Thesis (PhD (Conservation Ecology and Entomology))--Stellenbosch University, 2008. / Dispersal plays an essential role in determining the distribution of populations of species, especially species expanding their ranges. Two disciplines are concerned with gaining understanding of spread of species, namely restoration ecology and invasion biology. Con- ceptual understanding of dispersal, its mechanisms and its management is essential to both disciplines. Nevertheless, the disciplines have quite opposite objectives: in restoration ecol- ogy, spread of indigenous species into transformed landscapes is promoted, while invasion biology aims to prevent the (further) spread of alien species into pristine or restored habi- tats. Despite these two opposite objectives of facilitating spread and preventing spread of their respective target species, these disciplines have essentially the same requirements in terms of information needed for restoration. In this thesis, I will present two modelling studies—one looking at the impact of two different seed-feeding alien control agents on the spread of Hakea sericea, the other investigating the recolonisation by Dicerothamnus rhinocerotis of an old field dominated by Cynodon dactylon. Based on these studies, I will draw conclusions for the management in each case. In a second step, I will compare these two seemingly-different studies and draw conclusions on how these two disciplines can learn from each other, and how conclusions drawn and management recommendations developed for the one discipline can be translated for the other. The invasion biology study concluded that seed-feeding biocontrol agents do have a considerable impact on the velocity of the spread of the target species. In addition, management recommendations included the possibility of substituting seed-feeding biocontrol agents with an increased fire frequency where the negative impact on natural vegetation, on the site invaded by the target species, is acceptable. The restoration study concluded that the main impact on the velocity of spread, and the speed of the return of the shrub species onto the old fields, is the availability of micro-sites. A sensitivity analysis showed the even a slight change from 1% to 2% increases the velocity and pattern of spread dramatically. The other parameters playing an important role are the mean rate of establishment and the time span between

The Tonle Sap is the largest and most important natural wetland in Southeast Asia. It covers an area of more than 15,000 km2 with a unique mosaic of natural and agricultural floodplain habitats that coexist with the largest fishery in the Mekong Basin. Accelerating hydropower development and climate change, however, are altering the Mekong’s hydrology, which could negatively affect downstream ecosystems. The Tonle Sap is facing a two-fold problem. First, the link between its hydrology and ecosystem properties is not well understood. Second, potential ecological changes caused by future hydrological disruptions related to hydropower and climate change are unknown. Thus, the main objective of this thesis was to quantify how alterations to the Mekong hydrology could affect the Tonle Sap ecosystem. The following studies were performed to addressed the objective: (1) an assessment of landscape patterns using geographical information and remote sensing tools; (2) an assessment of habitat patterns based on field surveys of water, vegetation, and soils; (3) ecosystem function modelling to simulate net primary production (NPP) as a function of water quantity, sediments, and habitat type; and (4) fauna habitat modelling linking the results from the assessment of landscape patterns to fauna species. The assessment of landscape patterns revealed a distinct relationship between inundation and vegetation. Habitats in the Tonle Sap were divided into five groups based on annual flood duration, as well as physiognomic factors and human activity: (1) open water, (2) gallery forest, (3) seasonally flooded habitats, (4) transitional habitats, and (5) rainfed habitats. Large habitat shifts could occur as a result of hydropower development scenarios by the 2030s; areas optimal for gallery forest could decrease by 82% from baseline conditions, whereas areas of rainfed habitats could increase by 10-13 % (813-1061 km2). The assessment of habitat patterns demonstrated that despite the complexity and intense human use of this ecosystem, the flood-pulse is the underlying driver of habitat characteristics by (1) determining inundation depth and duration; (2) creating the main soils gradient; (3) limiting the area cleared for agriculture; (4) influencing vegetation structure and water quality; and (5) shaping the composition of plant species. The ecosystem function model was used to estimate a reduction of 9-39% in annual NPP caused by different scenarios of hydropower development and/or climate change during 2032-2042. Cumulative impacts from hydropower would disrupt NPP to a greater extent than climate change. The fauna habitat model revealed that species richness was greatest in the gallery forests and seasonally flooded habitats. Animals that permanently reside in or that rely on these habitats to complete essential life-history stages would be the most affected by future changes. This thesis provides the first quantitative formulation that directly links fundamental components of the Tonle Sap ecosystem to its flood-pulse hydrology. It also provides a comprehensive assessment of the impacts of expected hydrological alterations. Hydropower is expected to bring more abrupt and distinct ecological alterations than climate change in future decades. Relative aerial changes to the gallery forests are expected to be greater than in other habitats. A decline of the Tonle Sap’s ecosystem services will occur if appropriate measures are not implemented. These measures include mitigating hydropower alterations, conserving natural habitats in areas that are likely to remain hydrologically undisturbed, restoring natural habitats in projected areas for optimal growth, and optimizing agricultural practices in the floodplain. Research findings from this thesis focused on the Tonle Sap, but given the fundamental commonalities between this system and other large floodplains, the information presented is highly informative to other large flood-pulse driven systems around the globe.

