cette théorie mais plutôt sur ses applications à des modèles combinatoires. Elle s'avère en effet très utile pour étudier différents modèles de triangulations récursives du disque (travail effectué avec J.F. Le Gall) et les recherches partielles dans les quadtrees (travail effectué avec A. Joseph). / The subject of this thesis is the asymptotic study of large random combinatorial objects. This is obviously very broad, and we focused particularly on two themes: random planar maps and their limits, and combinatorial models that are in a way linked to fragmentation theory. The mathematical theory of random planar maps is quite young and was triggered by works of Benjamini & Schramm, Angel & Schramm and Chassaing & Schaeffer. This fascinating field is still growing and fundamental problems remain unsolved. We present some new results in both the scaling limit and local limit theories by introducing and studying the Brownian Cactus (with J.F. Le Gall and G. Miermont), giving a new view point, a view from infinity, at the Uniform Infinite Planar Quadrangulation (UIPQ) and bringing more theoretical contributions on stationary random graphs and sphere packable graphs (with I. Benjamini). Fragmentation theory is much older and can be tracked back to Kolmogorov and Filippov. Our goal was not to give a new abstract contribution to this well-developed theory (see the beautiful book of J. Bertoin) but rather to apply it to random combinatorial objects. Indeed, fragmentation theory turned out to be useful in the study of the so-called random recursive triangulations of the disk (joint work with J.F. Le Gall) and partial match queries in random quadtrees (joint work with A. Joseph).

Nous nous intéressons à trois problèmes issus du monde des arbres aléatoires discrets et continus. Dans un premier lieu, nous faisons une étude générale des arbres de fragmentation auto-similaires, étendant certains résultats de Haas et Miermont en 2006, notamment en calculant leur dimension de Hausdorff sous des hypothèses malthusiennes. Nous nous intéressons ensuite à une suite particulière d’arbres discrets k-aires, construite de manière récursive avec un algorithme similaire à celui de Rémy de 1985. La taille de l’arbre obtenu à la n-ième étape est de l’ordre de n^(1/k), et après renormalisation, on trouve que la suite converge en probabilité vers un arbre de fragmentation. Nous étudions également des manières de plonger ces arbres les uns dans les autres quand k varie. Dans une dernière partie, nous démontrons la convergence locale en loi d’arbres de Galton-Watson multi-types critiques quand on les conditionne à avoir un grand nombre de sommets d’un certain type fixé. Nous appliquons ensuite ce résultat aux cartes planaires aléatoire pour obtenir la convergence locale en loi de grandes cartes de loi de Boltzmann critique vers une carte planaire infinie. / We study three problems related to discrete and continuous random trees. First, we do a general study of self-similar fragmentation trees, extending some results established by Haas and Miermont in 2006, in particular by computing the Hausdorff dimension of these trees under some Malthusian hypotheses. We then work on a particular sequence of k-ary growing trees, defined recursively with a similar method to Rémy’s algorithm from 1985. We show that the size of the tree obtained at the n-th step if of order n^(1/k), and, after renormalization, we prove that the sequence convergences to a fragmentation tree. We also study embeddings of the limiting trees as k varies. In the last chapter, we show the local convergence in distribution of critical multi-type Galton-Watson trees conditioned to have a large number of vertices of a fixed type. We then apply this result to the world of random planar maps, obtaining that large critical Boltzmann-distributed maps converge locally in distribution to an infinite planar map.

Face à l'uniformisation marchande du monde, aux épreuves de synthèses technologiques et industrielles, aux discrédits de la nature méthodiquement consentis, à l'asservissement machinique de notre sensibilité, l'homme apparaît, comme les artefacts qu'il a créés, semblable à une entité idéalisée et artificielle, inadaptée à son milieu. Comment ré-envisager alors nos modalités de créativité et de générativité au-delà du paradigme humaniste finissant, par-delà cette crise multiple de la pensée et de l'identité, de l'espace et de la représentation ? La conversion des principes métaphysiques et substantialistes nous oriente auprès de sentirs transindividuels, vers un devenir non humain de l'être-par-delà-le-sujet. Penser les valences entre intelligibilité et sensibilité, les médiances entre être technique et naturel, les transjections entre entendement et imagination suppose de reconsidérer la nature de l'être au-delà de ses forces de détermination, dans la puissance de sa vulnérabilité. Cet essai d'esthétique sur la créativité spatiale contemporaine s'attache à établir à l'aune de la révolution numérique une poïétique géosensible, transverse et trilogique, qui questionne conjointement les processus actuels de création artistique, de conception architecturale et de conceptualisation philosophique. Il s'accorde à penser l'acte créatif à partir du concept de dispositif géosensible, application spatiale du devenir-frémissant et de ses modalités que sont les fragmentations, les micro-hétérogénéisations et les fluences. / Faced with the mercantile globalisation of the world, the processes of technological and industrial syntheses, the discreditation of nature which has been methodologically consented to, and the mechanical subservience of our sensibilities, human beings appear, like the artefacts they have created, like idealised and artificial entities, ill-adapted to their surrounding world.How can we then re-envisage our creative and generative modalities beyond the humanist paradigm which is drawing to an end, outside this multiple crisis of identity and thought, and of space and representation. The conversion of metaphysical and substantialist principles orients us towards transindividual feelings for a post-human becoming of the being-outside-the-subject. Thinking the valencies between intelligibility and sensibility, the mediancies between technical and natural being, the transjections between comprehension and imagination, postulates reconsidering the nature of the being beyond its powers of determination, in the strength of its vulnerability.This essay on the aesthetics of contemporary spatial creativity seeks to establish, in light of the digital revolution, a geosensible poïetics, transversal and trilogical, which conjointly questions the current processes of artistic creation, architectural conception, and philosophical conceptualisation. It seeks to think the creative act from the point of view of the concept of the geosensible device, a spatial application of a becoming-trembling, and of its modalities which are fragmentations, micro-heterogenisations, and fluences.

