[Truncated abstract] During the last fifty years mathematical models of catchment hydrology have been widely developed and used for hydrologic forecasting, design and water resources management. Most of these models need large numbers of parameters to represent the flow generation process. The model parameters are estimated through calibration techniques and often lead to ‘unrealistic’ values due to structural error in the model formulations. This thesis presents a new strategy for developing catchment hydrology models for representing streamflow and salinity generation processes. The strategy seeks to ‘learn from data’ in order to specify a conceptual framework that is appropriate for the particular space and time scale under consideration. Initially, the conceptual framework is developed by considering large space and time scales. The space and time scales are then progressively reduced and conceptual model complexity systematically increased until ultimately, an adequate simulation of daily streamflow and salinity is achieved. This strategy leads to identification of a few key physically meaningful parameters, most of which can be estimated a priori and with minimal or no calibration. Initially, the annual streamflow data from ten experimental catchments (control and cleared for agriculture) were analysed. The streamflow increased in two phases: (i) immediately after clearing due to reduced evapotranspiration, and (ii) through an increase in stream zone saturated area. The annual evapotranspiration losses from native vegetation and pasture, the ‘excess’ water (resulting from reduced transpiration after land use change), runoff and deep storage were estimated by a simple water balance model. The model parameters are obtained a priori without calibration. The annual model was then elaborated by analysing the monthly rainfall-runoff, groundwater and soil moisture data from four experimental catchments. Ernies (control, fully forested) and Lemon (53% cleared) catchments are located in zone with a mean annual rainfall of 725 mm. Salmon (control, fully forested) and Wights (100% cleared) are located in zone with mean annual rainfall of 1125 mm. Groundwater levels rose and the stream zone saturated area increased significantly after clearing. From analysis of this data it was evident that at a monthly time step the conceptual model framework needed to include a systematic gain/loss to storage component in order to adequately describe the observed lags between peak monthly rainfall and runoff.

No description available.

The objective of this study was to analyze patterns, dynamics and processes of land-use/cover changes in the transboundary Mara River Basin in East Africa. We specifically focused on deforestation and expansion of agriculture in the watershed. The intensity analysis approach was used to analyze data from satellite imagery-derived land-use/cover maps. Results indicate that swap change accounted for more than 50% of the overall change, which shows a very dynamic landscape transformation. Transition from closed forest to open forest was found to be a dominant landscape change, as opposed to a random change. Similarly, transition from open forest to small-scale agriculture was also found to be a dominant transition. This suggests a trend (pathway) of deforestation from closed forest to small-scale agriculture, with open forest as a transitional land cover. The observed deforestation may be attributed to continuous encroachment and a series of excisions of the forest reserve. Transition from rangeland to mechanized agriculture was found to be a dominant land-use change, which was attributed to change in land tenure. These findings are crucial for designing strategies and integrated watershed management policies to arrest further deforestation in the forest reserves as well as to sustainably control expansion of agriculture.

Esse estudo analisou a dinâmica da produção de alimentos no estado do Rio Grande do Sul e suas consequências para as alterações nas áreas de pastagens naturais no Bioma Pampa. Para tanto, foram analisadas as mudanças no uso da terra nos municípios conforme os Censos Agropecuários de 1975, 1985, 1995/1996 e 2006 e da produção de alimentos de acordo com os relatórios de produção agrícola e pecuária municipais do Instituto Brasileiro de Geografia e Estatística (IBGE). Para analisar as variações na área das culturas e quantidade produzida foi calculado um índice de crescimento relativo. Os municípios foram ordenados conforme a área das categorias em seu território e sua localização por Escalonamento Multidimensional Não Métrico (NMDS). A influência do bioma sobre o uso da terra foi analisada por ENVIFIT e a diferença entre as categorias por PERMANOVA, no software R. No Bioma Pampa ocorreu um decréscimo de 26% nas pastagens naturais desde 1975, apresentando decréscimo de até 12.5%, entre 1975 e 