The use of spatial statistics to investigate ecological processes in plant communities is becoming increasingly widespread. In diverse communities such as tropical rainforests, analysis of spatial structure may help to unravel the various processes that act and interact to maintain high levels of diversity. In particular, a number of contrasting mechanisms have been suggested to explain species coexistence, and these differ greatly in their practical implications for the ecology and conservation of tropical forests. Traditional first-order measures of community structure have proved unable to distinguish these mechanisms in practice, but statistics that describe spatial structure may be able to do so. This is of great interest and relevance as spatially explicit data become available for a range of ecological communities and analysis methods for these data become more accessible. This thesis investigates the potential for inference about underlying ecological processes in plant communities using spatial statistics. Current methodologies for spatial analysis are reviewed and extended, and are used to characterise the spatial signals of the principal theorised mechanisms of coexistence. The sensitivity of a range of spatial statistics to these signals is assessed, and the strength of such signals in natural communities is investigated. The spatial signals of the processes considered here are found to be strong and robust to modelled stochastic variation. Several new and existing spatial statistics are found to be sensitive to these signals, and offer great promise for inference about underlying processes from empirical data. The relative strengths of particular processes are found to vary between natural communities, with any one theory being insufficient to explain observed patterns. This thesis extends both understanding of species coexistence in diverse plant communities and the methodology for assessing underlying process in particular cases. It demonstrates that the potential of spatial statistics in ecology is great and largely unexplored.

There is a need for ecological modelling to help understand the dynamics in ecological systems, and thus aid management decisions to maintain or improve the quality of the ecological systems. This research focuses on non linear statistical modelling of observations from an estuarine system, Gippsland Lakes, on the south-eastern coast of Australia. Feed forward neural networks are used to model chlorophyll time series from a fixed monitoring station at Point King. The research proposes a systematic approach to modelling in ecology using feed forward neural networks, to ensure: (a) that results are reliable, (b) to improve the understanding of dynamics in the ecological system, and (c) to obtain a prediction, if possible. An objective filtering algorithm to enable modelling is presented. Sensitivity analysis techniques are compared to select the most appropriate technique for ecological models. The research generated a chronological profile of relationships between biophysical parameters and chlorophyll level for different seasons. A sensitivity analysis of the models was used to understand how the significance of the biophysical parameters changes as the time difference between the input and predicted value changes. The results show that filtering improves modelling without introducing any noticeable bias. Partial derivative method is found to be the most appropriate technique for sensitivity analysis of ecological feed forward neural networks models. Feed forward neural networks show potential for prediction when modelled on an appropriate time series. Feed forward neural networks also show capability to increase understanding of the ecological environment. In this research, it can be seen that vertical gradient and temperature are important for chlorophyll levels at Point King at time scales from a few hours to a few days. The importance of chlorophyll level at any time to chlorophyll levels in the future reduces as the time difference between them increases.

