Economic changes such as rising açaí prices and the availability of off-farm employment are transforming the landscape of the Amazonian várzea, subject to decision-making at the farming household level. Land use change results from complex human-environment interactions which can be addressed by an agent-based model. An agent-based model is a simulation model composed of autonomous interacting entities known as agents, built from the bottom-up. Coupled with cellular automata, which forms the agents’ environment, agent-based models are becoming an important tool of land use science, complementing traditional methods of induction and deduction. The decision-making methods employed by agent-based models in recent years have included optimization, imitation, heuristics, classifier systems and genetic algorithms, among others, but multiple methods have rarely been comparatively analyzed. A modular agent-based model is designed to allow the researcher to substitute alternative decision-making methods. For a smallholder farming community in Marajó Island near Ponta de Pedras, Pará, Brazil, 21 households are simulated over a 40-year period. In three major scenarios of increasing complexity, these households first face an environment where goods sell at a constant price throughout the simulated period and there are no outside employment opportunities. This is followed by a scenario of variable prices based on empirical data. The third scenario combines variable prices with limited employment opportunities, creating multi-sited households as members emigrate. In each scenario, populations of optimizing agents and heuristic agents are analyzed in parallel. While optimizing agents allocate land cells to maximize revenue using linear programming, fast and frugal heuristic agents use decision trees to quickly pare down feasible solutions and probabilistically select between alternatives weighted by expected revenue. Using distributed computing, the model is run through several parameter sweeps and results are recorded to a cenral database. Land use trajectories and sensitivity analyses highlight the relative biases of each decision-making method and illustrate cases where alternative methods lead to significantly divergent outcomes. A hybrid approach is recommended, employing alternative decision-making methods in parallel to illustrate inefficiencies exogenous and endogenous to the decision-maker, or allowing agents to select among multiple methods to mitigate bias and best represent their real-world analogues.

Economic changes such as rising açaí prices and the availability of off-farm employment are transforming the landscape of the Amazonian várzea, subject to decision-making at the farming household level. Land use change results from complex human-environment interactions which can be addressed by an agent-based model. An agent-based model is a simulation model composed of autonomous interacting entities known as agents, built from the bottom-up. Coupled with cellular automata, which forms the agents’ environment, agent-based models are becoming an important tool of land use science, complementing traditional methods of induction and deduction. The decision-making methods employed by agent-based models in recent years have included optimization, imitation, heuristics, classifier systems and genetic algorithms, among others, but multiple methods have rarely been comparatively analyzed. A modular agent-based model is designed to allow the researcher to substitute alternative decision-making methods. For a smallholder farming community in Marajó Island near Ponta de Pedras, Pará, Brazil, 21 households are simulated over a 40-year period. In three major scenarios of increasing complexity, these households first face an environment where goods sell at a constant price throughout the simulated period and there are no outside employment opportunities. This is followed by a scenario of variable prices based on empirical data. The third scenario combines variable prices with limited employment opportunities, creating multi-sited households as members emigrate. In each scenario, populations of optimizing agents and heuristic agents are analyzed in parallel. While optimizing agents allocate land cells to maximize revenue using linear programming, fast and frugal heuristic agents use decision trees to quickly pare down feasible solutions and probabilistically select between alternatives weighted by expected revenue. Using distributed computing, the model is run through several parameter sweeps and results are recorded to a cenral database. Land use trajectories and sensitivity analyses highlight the relative biases of each decision-making method and illustrate cases where alternative methods lead to significantly divergent outcomes. A hybrid approach is recommended, employing alternative decision-making methods in parallel to illustrate inefficiencies exogenous and endogenous to the decision-maker, or allowing agents to select among multiple methods to mitigate bias and best represent their real-world analogues.