1985. Destacam-se as taxas das lavouras e matas artificiais, em que para lavouras temporárias, apresentaram um crescimento considerável de 1985 para 2005. A influência do Bioma Pampa na composição das categorias de uso da terra manteve-se entre 14 e 15%, havendo diferença entre municípios localizados no Bioma Pampa em todos os anos analisados. Ao longo do tempo foi possível observar uma movimentação significativa das lavouras temporárias e das matas artificiais que principalmente entre 1995 e 2005 começam a integrar a paisagem do bioma. Para enfrentar o desafio de manter atividades agroecológicas em biomas ameaçados o Brasil deve investir em fiscalização, desenvolvendo sistemas de monitoramento capazes de detectar sutis alterações no uso da terra. Por outro lado, a produção de alimentos nesse estado apresentou crescimento na quantidade produzida. Foi observado um crescimento elevado na quantidade produzida de soja, sendo distribuição homogênea em todo o estado. O arroz apresentou redução na quantidade produzida no norte do estado e uma concentração expressiva nas regiões sul e fronteira oeste do estado. Os bovinos mantiveram seu rebanho estável com grande concentração na fronteira oeste. A silvicultura apresentou crescimento em praticamente todo estado, estando sua produção centrada na região sudeste do estado. As lavouras analisadas contribuíram para o PIB do estado e são capazes, hoje e no futuro, de suprir as demandas calóricas do Rio Grande do Sul em caso de necessidades, caso sejam mantidas as características atuais do agronegócio gaúcho. Dessas culturas, a soja foi a que mais disponibilizou calorias e retorno financeiro ao estado, sendo capaz de suprir a demanda local e oferecer excedentes para a exportação. Em todas as culturas, com exceção do milho, o aumento da quantidade de calorias disponibilizadas esteve relacionado ao aumento da área plantada, demandando maiores investimentos e incentivos ao incremento do rendimento das culturas. / This study analyzed the food production dynamics in the state of Rio Grande do Sul and its consequences for the changes in the natural grassland areas in the Pampa Biome. To this end, data for land use in the municipalities of Pampa Biome were collected from the Agricultural Censuses of 1975, 1985, 1995/1996 and 2006 as well as the food production data presented at the municipal agricultural and livestock reports published by the Brazilian Institute of Geography and Statistics (IBGE). To analyze the dynamics of the natural grasslands area of natural pastures, the micro-regions were compared every 10 years, the relative growth rate. The comparison between the municipalities was made by the adjusted mean the area allocated for the categories of land use. All municipalities were ranked as the area of the categories in its territory and its location by Multidimensional Scaling Not Metric (NMDS). The influence of the biome on land use was analyzed by ENVIFIT and the difference between the categories by PERMANOVA in Pampa Biome R. In software, there was a 26% decrease in natural pastures since 1975, presenting decrease of up to 12.5% between 1975 and 1985. Noteworthy are the rates of crops and artificial forest in which to temporary crops showed considerable growth from 1985 to 2005. The influence of the Pampa biome in the composition of land use categories remained between 14 and 15%, with significant differences between municipalities in the Pampa biome in all the years analyzed. Over time, it observed a significant movement of temporary crops and artificial forests that mainly between 1995 and 2005 begin to integrate the biome landscape. To face the challenge of maintaining agro-ecological activities threatened biomes in Brazil should invest in surveillance, developing monitoring systems capable of detecting subtle changes in land use. Moreover, the production of food in this state showed an increase in the amount produced. High growth for soy produced, being homogeneously distributed across the state was observed. Rice declined on the amount produced in the northern state and a significant concentration in the south and west of the state border. Cattle kept its stable herd with great concentration on the western border. Forestry grew in almost every state, with its production centered in the southeastern region of the state. The crops analyzed contributed to the state's GDP and are able, today and in the future, to meet the caloric demands of Rio Grande do Sul in the event purposes, if the current features of the gaucho agribusiness are maintained. These crops, soybean was the one that provided calories and financial return to the state, being able to meet local demand and provide surplus for exports. In all crops, except corn, increasing the amount of calories available was related to the increased planted area, requiring greater investments and incentives to increase crop yields.