The fragmentation of natural habitat caused by anthropogenic land use changes is one of the main drivers of the current rapid loss of biodiversity. In face of this threat, ecological research needs to provide predictions of communities' responses to fragmentation as a prerequisite for the effective mitigation of further biodiversity loss. However, predictions of communities' responses to fragmentation require a thorough understanding of ecological processes, such as species dispersal and persistence. Therefore, this thesis seeks an improved understanding of community dynamics in fragmented landscapes. In order to approach this overall aim, I identified key questions on the response of plant diversity and plant functional traits to variations in species' dispersal capability, habitat fragmentation and local environmental conditions. All questions were addressed using spatially explicit simulations or statistical models. In chapter 2, I addressed scale-dependent relationships between dispersal capability and species diversity using a grid-based neutral model. I found that the ratio of survey area to landscape size is an important determinant of scale-dependent dispersal-diversity relationships. With small ratios, the model predicted increasing dispersal-diversity relationships, while decreasing dispersal-diversity relationships emerged, when the ratio approached one, i.e. when the survey area approached the landscape size. For intermediate ratios, I found a U-shaped pattern that has not been reported before. With this study, I unified and extended previous work on dispersal-diversity relationships. In chapter 3, I assessed the type of regional plant community dynamics for the study area in the Southern Judean Lowlands (SJL). For this purpose, I parameterised a multi-species incidence-function model (IFM) with vegetation data using approximate Bayesian computation (ABC). I found that the type of regional plant community dynamics in the SJL is best characterized as a set of isolated “island communities” with very low connectivity between local communities. Model predictions indicated a significant extinction debt with 33% - 60% of all species going extinct within 1000 years. In general, this study introduces a novel approach for combining a spatially explicit simulation model with field data from species-rich communities. In chapter 4, I first analysed, if plant functional traits in the SJL indicate trait convergence by habitat filtering and trait divergence by interspecific competition, as predicted by community assembly theory. Second, I assessed the interactive effects of fragmentation and the south-north precipitation gradient in the SJL on community-mean plant traits. I found clear evidence for trait convergence, but the evidence for trait divergence fundamentally depended on the chosen null-model. All community-mean traits were significantly associated with the precipitation gradient in the SJL. The trait associations with fragmentation indices (patch size and connectivity) were generally weaker, but statistically significant for all traits. Specific leaf area (SLA) and plant height were consistently associated with fragmentation indices along the precipitation gradient. In contrast, seed mass and seed number were interactively influenced by fragmentation and precipitation. In general, this study provides the first analysis of the interactive effects of climate and fragmentation on plant functional traits. Overall, I conclude that the spatially explicit perspective adopted in this thesis is crucial for a thorough understanding of plant community dynamics in fragmented landscapes. The finding of contrasting responses of local diversity to variations in dispersal capability stresses the importance of considering the diversity and composition of the metacommunity, prior to implementing conservation measures that aim at increased habitat connectivity. The model predictions derived with the IFM highlight the importance of additional natural habitat for the mitigation of future species extinctions. In general, the approach of combining a spatially explicit IFM with extensive species occupancy data provides a novel and promising tool to assess the consequences of different management scenarios. The analysis of plant functional traits in the SJL points to important knowledge gaps in community assembly theory with respect to the simultaneous consequences of habitat filtering and competition. In particular, it demonstrates the importance of investigating the synergistic consequences of fragmentation, climate change and land use change on plant communities. I suggest that the integration of plant functional traits and of species interactions into spatially explicit, dynamic simulation models offers a promising approach, which will further improve our understanding of plant communities and our ability to predict their dynamics in fragmented and changing landscapes. / Die Fragmentierung von Landschaften umfasst die Zerschneidung und den Verlust von Flächen mit natürlicher Vegetationsentwicklung und ist eine der Hauptursachen für den gegenwärtigen drastischen Verlust an Biodiversität. Diese Dissertation soll zu einem besseren Verständnis der Vegetationsdynamik in fragmentierten Landschaften beitragen. Damit verbunden ist das Ziel, Vorhersagen über die Reaktion von Pflanzengemeinschaften auf Fragmentierung zu verbessern. Diese Vorhersagen sind notwendig, um gezielte Naturschutzmaßnahmen zur Verminderung eines weiteren Verlustes an Biodiversität umsetzen zu können. In Kapitel 2 der Dissertation wird mit einem Simulationsmodell