As ações humanas têm sido reconhecidas como as principais forças geradoras de impactos ambientais. Dentre essas ações ressaltam-se as mudanças de uso e cobertura da terra, as quais têm contribuído significativamente para a degradação e/ou conservação de ecossistemas e recursos naturais. Desse modo, devido à necessidade de assegurar um ambiente ecologicamente equilibrado para a população humana, a gestão ambiental pública (GAP) vem se desenvolvendo cada vez mais no Brasil. Esta recorre frequentemente aos conhecimentos técnicos e científicos para construir formas de gerir o ambiente. Entre estes, destacam-se os relacionados à Ecologia da Paisagem, com diversos trabalhos realizados na região onde se realizou a presente pesquisa (Bacia do Rio Corumbataí, SP). No entanto, ao basear-se somente nesses conhecimentos a GAP pode estar desconsiderando outra importante forma de gestão ambiental (GA), a gestão ambiental local (GAL). GAL é a gestão ambiental praticada por grupos familiares de pequenos produtores rurais e populações locais, em suas práticas cotidianas. A GAL está relacionada ao uso de recursos naturais em busca da sobrevivência (manutenção e reprodução física e cultural), incluindo maior ou menor inserção no mercado, e seu impacto ambiental. A desconsideração da GAL pela GAP pode comprometer a efetividade desta última e acarretar conflitos socioambientais. Sendo assim, visando gerar subsídios para adequação entre GAP e GAL, em Ipeúna, SP, esta pesquisa teve os seguintes objetivos: caracterizar a GAL realizada por proprietários rurais de Ipeúna, SP; comparar as decisões, principalmente sobre o uso da terra (e recursos associados), desse nível de gestão (GAL) com as formas de GAP mais presentes no cotidiano dos proprietários rurais; verificar se alguma dessas formas de GAP está fundamentada em princípios da Ecologia da Paisagem; e comparar a visão dos pequenos produtores rurais e da Ecologia da Paisagem sobre o processo de fragmentação florestal no município de Ipeúna, SP. A GAL praticada pelos proprietários rurais de Ipeúna, SP, foi analisada sob uma perspectiva histórica. Observou-se que aquela se modificou em função do processo de industrialização da agricultura no Brasil. As decisões sobre o uso da terra tomadas pelos pequenos produtores rurais integrantes dessa pesquisa foram guiadas principalmente por aspectos econômicos. As formas de GAP mais presentes no cotidiano daqueles produtores foram do tipo comando e controle, elaboradas para serem aplicadas em grandes territórios como o Estado de São Paulo e no País todo. O conhecimento científico teve importante papel em pelo menos alguma etapa da elaboração, construção e aplicação dessas formas de GAP. Nenhuma delas foi elaborada considerando o contexto e a GAL praticada na área estudada. Com relação à Ecologia da Paisagem, nenhuma forma de GAP mencionada pelos produtores rurais que fizeram parte desta pesquisa foi elaborada com base em suas recomendações. A visão dos produtores integrantes desta pesquisa e da Ecologia da Paisagem com relação ao processo de fragmentação florestal apresentou complementaridades. Estas complementaridades podem ser úteis para a construção de formas de GA mais condizentes com a realidade local e, portanto, mais eficazes para a conservação ambiental e equidade social. / Human actions have been recognized as the main driving forces of environmental impacts. Among these actions, the land use/cover changes have been contributed significantly to the degradation and/or conservation of ecosystems and natural resources. Due to the need of assuring an ecologically balanced environment to the human population, the public environmental management (PEM) has been developing increasingly in Brazil. The PEM has often used technical and scientific knowledge to build forms of environmental management. Among these technical and scientific knowledge, we highlight those related to Landscape Ecology, with several researches performed in the region where the present study took place (Corumbataí River Basin, SP). However, by relying only on this kind of knowledge the PEM could take apart another important form of environmental management (EM), the local environmental management (LEM). The LEM is the environmental management practiced daily by small farmers families and local people. The LEM concerns to the use of natural resources in order to achieve their social and cultural reproduction in a specific place. The LEM includes more or less market insertion and its impacts on the environment. When the PEM do not consider the LEM, the effectiveness of PEM can decrease and cause social and environmental conflicts. Therefore, in order to generate subsidies for coordinating LEM and PEM in Ipeúna, SP, this study had the following aims: to characterize the LEM practiced by farmers in Ipeúna, SP; to compare the decisions, mainly about the land use (and associated resources), taken in this management level (LEM) to the more common types of PEM in the small farmers routines; to verify if the PEM mentioned by the small farmers is based on principles of Landscape Ecology, and to compare their and the Landscape Ecology views about the process of forest fragmentation in Ipeúna, SP. The LEM practiced by farmers in Ipeúna, SP, was analyzed from a historical perspective. It was observed that the LEM has changed due to the process of agriculture industrialization in Brazil. Mainly economics aspects guided the small farmers land use decisions. The more frequent PEM forms in the daily life of small farmers were command and control type. This type of PEM was designed to large areas such as São Paulo and all over Brazil. Scientific knowledge played an important role in some parts of PEM elaboration, construction and application forms. None of the PEM identified by the small farmers have been developed considering the context and the LEM practiced in the area. About the Landscape Ecology, none of PEM forms mentioned by small farmers was developed based on its prescriptions. The view of the small farmers who integrated this research and of the Landscape Ecology related to the process of forest fragmentation showed complementarities. Such complementarities could be useful to the building of more consistent to the local reality and so, more effective to the environmental conservation and social equality.