Le concept de services écosystémiques est de plus en plus utilisé par différentes disciplines scientifiques et pris en compte dans les sphères politiques pour attirer l'attention sur les bénéfices que l'Homme reçoit des écosystèmes. Ce concept mène à étudier les liens complexes entre l'homme et son environnement. Cependant, la majorité des recherches actuelles restent théorique et peu de cas d'étude mettent à l'épreuve ce concept dans une démarche transdisciplinaire. Cette thèse à donc pour objectif principal de combler ce manque en explorant et analysant les dynamiques et processus des services écosystémiques en terme d'offre et de demande, y compris les effets de rétroactions, par une approche socio-écologique dans un contexte de changement planétaire. A cette fin, une étude transdisciplinaire a été conduite sur les prairies subalpines de la commune de Villar d'Arène (Hautes-Alpes) où l'élevage ovin et bovin domine. Des techniques de modélisations statistiques et de systèmes d'informations géographiques ont été combinées pour analyser la fourniture potentielle de services écosystémiques, conduisant à une cartographie de ceux-ci à l'échelle du paysage. La demande en services écosystémiques à quant à elle été étudiée par l'intermédiaire d'entretiens individuels et de groupes avec les acteurs locaux ainsi que des jeux de rôle avec les éleveurs de la commune. La co-construction de scénarios prospectifs avec les acteurs locaux ont permis d'étudier l'évolution de l'offre et la demande en services à un horizon 2030. / The ecosystem service (ES) concept is increasingly used in different scientific disciplines and is spreading into policy and business circles to draw attention to the benefits that people receive from biodiversity and ecosystems. Nevertheless, while the number of case studies considering various dimensions of the interactions between ecosystems and land use via ES has been steadily increasing, integrated research addressing interrelationships between biodiversity, ES and land use has remained mostly theoretical. This thesis aims through a socio-ecological approach to understand: (1) Which ES are potentially delivered given ecological dynamics, (2) how these ES are perceived by stakeholders in terms of value and knowledge, (3) how human management affects ES delivery, and (4) how ES are taken into account in land management decisions, thereby considering feedbacks from ecosystem to the land use system through ES. To address these questions, an interdisciplinary study was conducted on Villar d'Arène (French Alps) a municipality where the subalpine landscape is shaped by extensive mountain livestock farming. Statistical modelling and geographical information systems where combined to analyse the determinants of the spatial distribution of biodiversity and ES within the landscape using ecological (including plant functional traits), biophysical and land-use data. The following ES were mapped: agronomic value, aesthetic value, water quality, carbon storage, soil fertility, soil moisture, conservation of plant diversity and pollination. These allowed us to quantify trade-offs and synergies in the current landscape and to identify key management types supporting multifunctionality. The dynamics of ES was projected under four different scenarios integrating climatic, socio-economic and land-use changes, which were developed using a participative approach with regional experts and local farmers. Analyses of projected scenario impacts showed that ES synergies and trade-offs evolve differently when considering direct effects of climate on ecosystems, and/or their indirect effects through farmers adaptive responses. Interviews with local stakeholders (experts from nature conservation and agricultural extension, farmers and inhabitants) of mountain grasslands showed that the ES concept is still relatively unknown in explicit terms. Nevertheless after defining ES to interviewees, they expressed a variety of relevant interests and knowledge. Although all stakeholders valued a common set of ecosystem services (agronomic value, aesthetic value, water quality, and conservation of plant diversity), we identified negative and positive representations of the effects of grassland management on ecosystem services, depending on stakeholders perceptions of the relationships between soil fertility and biodiversity, and biodiversity and the other services. Finally, a role-playing game explored how ES cognition mediated environmental feedbacks on farmers' behaviours. Results emphasized the influence of other factors such as socio-economic or climatic context, topographic constraints, social value of farming or farmer individual and household characteristics, on the link between ES and land-management decisions. This case study demonstrates the interest of an integrated approach decomposing the feedback loop from ecosystems to land use when studying ES for scientific or policy purposes.