untersucht, wie sich die Ausbreitungsdistanz von Samen auf die lokale Artenzahl von Pflanzengemeinschaften auswirkt. Dabei zeigte sich, dass längere Ausbreitungsdistanzen die lokale Artenvielfalt sowohl erhöhen, als auch verringern können. Der wichtigste Einflussfaktor war dabei die Artenvielfalt der über-geordneten Pflanzengemeinschaft, in der die betrachtete lokale Gemeinschaft eingebettet war. Im dritten Kapitel wird die Konnektivität zwischen Pflanzengemeinschaften in Habitat-fragmenten, d.h. der Austausch von Arten und Individuen durch Samenausbreitung, im Unter-suchungsgebiet in Israel analysiert. Dafür wurde ein zweites räumliches Simulationsmodell mit statistischen Verfahren an Felddaten angepasst. Der Vergleich des Modells mit den Daten wies auf eine sehr geringe Konnektivität zwischen den Habitatfragmenten hin. Das Modell sagte vorher, dass innerhalb von 1000 Jahren 33% - 60% der Arten aussterben könnten. In Kapitel 4 wird zuerst analysiert, welche Prozesse die Verteilung von funktionellen Eigenschaften in Pflanzengemeinschaften bestimmen. In einem zweiten Schritt wird dann unter-sucht, wie sich funktionelle Eigenschaften von Pflanzengemeinschaften mit dem Niederschlag und der Fragmentierung im Untersuchungsgebiet in Israel verändern. Der Zusammenhang zwischen den Eigenschaften Pflanzenhöhe, sowie spezifischer Blattfläche und der Fragmentierung änderte sich nicht entlang des Niederschlagsgradienten. Im Gegensatz dazu, änderte sich der Zusammenhang zwischen der Samenmasse bzw. der Samenzahl und der Fragmentierung mit dem Niederschlag. Aus den Ergebnissen der ersten Teilstudie wird deutlich, dass Naturschutzmaßnahmen, die natürliche Habitate stärker vernetzen sollen, die Diversität, sowie die Zusammensetzung der übergeordneten Artengemeinschaft berücksichtigen müssen, um Verluste an Biodiversität zu vermeiden. Die Verknüpfung eines räumlichen Simulationsmodells mit Felddaten in der zweiten Teilstudie stellt einen neuen und vielversprechenden Ansatz für die Untersuchung der Auswirkungen verschiedener Management-Szenarien dar. Die dritte Teilstudie ist die erste Analyse der gemeinsamen Auswirkungen von Klima und Fragmentierung auf funktionelle Pflanzen-eigenschaften und zeigt die hohe Bedeutung der Untersuchung von Synergie-Effekten verschiedener Umweltfaktoren. Für zukünftige Forschung legt diese Dissertation nahe, funktionelle Eigenschaften und Konkurrenz zwischen Arten in räumlichen Simulationsmodellen zu berücksichtigen, um das Verständnis von Artengemeinschaften in fragmentierten Landschaften noch weiter zu verbessern.

Coral reefs are experiencing increasing levels of stress due to climate change, overfishing, coastal development and nutrient runoff from agriculture to name a few. They are however, economically vital ecosystems in terms of both their income generating capacity and as a source of food for millions of people around the world. This predicament emphasises the need for effective ecosystem management to be able to balance the benefits of coral reefs with the inherent stressors associated with people utilising their resources. It is particularly important given the potential large scale impacts associated with climate change such as mass coral bleaching events. Similarly, much of the need for direct management of coral reefs exists in developing countries where the resources, information, and technology are limited or unavailable for such a task. This places them, in particular, at the high end of management uncertainty and impact vulnerability. Accordingly, there is a pronounced need to improve this capacity to understand coral reef ecosystem function and to use this to better predict the overall systems level outcome of management options. This thesis has sought to improve our understanding of key ecological elements of coral reef ecosystems and to build on this new knowledge to produce a widely applicable ecosystem model that will allow managers to better understand and predict the outcomes of their actions. Coral reef ecosystem behaviour is far from understood in its entirety and there are many facets that require detailed further investigations to be able to more confidently predict ecosystem response to any given disturbance. To enhance the current understanding of coral reef ecosystems prior to the model development, investigations were undertaken into the dynamics of nitrogen on a coral reef following bleaching induced coral mortality. The results showed that the rates of nitrogen fixation on surfaces made available due to a coral mortality event increased dramatically in the three months following coral mortality, potentially acting as a driving force for the ecosystem to pass through a phase shift to algal dominance. Application of these nitrogen dynamics to entire coral reef ecosystems required a methodology for scaling these sub coral colony processes to entire reefs. This scaling issue is particularly pertinent given the improved understanding of the overwhelming significance of micro-scale processes to community dynamics. The surface index (SI) concept, relating the two-dimensional projected area to the three-dimensional area of corals was refined and developed for variations of gross coral morphologies. This allowed for the scaling of nitrogen flux estimates to be made over entire reef systems, enabling the incorporation of these fluxes into an ecosystem scale model. One of the key factors associated with the potential for a coral reef to recover from a mass coral mortality event is the potential for new corals to successfully recruit. The process of coral recovery could