Abstract Rivers create unique habitat for aquatic life and provide ecosystem services for humans. Thus, degradation of river water quality is a serious, global problem. Water quality is the outcome of anthropogenic and natural landscape factors and the interaction of these two. To improve water quality, robust and quick methods are needed to study the complex, spatio-temporally dependent relation between water quality and environment conditions across extensive areas. This thesis aimed to study the relationship between water quality (total phosphorus and nitrogen, pH, water colour and dissolved oxygen) and environmental factors in boreal rivers combining grid-based data and statistical methods. The study comprised of 34 Finnish rivers with their catchments. First, the effect of natural and human-induced environmental factors on water quality was studied. Then, (a) the ability of the characteristics of different spatial scales around the river channel and under different discharge conditions to predict water quality was explored and (b) the suitability of the applied statistical methods (generalized linear and additive models, partitioning methods, non-metric multidimensional scaling) in water quality studies was evaluated. As expected, the results highlighted the impact of agricultural activities on water quality as nutrients and pH increased, together with the cover of agricultural activities. However, when studied as a group, natural factors explained water quality better than land use/cover. Lakes were strongly related to decreased nutrients and water colour. The effect of fine-grained soils on nutrients and pH was positive. In the scale studies, nutrients and water colour were best explained by the characteristics of the entire catchment but pH was mostly predicted by the characteristics of the 50 m riparian zone. The connection between water quality and environment was strongest during high-flow discharge periods. The results encourage the use of the applied methods, showing that the combination of grid-based data and advanced statistical methods provide an efficient first-filter estimate of water quality-environment relations. Spatial-based statistical modelling provides a crucial framework for river, water resources and land use management. The applied methods can also be seen as essential tools when predicting the impacts of global change on water quality. / Tiivistelmä Joet luovat vesielämälle ainutlaatuisen elinympäristön ja tarjoavat ihmisille ekosysteemipalveluja. Jokien vedenlaadun huononeminen onkin vakava, maailmanlaajuinen uhka. Vedenlaatua määrittävät luonnolliset ja ihmisen muokkaamat ympäristötekijät sekä näiden yhteisvaikutus. Vedenlaadun kohentamiseksi tarvitaan luotettavia ja nopeita menetelmiä, joiden avulla voidaan tutkia monimutkaista, alueellisesti ja ajallisesti riippuvaa yhteyttä vedenlaadun ja ympäristöolojen välillä laajoilla alueilla. Tässä väitöskirjassa oli tavoitteena tutkia vedenlaadun (kokonaisfosfori ja -typpi, pH, väriluku ja liukoinen happi) ja ympäristön yhteyttä boreaalisella vyöhykkeellä käyttäen paikkatietoaineistoja ja tilastollisia menetelmiä. Tutkimusalueena oli 34 suomalaista jokea valuma-alueineen. Ensinnäkin tutkittiin luonnollisten ympäristötekijöiden ja ihmistoiminnan vaikutusta vedenlaatuun. Tavoitteena oli myös selvittää, miten joen ympärillä olevien erikokoisten vyöhykkeiden ominaisuudet ja vaihtelevat virtaamaolosuhteet selittävät vedenlaatua. Lopuksi arvioitiin käytettyjen tilastollisten menetelmien (yleistetyt lineaariset ja additiiviset mallit, hajonnan ositusmenetelmät, ordinaatioanalyysi) soveltuvuutta vedenlaatututkimuksissa. Tuloksissa korostui odotetusti valuma-alueen maatalouden vaikutus vedenlaatuun. Ravinteiden määrä ja pH-luku kasvoivat maatalouden lisääntyessä. Muuttujien ryhmittäisessä tarkastelussa ympäristön luonnolliset ominaisuudet selittivät vedenlaatua maankäyttöä/-peittoa paremmin. Järvisyyden lisääntyminen oli yhteydessä ravinteisuuden ja väriluvun laskuun. Hienorakeiset maalajit olivat yhteydessä ravinteisuuden ja pH-luvun nousuun. Mittakaavatarkastelussa ravinteisuutta ja värilukua ennustivat parhaiten koko valuma-alueen ominaisuudet, mutta pH-lukua selittivät parhaiten ominaisuudet 50 m:n vyöhykkeellä joen ympärillä. Ympäristön ja vedenlaadun yhteys oli voimakkaimmillaan, kun jokien virtaamat olivat korkeimmillaan. Tämä väitöskirja osoittaa, että paikkatietoaineistojen ja sovellettujen tilastollisten menetelmien yhteiskäyttö tuottaa tehokkaita malleja vedenlaadun ja ympäristön välisestä yhteydestä. Spatiaalis-tilastollinen mallinnus tarjoaa tärkeän viitekehityksen jokien ja vesistöjen käytön sekä maankäytön suunnitteluun. Lähestymistapa voidaan nähdä myös tärkeänä välineenä ennustettaessa globaalimuutoksen vaikutusta vedenlaatuun.