Fundação de Amparo a Pesquisa do Estado de Minas Gerais / The Araguari river hydrographic basin (BHRA) presents an area of 22,146.23 km2 and is located in a region of Minas Gerais State with high water demand by agriculture, industry and humans, which leads to high environmental costs, including damage of natural water resources systems. The frequent land use changes related to such activities often degrade water resources and therefore should be monitored in order to provide effective environmental management strategies, consistent with conservation activities of existing water streams. In this context, considering the need of efficient water resources management programs, the hydrological modeling becomes an important tool to help monitoring and management activities of water resources found in the watershed. Among numerous hydrological models, we can highlight (as used in this research) the ones having functions to overcome the demand for reliable and continuous data sets. In other words, these models can overcome situations where there is total or partial lack of data collected by hydrometeorological stations and even simulate situations/hypothetical scenarios for verification of some phenomena, for example, the behavior of stream flows. Thus, the objective of this research is to use a hydrological modeling to study the stream flow of BHRA, considering changes in land use and natural vegetation. The selected hydrological model was the Modelo de Grandes Bacias (MGB), developed by Instituto de Pesquisas Hidráulicas (IPH) from the Federal University of Rio Grande do Sul (UFRGS). The effects of land use and land cover (LULC) changes (time period of 1975-2010) in the stream flow regime were evaluated. The main LULC changes in the study area in this time period was the replacement of natural grasslands (Campo limpo, Campo sujo and Campo Cerrado) by croplands and cultivated pasturelands, as result of great increase of economic activities in this region. In 2010, the natural vegetation was reduced to nearly half of found in 1975 (88,93% and 42,97% of basin covered by natural vegetation in 1975 and 2010, respectively). Cultivated pasturelands increased from 8,41% in 1975 to 30,59% in 2010. Croplands increased from 2% in 1975 to 23,02% in 2010. The MGB was adjusted to simulate daily stream flows of the BHRA based on rainfall, climate and stream flow data gathered by eight rain gauged stations, 54 rainfall stations, besides the following spatial data: digital elevation models derived from the Shuttle Radar Topography Mission (SRTM); soil, geomorphology, geology and two-year LULC map (1975 and 2010). The calibration and simulation of MGB for BHRA showed good fit between observed (ANA) and calculated stream flow (MGB). The BHRA was then divided into sub-basins for further analysis. The Ibiá sub-basin showed better results of calibration and simulation of daily stream flows. The stream flows of Fazenda São Mateus, Desemboque and Fazenda Letreiro sub-basins were also analyzed. The Fazenda Letreiro sub-basin, the one that presented highest levels of LULC changes in the time period considered (conversion of 53,56% of natural vegetation into croplands and cultivated pasturelands) showed an increase of maximum (38,8 m3/s), minimum (2,29 m3/s) and average (5,78 m3/s) stream flow and for annual laminar flow (231,35 mm/year) as well. Hypothetical scenarios were simulated for the Ibiá sub-basin and compared with the stream flow results estimated by MGB for the year 2010. For actual LULC conditions, the measured and the estimated mean stream flows were quite similar. In a scenario of 100% of natural vegetation cover, the minimum stream flow increased 2,28 m3/s. In a scenario of 100% of cultivated pasture, the maximum stream flow presented highest increase of 36,18 m3/s. The use of the model and the analyses conducted in this study were considered satisfactory to meet the proposed research objectives. / A bacia hidrográfica do rio Araguari (BHRA) - MG, apresenta área de 22.146,23 km2 localizados em uma região de grande consumo de água pela agricultura, indústria e pelas pessoas, o que acarreta altos custos ambientais, incluindo danos aos sistemas hídricos naturais. As constantes alterações de uso da terra relacionadas a tais atividades são potencialmente degradantes dos recursos hídricos e devem ser conhecidas para que se possa planejar a gestão dos ambientes de forma coerente com a preservação dos cursos d água existentes. Neste contexto, frente a necessidade de gestão dos recursos hídricos, a modelagem hidrológica é uma ferramenta importante e eficiente no sentido de auxiliar o monitoramento e gestão das águas presentes nas bacias hidrográficas. Dentre as inúmeras aplicações dos modelos hidrológicos, podemos destacar (como utilizado nesta pesquisa) que eles têm a função de suprir a demanda de dados confiáveis e contínuos em locais onde existe total ou parcial ausência de dados coletados pelas estações hidrometeorológicas e ainda de simular situações/cenários hipotéticos para verificação de alguns fenômenos, a exemplo do comportamento da vazão dos rios presentes área de estudo. Desta forma, o objetivo geral desta pesquisa é realizar uma modelagem hidrológica para o estudo da vazão da BHRA, a partir das mudanças do uso da terra e cobertura vegetal nativa. O modelo hidrológico utilizado foi o Modelo de Grandes Bacias (MGB), do Instituto de Pesquisas Hidráulicas (IPH), da UFRGS. Foram avaliados os efeitos das mudanças de uso da terra e cobertura vegetal nativa (período de 1975 a 2010) na resposta da vazão na bacia. As principais mudanças de uso da terra e cobertura vegetal nativa na BHRA neste período foram a substituição da classe campestre (Campo limpo/Campo sujo/Campo cerrado) pelas classes de atividades agropecuárias, devido ao grande avanço destas atividades econômicas na região. Em 2010, a cobertura vegetal nativa se reduziu praticamente a metade dos valores calculados para o ano de 1975 (passou de 88,93% para 42,97%). A classe de Pastagem aumentou de 8,41% (1975) para 30,59% (2010) e a classe de Agricultura aumentou de 2% (1975) para 23,02% (2010). O MGB foi ajustado para a simulação de vazões diárias a partir dos dados de chuva, clima e vazão observadas de 8 estações fluviométricas, 54 estações pluviométricas, além de dados espaciais: Modelo Digital de Elevação (SRTM), mapa de solos, mapa de unidades geomorfológicas, mapa geológico e mapas de uso da terra e cobertura vegetal nativa (1975 e 2010) da BHRA. O processo de calibração do modelo e simulação para a BHRA permitiu verificar um bom ajuste entre as vazões observadas (ANA) e calculadas (MGB). A BHRA foi dividida em