potentially be enhanced if recruitment is viable in the immediate aftermath of a mortality event. Although investigations in this area were inconclusive, extensive herbivore action on turf assemblages up to eight months old indicated that recruitment may be inhibited through the high palatability of turf assemblages in this age bracket. Integrating these processes with the many other published dynamics of coral reefs allowed for the development of the dynamic systems model. By constraining the model structure to known relationships between the modelling parameters, the model can be calibrated to replicate the dynamics of any coral reef ecosystem. This allows the model to be applied to systems where limited data and/or resources are available, making it widely implementable in developing countries such as the small island states scattered around the tropics. The model is ideally suited to the adaptive management framework whereby managers can continually assess the potential future outcomes of management interventions. In addition, due to the spatially inexplicit and generic nature of the model, it can be easily adapted and integrated into large scale regional modelling frameworks or combined with other modelling packages such as socio-economic or fisheries models to provide enhanced management packages. The culmination of the targeted research and integration of existing knowledge has allowed for the development of an ecosystem model for coral reefs that can be easily adopted by coral reef managers throughout the world. It is however, by no means a definitive coral reef ecosystem model and there are many facets that can and should continue to be refined to enhance the reliability of the model.

Coral reefs are experiencing increasing levels of stress due to climate change, overfishing, coastal development and nutrient runoff from agriculture to name a few. They are however, economically vital ecosystems in terms of both their income generating capacity and as a source of food for millions of people around the world. This predicament emphasises the need for effective ecosystem management to be able to balance the benefits of coral reefs with the inherent stressors associated with people utilising their resources. It is particularly important given the potential large scale impacts associated with climate change such as mass coral bleaching events. Similarly, much of the need for direct management of coral reefs exists in developing countries where the resources, information, and technology are limited or unavailable for such a task. This places them, in particular, at the high end of management uncertainty and impact vulnerability. Accordingly, there is a pronounced need to improve this capacity to understand coral reef ecosystem function and to use this to better predict the overall systems level outcome of management options. This thesis has sought to improve our understanding of key ecological elements of coral reef ecosystems and to build on this new knowledge to produce a widely applicable ecosystem model that will allow managers to better understand and predict the outcomes of their actions. Coral reef ecosystem behaviour is far from understood in its entirety and there are many facets that require detailed further investigations to be able to more confidently predict ecosystem response to any given disturbance. To enhance the current understanding of coral reef ecosystems prior to the model development, investigations were undertaken into the dynamics of nitrogen on a coral reef following bleaching induced coral mortality. The results showed that the rates of nitrogen fixation on surfaces made available due to a coral mortality event increased dramatically in the three months following coral mortality, potentially acting as a driving force for the ecosystem to pass through a phase shift to algal dominance. Application of these nitrogen dynamics to entire coral reef ecosystems required a methodology for scaling these sub coral colony processes to entire reefs. This scaling issue is particularly pertinent given the improved understanding of the overwhelming significance of micro-scale processes to community dynamics. The surface index (SI) concept, relating the two-dimensional projected area to the three-dimensional area of corals was refined and developed for variations of gross coral morphologies. This allowed for the scaling of nitrogen flux estimates to be made over entire reef systems, enabling the incorporation of these fluxes into an ecosystem scale model. One of the key factors associated with the potential for a coral reef to recover from a mass coral mortality event is the potential for new corals to successfully recruit. The process of coral recovery could potentially be enhanced if recruitment is viable in the immediate aftermath of a mortality event. Although investigations in this area were inconclusive, extensive herbivore action on turf assemblages up to eight months old indicated that recruitment may be inhibited through the high palatability of turf assemblages in this age bracket. Integrating these processes with the many other published dynamics of coral reefs allowed for the development of the dynamic systems model. By constraining the model structure to known relationships between the modelling parameters, the model can be calibrated to replicate the dynamics of any coral reef ecosystem. This allows the model to be applied to systems where limited data and/or resources are available, making it widely implementable in developing countries such as the small island states scattered around the