A expansão da cana-de-açúcar tem ocorrido de maneira intensa no cenário brasileiro, especialmente no estado de São Paulo e região Centro-Oeste do Brasil, com potencial de causar significativos impactos destacando-se a degradação do solo, a poluição dos recursos hídricos, a pressão sobre outras culturas ou matas nativas, conflitos pelo uso do solo, entre outros. O objetivo dessa pesquisa é avaliar a aplicação de técnicas de modelagem do uso do solo em conjunto com instrumentos de política ambiental por meio da modelagem da expansão da cana-de-açúcar na bacia do Tietê/Jacaré (UGRHI 13). A bacia hidrográfica em questão caracteriza-se pela presença de regiões de alta vulnerabilidade ambiental, em termos da fragilidade do território e presença de atributos ambientais de significativo interesse, bem como pelo fato de suas principais atividades econômicas estarem vinculadas ao agronegócio. Para a consecução dos objetivos estabelecidos, o modelo GEOMOD presente na plataforma IDRISI foi aplicado para a projeção da expansão da cana-de-açúcar em 2019. De modo a verificar a aplicabilidade das informações geradas, a projeção gerada foi cotejada a dois instrumentos de planejamento regional: o Zoneamento Agroambiental do Setor Sucroalcooleiro e o Plano de Bacia da UGRHI 13. Constatou-se que a aplicação de modelos de uso do solo para a projeção de alterações nos padrões de ocupação do território apresenta-se como prática promissora que pode ser integrada ao planejamento e gestão ambiental a nível regional, sendo, entretanto, necessário melhorar a performance do modelo elaborado a fim de gerar informações mais confiáveis para os tomadores de decisão. / Sugarcane expansion has been significant in Brazil, especially in São Paulo State and Midwest Region. This crop has potential of causing significant impacts, such as: soil degradation, pollution of water resources, pressure on other crops or native forests, conflicts regarding land use, among others. The purpose of this study is to verify the application of land use cover change models with environmental policy instruments by modeling sugarcane expansion in Tietê/Jacaré watershed (UGRHI 13). This water basin is characterized by high vulnerability environmental areas in terms of fragility territorial and presence of significant environmental attributes, as well as the fact that its main economic activities are linked to agribusiness. GEOMOD land use cover change model was applied to project sugarcane expansion in 2019. In order to verify the applicability of sugarcane projection, it was used in the context of the Sugarcane Agro-Industry Zoning and the Water Basin Plan. The results showed that land use cover change models application for sugarcane projection is presented as a promising practice that can be integrated to the regional environmental planning. However, it´s necessary to improve the performance of the model in order to generate more reliable information to the decision makers.

>Magister Scientiae - MSc

>Magister Scientiae - MSc / The Kuils River Catchment is an urban river catchment that forms part of the larger Kuils-Eerste River system draining the eastern half of the Cape Metropolitan Authority area and Stellenbosch Municipality. Rapid urbanisation has resulted in the encroachment of residential and industrial areas into the river system through channelization and sewage disposal. This research project intends to assess the quality of surface runoff in the Kuils River catchment and determining non-point source pollutant loading rates in the catchment using GIS-based modelling. The study results show how modelled potential sources of surface runoff and NPS pollutants using desktop GIS analysis tools in a sequential process that involved different levels of software applications could explain the characteristics of the catchment. With the help of the Expected Mean Concentration (EMC) values associated with surface runoff from land use/covers, NPS pollutant loads were assessed downstream towards the Kuils River Catchment outlet using the Nonpoint Source Pollution and Erosion Comparison Tool (N-SPECT) based in ArcGIS. The outputs from this model consist of predicted annual pollutant loading (mg/mvyear) for each Kuils-Eerste River that occurs in the catchment. The results have shown clearly the spatial distribution of sources of particular pollutants in the catchment. Further or advanced processing knowhow with this model might provide far reaching insights into the problem and it is however recommended that these results produced using N-SPECT be compared to those of other hydrologic models using the same inputs.