sub-bacias para melhor análise, sendo a sub-bacia Ibiá aquela que apresentou melhores resultados na calibração e simulação de vazões diárias. Além da sub-bacia Ibiá, foram analisadas as vazões das sub-bacias Fazenda São Mateus, Desemboque e Fazenda Letreiro pertencentes à BHRA. A sub-bacia Fazenda Letreiro, sub-bacia que teve mais mudanças de uso da terra e cobertura vegetal nativa no período considerado (com a substituição de 53.56% da cobertura vegetal nativa transformada em agricultura e pastagem) apresentou um aumento para as vazões máxima (em 38,8 m3/s), mínima (em 2,29 m3/s), média (em 5,78 m3/s) e lâmina escoada anual (231,35 mm/ano), no período de 1975 a 2010. Foram simulados cenários hipotéticos para a sub-bacia Ibiá e comparados os resultados com os dados de vazão calculados pelo MGB, do ano de 2010. Em relação aos dados de vazão do cenário real de 2010: a vazão média para os cenários hipotéticos propostos se manteve praticamente equiparada, o cenário 100% Cobertura vegetal nativa apresentou maior alteração em relação a vazão mínima (aumento de 2,28 m3/s) e o cenário de 100% Pastagem apresentou a maior alteração na vazão máxima (aumento de 36,18 m3/s). A utilização do modelo e as análises realizadas foram consideradas satisfatórias para cumprir os objetivos da pesquisa. / Doutor em Geografia

Esse estudo analisou a dinâmica da produção de alimentos no estado do Rio Grande do Sul e suas consequências para as alterações nas áreas de pastagens naturais no Bioma Pampa. Para tanto, foram analisadas as mudanças no uso da terra nos municípios conforme os Censos Agropecuários de 1975, 1985, 1995/1996 e 2006 e da produção de alimentos de acordo com os relatórios de produção agrícola e pecuária municipais do Instituto Brasileiro de Geografia e Estatística (IBGE). Para analisar as variações na área das culturas e quantidade produzida foi calculado um índice de crescimento relativo. Os municípios foram ordenados conforme a área das categorias em seu território e sua localização por Escalonamento Multidimensional Não Métrico (NMDS). A influência do bioma sobre o uso da terra foi analisada por ENVIFIT e a diferença entre as categorias por PERMANOVA, no software R. No Bioma Pampa ocorreu um decréscimo de 26% nas pastagens naturais desde 1975, apresentando decréscimo de até 12.5%, entre 1975 e 1985. Destacam-se as taxas das lavouras e matas artificiais, em que para lavouras temporárias, apresentaram um crescimento considerável de 1985 para 2005. A influência do Bioma Pampa na composição das categorias de uso da terra manteve-se entre 14 e 15%, havendo diferença entre municípios localizados no Bioma Pampa em todos os anos analisados. Ao longo do tempo foi possível observar uma movimentação significativa das lavouras temporárias e das matas artificiais que principalmente entre 1995 e 2005 começam a integrar a paisagem do bioma. Para enfrentar o desafio de manter atividades agroecológicas em biomas ameaçados o Brasil deve investir em fiscalização, desenvolvendo sistemas de monitoramento capazes de detectar sutis alterações no uso da terra. Por outro lado, a produção de alimentos nesse estado apresentou crescimento na quantidade produzida. Foi observado um crescimento elevado na quantidade produzida de soja, sendo distribuição homogênea em todo o estado. O arroz apresentou redução na quantidade produzida no norte do estado e uma concentração expressiva nas regiões sul e fronteira oeste do estado. Os bovinos mantiveram seu rebanho estável com grande concentração na fronteira oeste. A silvicultura apresentou crescimento em praticamente todo estado, estando sua produção centrada na região sudeste do estado. As lavouras analisadas contribuíram para o PIB do estado e são capazes, hoje e no futuro, de suprir as demandas calóricas do Rio Grande do Sul em caso de necessidades, caso sejam mantidas as características atuais do agronegócio gaúcho. Dessas culturas, a soja foi a que mais disponibilizou calorias e retorno financeiro ao estado, sendo capaz de suprir a demanda local e oferecer excedentes para a exportação. Em todas as culturas, com exceção do milho, o aumento da quantidade de calorias disponibilizadas esteve relacionado ao aumento da área plantada, demandando maiores investimentos e incentivos ao incremento do rendimento das culturas. / This study analyzed the food production dynamics in the state of Rio Grande do Sul and its consequences for the changes in the natural grassland areas in the Pampa Biome. To this end, data for land use in the municipalities of Pampa Biome were collected from the Agricultural Censuses of 1975, 1985, 1995/1996 and 2006 as well as the food production data presented at the municipal agricultural and livestock reports published by the Brazilian Institute of Geography and Statistics (IBGE). To analyze the dynamics of the natural grasslands area of natural pastures, the micro-regions were compared every 10 years, the relative growth rate. The comparison between the municipalities was made by the adjusted mean the area allocated for the categories of land use. All municipalities were ranked as the area of the categories in its territory and its location by Multidimensional Scaling Not Metric (NMDS). The influence of the biome on land use was analyzed by ENVIFIT and the difference between the categories by PERMANOVA in Pampa Biome R. In software, there was a 26% decrease in natural pastures since 1975, presenting decrease of up to 12.5% between 1975 and 1985. Noteworthy are the rates of crops and artificial forest in which to temporary crops showed considerable growth from 1985 to 2005. The influence of the Pampa biome in the composition of land use categories remained between 14 and 15%, with significant differences between municipalities in the Pampa biome in all the years analyzed. Over time, it