tropics. The model is ideally suited to the adaptive management framework whereby managers can continually assess the potential future outcomes of management interventions. In addition, due to the spatially inexplicit and generic nature of the model, it can be easily adapted and integrated into large scale regional modelling frameworks or combined with other modelling packages such as socio-economic or fisheries models to provide enhanced management packages. The culmination of the targeted research and integration of existing knowledge has allowed for the development of an ecosystem model for coral reefs that can be easily adopted by coral reef managers throughout the world. It is however, by no means a definitive coral reef ecosystem model and there are many facets that can and should continue to be refined to enhance the reliability of the model.

Coral reefs are experiencing increasing levels of stress due to climate change, overfishing, coastal development and nutrient runoff from agriculture to name a few. They are however, economically vital ecosystems in terms of both their income generating capacity and as a source of food for millions of people around the world. This predicament emphasises the need for effective ecosystem management to be able to balance the benefits of coral reefs with the inherent stressors associated with people utilising their resources. It is particularly important given the potential large scale impacts associated with climate change such as mass coral bleaching events. Similarly, much of the need for direct management of coral reefs exists in developing countries where the resources, information, and technology are limited or unavailable for such a task. This places them, in particular, at the high end of management uncertainty and impact vulnerability. Accordingly, there is a pronounced need to improve this capacity to understand coral reef ecosystem function and to use this to better predict the overall systems level outcome of management options. This thesis has sought to improve our understanding of key ecological elements of coral reef ecosystems and to build on this new knowledge to produce a widely applicable ecosystem model that will allow managers to better understand and predict the outcomes of their actions. Coral reef ecosystem behaviour is far from understood in its entirety and there are many facets that require detailed further investigations to be able to more confidently predict ecosystem response to any given disturbance. To enhance the current understanding of coral reef ecosystems prior to the model development, investigations were undertaken into the dynamics of nitrogen on a coral reef following bleaching induced coral mortality. The results showed that the rates of nitrogen fixation on surfaces made available due to a coral mortality event increased dramatically in the three months following coral mortality, potentially acting as a driving force for the ecosystem to pass through a phase shift to algal dominance. Application of these nitrogen dynamics to entire coral reef ecosystems required a methodology for scaling these sub coral colony processes to entire reefs. This scaling issue is particularly pertinent given the improved understanding of the overwhelming significance of micro-scale processes to community dynamics. The surface index (SI) concept, relating the two-dimensional projected area to the three-dimensional area of corals was refined and developed for variations of gross coral morphologies. This allowed for the scaling of nitrogen flux estimates to be made over entire reef systems, enabling the incorporation of these fluxes into an ecosystem scale model. One of the key factors associated with the potential for a coral reef to recover from a mass coral mortality event is the potential for new corals to successfully recruit. The process of coral recovery could potentially be enhanced if recruitment is viable in the immediate aftermath of a mortality event. Although investigations in this area were inconclusive, extensive herbivore action on turf assemblages up to eight months old indicated that recruitment may be inhibited through the high palatability of turf assemblages in this age bracket. Integrating these processes with the many other published dynamics of coral reefs allowed for the development of the dynamic systems model. By constraining the model structure to known relationships between the modelling parameters, the model can be calibrated to replicate the dynamics of any coral reef ecosystem. This allows the model to be applied to systems where limited data and/or resources are available, making it widely implementable in developing countries such as the small island states scattered around the tropics. The model is ideally suited to the adaptive management framework whereby managers can continually assess the potential future outcomes of management interventions. In addition, due to the spatially inexplicit and generic nature of the model, it can be easily adapted and integrated into large scale regional modelling frameworks or combined with other modelling packages such as socio-economic or fisheries models to provide enhanced management packages. The culmination of the targeted research and integration of existing knowledge has allowed for the development of an ecosystem model for coral reefs that can be easily adopted by coral reef managers throughout the world. It is however, by no means a definitive coral reef ecosystem model and there are many facets that can and should continue to be refined to enhance the reliability of the model.