Contexte : La chromodynamique quantique (CDQ), théorie de l’interaction forte, prédit un nouvel état de la matière, le plasma de quarks et de gluons (PQG) dont les degrés de liberté fondamentale, les quarks et les gluons, peuvent bouger quasi-librement. Les hautes températures et densités de particules, qui sont nécessaires, sont supposées être les conditions de l’univers dans ses premiers moments ou dans les étoiles à neutrons. Récemment elles ont été recrées par des collisions de noyaux d’ions lourdes à hautes énergies. Ces expériences étudient le PQG par la détection des particules de hautes énergies qui traversent le milieu, notamment, les quarks lourds. Les mécanismes de leur perte d’énergie dans le PQG ne sont pas compris complètement. Particulièrement, ils sont attribués aux processus soit de radiation induite par le milieu, soit de collisions de particules de type 2 vers 2, ou des combinaisons.Méthodes : Afin de trouver de nouvelles observables pour pouvoir distinguer les mécanismes de la perte d’énergie, on a implémenté un algorithme Monte-Carlo, qui simule la formation des cascades des particules à partir d’une particule initiale. Pour traiter le milieu, on a introduit des interactions PQG-jets, qui correspondent aux processus collisionnels et radiatifs. Les corrélations entre deux particules finales des cascades, dont une représente un quark trigger, ont été examinées comme moyen pour distinguer les modèles.Résultats : La dépendance de l’ouverture angulaire pour des corrélations entre deux particules en fonction des énergies des particules peut servir comme moyen pour séparer les mécanismes collisionnels et radiatifs de la perte d’énergie dans le milieu. / Context: Quantum chromodynamics (QCD), the theory of the strong interactions, predicts a new state of matter, the quark-gluon plasma (QGP), where its fundamental degrees of freedom, the quarks and gluons, behave quasi-freely. The required high temperatures and/orparticle densities can be expected for the early stages of the universe and in neutron stars, but have lately become accessible by highly energetic collisions of heavy ion cores. Commonly, these experiments study the QGP by the detection of hard probes, i.e. highly energetic particles, most notably heavy quarks, that pass the medium. The mechanisms of their energy-loss in the QGP are not yet completely understood. In particular, they are attributed to processes of either additional, medium induced radiation or 2 to 2 particle scattering, or combinations thereof.Methods: In a theoretical, phenomenological approach to search for new observables that allow discriminating between these collisional and radiative energy-loss mechanisms a Monte-Carlo algorithm that simulates the formation of particle cascades from an initial particle was implemented. For the medium, different types of QGP-jet interactions, corresponding to collisional and/orradiative energy loss, were introduced. Correlations between pairs of final cascade particles, where one represents a heavy trigger quark, were investigated as a means to differentiate between these models.Findings: The dependence of angular opening for two particle correlations as a function of particle energy may provide a means to disentangle collisional and radiative mechanisms of in-medium energy loss.

The increasing reliance on space infrastructure and its rapid expansion necessitate the development and enhancement of tools for space debris and fragmentation research. Accurate prediction of the risks associated with satellite fragmentation requires comprehensive understanding of the dynamics involved. To address this need, the widely used NASA Standard Breakup Model (SBM) is employed in this thesis to predict fragment characteristics resulting from breakup events. Additionally, a novel method is introduced to determine the direction of these fragments, something not directly covered by the SBM. Furthermore, the principle of kinetic gas theory is applied to calculate the overall, long-term collision risk between debris and a predetermined satellite population. The results from this reveal the limitations of the SBM in accurately simulating fragmentations for certain satellite types. However, the newly implemented fragment directionality method aligns well with observed data, suggesting its potential for further research. Similarly, the risk model exhibits strong correspondence with ESA's MASTER, a model used for assessing collision risks with debris, with the deviations likely due to different impact velocity models used. Finally, the validated fragmentation and risk models are combined, and the combined model is used to analyse a real-world fragmentation event. / Det ökande beroendet av rymdinfrastruktur, samt dess snabba expansion kräver utveckling och förbättring av verktyg för forskning och analys kring rymdskräp och fragmentering. För att förstå risken förknippad med satellitfragmentationer så krävs förståelse för den involverade dynamiken. För att tillgodose detta används NASA:s Standard Breakup Model (SBM) i denna avhandling för att bestämma fragmentegenskaper som bildas från olika sorters fragmentationshändelser. Dessutom introduceras en ny metod för att bestämma riktningen för dessa fragment, något som inte direkt täcks av SBM. Dessutom tillämpas principen för kinetisk gasteori för att beräkna den totala, långsiktiga kollisionsrisken mellan rymdskrot och en förutbestämd satellitpopulation. Resultaten från detta avslöjar SBM:s begränsningar när det gäller att simulera fragmenten för vissa satellittyper. Hursomhelst så kan man se att den nyligen implementerade fragmentriktningsmetoden stämmer väl överens med den observerade datan, vilket tyder på dess potential för ytterligare forskning. På samma sätt uppvisar riskmodellen överensstämmelse med ESA:s MASTER, en modell som används för att bedöma kollisionsrisker med rymdskrot, där avvikelser sannolikt beror på att olika kollisionshastighetmodeller används. Slutligen kombineras de validerade fragmenterings- och riskmodellerna, som sedan används för att bidra med analyser till en riktig fragmentationshändelse.