observed a significant movement of temporary crops and artificial forests that mainly between 1995 and 2005 begin to integrate the biome landscape. To face the challenge of maintaining agro-ecological activities threatened biomes in Brazil should invest in surveillance, developing monitoring systems capable of detecting subtle changes in land use. Moreover, the production of food in this state showed an increase in the amount produced. High growth for soy produced, being homogeneously distributed across the state was observed. Rice declined on the amount produced in the northern state and a significant concentration in the south and west of the state border. Cattle kept its stable herd with great concentration on the western border. Forestry grew in almost every state, with its production centered in the southeastern region of the state. The crops analyzed contributed to the state's GDP and are able, today and in the future, to meet the caloric demands of Rio Grande do Sul in the event purposes, if the current features of the gaucho agribusiness are maintained. These crops, soybean was the one that provided calories and financial return to the state, being able to meet local demand and provide surplus for exports. In all crops, except corn, increasing the amount of calories available was related to the increased planted area, requiring greater investments and incentives to increase crop yields.

Cette thèse vise analyser les déterminants (objectifs et subjectifs) à l’origine de l'hétérogénéité des évaluations de projets. Dans un premier temps, j'analyserai comment les quantités (objectives) sont incorporées dans l'analyse coût-bénéfice. Le premier chapitre principal traite de la manière dont l'analyse coût-bénéfice est affectée par la distribution temporelle des impacts considérée. À travers l'exemple du changement d'affectation des terres provenant de la production de biocarburants, il ressort que les décisions concernant les projets ayant des impacts environnementaux non constants dans le temps s'appuient sur des valeurs actuelles nettes biaisées, ce qui peut entraîner la mise en œuvre de projets en réalité non souhaitables ou la non-implémentation de projets réellement souhaitables. Ce travail est à la fois théorique et numérique.Ensuite, j'étudie le rôle de la structure des préférences sur la prise de décision individuelle dans un cadre individuel puis un cadre collectif. Au niveau individuel, nous développons un modèle théorique qui permet soit de la substituabilité, soit de la complémentarité entre les biens en fonction du contexte (revenu et qualité de l'environnement). Il en résulte que l'élasticité-revenu du consentement à payer marginal peut être négative dans un contexte de substituabilité inter-biens, ce qui contraste avec les modèles habituels qui ne permettent que des élasticités positives du revenu (le bien environnemental ne peut jamais être inférieur mais est toujours normal). Notre cadre théorique affecte également la façon dont la consommation et la qualité de l'environnement sont actualisées, ce qui est d'autant plus pertinent dans un contexte de chocs sur les revenus. Au niveau collectif, j’utilise une approche expérimentale pour analyser l'effet de l'interaction entre les individus ayant des structures de préférences différentes sur les contributions au bien public. Il en résulte que la substituabilité parfaite est associée à un plus grand nombre de passagers clandestins que la complémentarité. Cependant, une aversion à l'inégalité avantageuse émerge également des individus dont les préférences sont à l'origine d'une parfaite substituabilité vis-à-vis de ceux dont les préférences sont fondées sur la complémentarité. / This thesis aims at providing insights on the (objective and subjective) determinants of the heterogeneity of project evaluations. At first, I analyze how (objective) quantities are incorporated in the cost-benefit analysis. The first main chapter deals with the way cost-benefit analysis is affected by the time distribution of impacts considered. Through the example of land use change from biofuel production, I find that decisions regarding projects with non-constant environmental impacts rely on distorted net present values, which may result in the implementation of actually non-desirable projects or the non-implementation of actually desirable projects. This work is both theoretical and numerical.Second, I investigate the role of the structure of preferences, i.e. whether private goods and environmental goods are substitutable or complementary in providing utility, on individual decision-making in an individual framework and a collective framework. At the individual level (second main chapter), we develop a theoretical model which allows either for substitutability or complementarity depending on the context (income and environmental quality). It results that the income elasticity of willingness to pay can be negative in contexts of between-goods substitutability, which contrasts with usual frameworks which only allow for positive income elasticities (thus the environmental good can never be inferior but is always normal). Our framework also affects the way consumption and environmental quality are discounted, which is all the more relevant in the context of income shocks. At the collective level (third main chapter), we use an experimental approach to analyze the effect of the interaction between individuals with different structures of preferences on contributions to the public good. It results that perfect substitutability is associated with more free-riding than complementarity. However, an aversion to advantageous inequality also emerges from individuals whose preferences underlie perfect substitutability towards those whose preferences are based on complementarity.