Esta tese identificou vários aspectos importantes da estrutura trófica e dinâmica de nutrientes em ecossistemas aquáticos sujeitos a perturbações externas, tais como mudanças no nível da água, aumento de carga de nutrientes e matéria orgânica, biomanipulação e clima. Para descrever tais efeitos sobre a estrutura trófica, um complexo modelo ecológico, com base em processos físicos, químicos e biológicos, foi desenvolvido e aplicado em ecossistemas aquáticos tropicais, subtropicais e temperados. Em tais aplicações, uma análise integrada foi empregada considerando a transferência de impactos da bacia para o ecossistema, com a finalidade de entender a real dinâmica dos ecossistemas aquáticos. As simulações revelaram importantes aspectos sobre a estrutura e funcionalidade dos ecossistemas frente as pertubações. Por exemplo, a lagoa Mangueira, um lago raso subtropical no sul do Brasil, pode oscilar entre dois estados alternativos estáveis (alta e baixa transparência da água), dependendo da concentração de ortofosfato e da túrbidez. Foi sugerido também que podem ocorrer profundas alterações na composição do fitoplâncton com a aumento sistemático da concentração de ortofosfato. Foram constatados significativos gradientes horizontais e verticais na estrutura trófica da lagoa Mangueira, indicando a importância de considerar a heterogeneidade espacial em grandes lagos para melhorar a compreensão dos processos ecológicos e padrões. Cenários críticos de simulação indicaram que o aumento da carga de nutrientes afeta negativamente a transparência da água, diferentemente do efeito de mudanças climáticas. Pequenas alterações na estrutura trófica causadas por esses fatores indicam uma forte resistência da lagoa Mangueira. O modelo ecológico também foi útil para descrever a dinâmica do ecossistema, antes e depois da biomanipulação do lago Engelsholm localizado na Dinamarca. Nesse caso, o modelo conseguiu razoavelmente prever a mudança para um estado de alta transparência da após a biomanipulação, entretanto a mudança na composição do fitoplâncton foi muito mais difícil de representar. Além disso, aplicações mais simples também auxiliaram o processo de tomada de decisão e planejamento. A modelagem ecológica aplicada em ecossistemas aquáticos mostrou ser uma alternativa promissora para a gestão integrada dos recursos naturais. / This thesis identified several important features of the trophic structure and dynamics of nutrients in ecosystems subjected to external disturbances such as changes in the water level, nutrient and organic matter loading, fish communities and climate. To describe such effects on trophic structure, a complex ecological model, based on physical, chemical and biological processes has been developed and applied in tropical, subtropical and temperate aquatic ecosystems. In such applications, an integrated analysis was employed considering the transfer of impacts from the watershed to the ecosystem, in order to bring closer the understading of the ecosystem dynamics. Simulations revealed important features about system structure and functionality in front of disturbances. For instance, lake Mangueira a subtropical lake in Southern Brazil can oscilate between two alternative steady stables (clear or turbid water), depending orthophosphate concentration and water transparency. It was also suggested that can occurs profound changes in phytoplankton composition from increase of orthophosphate concentration. Horizontal and vertical gradients in the trophic structure are notable in lake Mangueira, indicating the importance to consider spatial heterogeneity in large lakes to improve understanding of ecological processes and patterns. Simulation scenarios indicated that increased nutrient loading negatively affects water transparency, differently from climate changes. Slight changes in trophic structure caused by those factors indicate a strong resistence of the lake Mangueira. The ecological model also was useful to describe the ecosystem dynamics before and after biomanipulation of lake Engelsholm located in Denmark. In that case, the model could reasonably well predict a shift to a clear water state, but the changed composition of phytoplankton functional groups was much more hard to represent. Furthermore, simpler applications also provided support for decision making and planning in the ecosystem. Ecological modeling applied in aquatic ecosystems showed be a promising alternative towards the integrated management of natural resources.