A agricultura de corte-e-queima é praticada há milhares de anos nas áreas florestadas do planeta, principalmente nas regiões tropicais. Muitos estudos atestam a sustentabilidade desses sistemas quando praticados tradicionalmente e sob baixas densidades populacionais, mantendo, ou mesmo, promovendo a biodiversidade local e garantindo a subsistência de milhões de pessoas pobres rurais. No entanto, é crescente na literatura acadêmica e no debate político o papel que a agricultura de corte-e-queima vem desempenhando no desmatamento e demais impactos ambientais e sócio-econômicos. Esse processo é conseqüência das mudanças no uso do solo, intensificação agrícola e aumento demográfico que estão alterando as práticas e comprometendo a sustentabilidade desses sistemas agrícolas tradicionais. No Vale do Ribeira, sudeste do Estado de São Paulo, comunidades quilombolas têm sofrido um processo acelerado de mudanças nos seus padrões de subsistência. Dessa forma, esse estudo traz uma revisão da literatura sobre a agricultura de corte-equeima com o objetivo de traçar um panorama geral do que foi produzido até o momento, identificar as principais correntes teóricas envolvidas e apontar as alternativas propostas para sua manutenção. Em seguida, são caracterizados o perfil demográfico e sócio-econômico de nove comunidades quilombolas do Vale do Ribeira e identificados os principais fatores responsáveis pelas mudanças recentes nos seus padrões de subsistência. Por fim, é feita uma análise etnográfica, através de informações levantadas em três comunidades quilombolas pré-selecionadas, para a caracterização das atividades agrícolas praticadas atualmente e no passado recente, bem como o processo de mudanças ocorridas e os impactos causados no sistema agrícola de corte-equeima. Os resultados mostram que as restrições impostas pela legislação ambiental, os conflitos de terra, a construção de uma rodovia na região, a crescente inserção no mercado regional, e a atuação de órgãos governamentais e não-governamentais de desenvolvimento são os principais fatores responsáveis pelas mudanças observadas no sistema agrícola de corte-e-queima e, conseqüentemente, na organização socioeconômica destas populações. / Slash-and-burn agriculture has been practiced for thousands of years in the forests around the world, especially in the tropics, where it provides for the livelihood of countless poor rural populations. Many studies claim that slash-andburn agriculture is sustainable when performed under conditions of low human demographic density, maintaining or even increasing local biodiversity. However, it is growing in the academic literature, as well as in development debates, the concern regarding the role that this system has been playing in the deforestation of the planet´s tropical forests. This process appears to be closely linked to changes in land use patterns (agricultural intensification) and urban and rural demographic growth. In Ribeira Valley, State of São Paulo, Quilombola communities have had a significant increase of changes in their subsistence system. On the thread of these concerns, this study presents a critical review of the international and national academic literature on slash-and-burn agriculture. Thus, this review intends to draw a broad scenario of the current academic debate on this issue, as well as to identify the main alternatives strategies proposed to maintain or replace this cultivation system. Following this study aims to characterize the socioeconomic and demographic profile of nine Quilombola populations in the Ribeira Valley, and to identify the main factors responsible for the recent changes in their subsistence system. Por fim, an ethnographic analysis is done, through informations gathered in three pre-selected Quilombola communities, to characterize the agricultural activities practiced nowadays and in the recent past, besides the change process and impacts on the slash-and-burn agriculture. The results show that restrictions imposed by environmental laws, conflict over land, the construction of a major road in the region, the growing insertion into a market economy, and the intervention of governmental and nongovernmental development agencies are the main factors behind the changes observed in the subsistence system and, consequently, in the socioeconomic organization of these populations.