Esta tese identificou vários aspectos importantes da estrutura trófica e dinâmica de nutrientes em ecossistemas aquáticos sujeitos a perturbações externas, tais como mudanças no nível da água, aumento de carga de nutrientes e matéria orgânica, biomanipulação e clima. Para descrever tais efeitos sobre a estrutura trófica, um complexo modelo ecológico, com base em processos físicos, químicos e biológicos, foi desenvolvido e aplicado em ecossistemas aquáticos tropicais, subtropicais e temperados. Em tais aplicações, uma análise integrada foi empregada considerando a transferência de impactos da bacia para o ecossistema, com a finalidade de entender a real dinâmica dos ecossistemas aquáticos. As simulações revelaram importantes aspectos sobre a estrutura e funcionalidade dos ecossistemas frente as pertubações. Por exemplo, a lagoa Mangueira, um lago raso subtropical no sul do Brasil, pode oscilar entre dois estados alternativos estáveis (alta e baixa transparência da água), dependendo da concentração de ortofosfato e da túrbidez. Foi sugerido também que podem ocorrer profundas alterações na composição do fitoplâncton com a aumento sistemático da concentração de ortofosfato. Foram constatados significativos gradientes horizontais e verticais na estrutura trófica da lagoa Mangueira, indicando a importância de considerar a heterogeneidade espacial em grandes lagos para melhorar a compreensão dos processos ecológicos e padrões. Cenários críticos de simulação indicaram que o aumento da carga de nutrientes afeta negativamente a transparência da água, diferentemente do efeito de mudanças climáticas. Pequenas alterações na estrutura trófica causadas por esses fatores indicam uma forte resistência da lagoa Mangueira. O modelo ecológico também foi útil para descrever a dinâmica do ecossistema, antes e depois da biomanipulação do lago Engelsholm localizado na Dinamarca. Nesse caso, o modelo conseguiu razoavelmente prever a mudança para um estado de alta transparência da após a biomanipulação, entretanto a mudança na composição do fitoplâncton foi muito mais difícil de representar. Além disso, aplicações mais simples também auxiliaram o processo de tomada de decisão e planejamento. A modelagem ecológica aplicada em ecossistemas aquáticos mostrou ser uma alternativa promissora para a gestão integrada dos recursos naturais. / This thesis identified several important features of the trophic structure and dynamics of nutrients in ecosystems subjected to external disturbances such as changes in the water level, nutrient and organic matter loading, fish communities and climate. To describe such effects on trophic structure, a complex ecological model, based on physical, chemical and biological processes has been developed and applied in tropical, subtropical and temperate aquatic ecosystems. In such applications, an integrated analysis was employed considering the transfer of impacts from the watershed to the ecosystem, in order to bring closer the understading of the ecosystem dynamics. Simulations revealed important features about system structure and functionality in front of disturbances. For instance, lake Mangueira a subtropical lake in Southern Brazil can oscilate between two alternative steady stables (clear or turbid water), depending orthophosphate concentration and water transparency. It was also suggested that can occurs profound changes in phytoplankton composition from increase of orthophosphate concentration. Horizontal and vertical gradients in the trophic structure are notable in lake Mangueira, indicating the importance to consider spatial heterogeneity in large lakes to improve understanding of ecological processes and patterns. Simulation scenarios indicated that increased nutrient loading negatively affects water transparency, differently from climate changes. Slight changes in trophic structure caused by those factors indicate a strong resistence of the lake Mangueira. The ecological model also was useful to describe the ecosystem dynamics before and after biomanipulation of lake Engelsholm located in Denmark. In that case, the model could reasonably well predict a shift to a clear water state, but the changed composition of phytoplankton functional groups was much more hard to represent. Furthermore, simpler applications also provided support for decision making and planning in the ecosystem. Ecological modeling applied in aquatic ecosystems showed be a promising alternative towards the integrated management of natural resources.