A pecuária é considerada uma das principais fontes de emissão de gases de efeito estufa (GEE) no Brasil. Sua participação no inventario nacional de emissões de GEE está relacionada tanto à conversão da vegetação nativa em pastagens, com a perda de biomassa vegetal e modificações nas propriedades físicas e químicas do solo, quanto à participação dos próprios animais, através da eructação e da deposição de dejetos. A quantificação das emissões de GEE em sistemas agropecuários possibilita avaliar o grau de impacto dessa atividade sobre o ambiente. Grande parte dos estudos realizados para a quantificação das emissões dos dejetos do gado foi desenvolvido em regiões de clima temperado, porém faltam informações para as condições tropicais. No Brasil os fatores de emissão obtidos são menores que o valor default de 2% proposto pelo IPCC. Em vista ao grau de incerteza associado ao valor default para os dejetos animais, confirma-se a necessidade da determinação de fatores de emissão específicos, com o objetivo de conferir maior precisão aos inventários nacionais. Os objetivos deste estudo foram (i) avaliar os sistemas vegetação nativa e pastagem quanto aos fluxos de GEE provenientes das respectivas fontes: solo, fezes e urina do gado, com a finalidade de verificar suas contribuições específicas no total de GEE emitidos; (ii) determinar os fatores de emissão dos dejetos animais para a região edafoclimática da região sudeste do Brasil. Esse estudo foi realizado durante as estações seca e chuvosa, para avaliar também o efeito da sazonalidade na emissão de GEE. Foi utilizado câmaras estáticas fixadas ao solo para quantificação dos fluxos de CO2, CH4 e N2O, por um período de trinta dias em cada estação. Os fluxos diários de emissão de GEE provenientes dos dejetos apresentaram pico de emissão logo após sua aplicação, que perduraram apenas durante os primeiros dez dias amostrados. Os fatores de emissão do N2O calculados neste estudo também foram inferiores ao default, de 0,05% para a urina e 0,001% para as fezes na estação seca, e de 0,4% e 0,004% na estação chuvosa, respectivamente. O fator de emissão do CH4 calculado para as fezes do gado foi de 0,012 kg CH4 cabeça-1 ano-1 na estação seca e 0,004 kg CH4 cabeça-1 ano-1 na estação chuvosa. Os fluxos acumulados de CH4 e N2O gerados nesse estudo foram convertidos em CO2 equivalente, para efeito de comparação, para a contabilização da contribuição total de cada fonte na emissão de GEE. O solo sob vegetação nativa emitiu um total de 274 kg CO2e ha-1 ano-1, enquanto que na pastagem esse valor foi de 657 kg CO2e ha-1 ano-1, sem contabilizar os dejetos. Os dejetos contribuíram com 9.853 kg CO2e ha-1 ano-1, e quando somado um valor default para a eructação do gado, esse valor aumentou para 27.878 kg CO2e ha-1 ano-1. Os resultados demonstram que a pastagem emite uma quantidade 2,5 maior de GEE para a atmosfera quando comparado com uma área de vegetação nativa. Além disso, verificou-se a influência da sazonalidade na emissão dos GEE e a importante contribuição dos dejetos no total das emissões contabilizadas para o sistema pastagem / Livestock is considered one of the main sources of greenhouse gases (GHGs) emission in Brazil. Its contribution is related either to conversion of native vegetation in pasture areas, with changes in physical and chemical soil properties, consequently changing the GHG fluxes into the atmosphere, or by the decomposition of livestock manure. Most studies conducted to quantify emissions from livestock manure have been developed in temperate regions with still a lack of information for tropical conditions. In Brazil, emission factors obtained were lower than the default value of 2% proposed by IPCC. Due to the uncertainty degree associated with the default value for animal manure, confirms the need for determination of specific emission factors, in order to give greater precision to national inventories. The objectives of this study were (i) to evaluate the systems native vegetation and pasture as the GHG fluxes from the respective sources: soil, faeces and cattle urine, in order to verify their specific contributions in the total GHG emissions; (ii) and determine the emission factors of animal manure. This study was conducted during the dry and rainy seasons, to evaluate the effect of seasonality in GHG emissions. Static chambers fixed to the ground were used to quantify CO2, CH4 and N2O fluxes for a period of thirty days in each season, with five replicates for each treatment. On each day of collection, sampling was performed at regular intervals (0, 10 and 20 minutes after chamber closure). The GHG emission daily flows from manure showed a peak of emission immediately after application, which lasted only during the first ten days sampled. The N2O emission factors calculated in this study were lower than the default, 0.05% for urine and 0.001% for faeces in the dry season, and 0.4% and 0.004% in the rainy season, respectively. The CH4 emission calculated factors for the cattle faeces were 0.012 kg CH4 head-1 yr-1 in the dry season and 0.004 kg CH4 head-1 yr-1 in the rainy season. The cumulative flows of CH4 and N2O generated in this study were converted into CO2 equivalent, for comparison, accounting for the total contribution of each source of GHG emissions. The soil under native vegetation issued a total of 274 kg CO2e ha-1 yr-1, while in the pasture this value was 657 kg CO2e ha-1 yr-1, not counting the manure. The manure contributed 9853 kg CO2e ha-1 yr-1, and when coupled with a default value for cattle belching, this value increased to 27,878 kg CO2e ha-1 yr-1. The results demonstrate that the pasture emits 2.5-fold greater quantity of GHG when compared to a native vegetation area. In addition, there was the influence of seasonality on GHG emissions and the important contribution of waste in total emissions accounted for pasture system