Filtern
Erscheinungsjahr
- 2016 (18) (entfernen)
Dokumenttyp
- Dissertation (18) (entfernen)
Sprache
- Englisch (18) (entfernen)
Schlagworte
- Ad-hoc-Netz (1)
- Algorithmische Geometrie (1)
- Beaconless (1)
- Bedrohte Tiere (1)
- Daphnia (1)
- Distributed Algorithm (1)
- Drahtloses Sensorsystem (1)
- Drahtloses vermachtes Netz (1)
- E-Participation (1)
- Ebener Graph (1)
- Enterprise Architecture Framework (1)
- Evacuation modeling (1)
- Flow decomposition (1)
- Geographic routing (1)
- Geometric spanner (1)
- Graph (1)
- IT-Security (1)
- Komplexität / Algorithmus (1)
- Local algorithm (1)
- Maifisch (1)
- Metamodel (1)
- Nachbarschaftsgraph (1)
- Netzwerktopologie (1)
- Optimization (1)
- Planar graphs (1)
- Einbringen (1)
- Quasi unit disk graph (1)
- Reactive algorithm (1)
- Reference Model (1)
- Rhein (1)
- Routing (1)
- Security Requirements (1)
- Semantic Web (1)
- Semantic Web Data (1)
- Southern Amazonia (1)
- Unit disk graph (1)
- Verteilter Algorithmus (1)
- Wanderfische (1)
- Wireless sensor network (1)
- Zooplankton (1)
- Zusammenhängender Graph (1)
- decomposition (1)
- description logic (1)
- ecotoxicology (1)
- energetics (1)
- field experiment (1)
- flows over time (1)
- fluid disturbances (1)
- fungicide (1)
- greenhouse gases (1)
- kinematics (1)
- land use change (1)
- modeling (1)
- olive mill wastewater (1)
- pesticide (1)
- reasoning (1)
- soil water repellency (1)
- soils (1)
- stream (1)
- swarming (1)
- swimming behaviour (1)
- thermal analysis (1)
- zooplankton (1)
Institut
Aktuelle Schätzungen bestätigten, dass Binnengewässer eine erhebliche Menge Methan (CH4) und Kohlendioxid (CO2) sowohl auf regionaler Ebene, als auch global freisetzen. Jedoch basieren diese Schätzungen auf extrapolierten gemessenen Daten, ungenügender Auflösung der räumlich-zeitlichen Variabilität und es mangelt an Daten aus ariden und semi-ariden Gebieten, sowie den Kohlestoffquellen aus Kläranlagen.
Für die hier vorliegende Studie analysierten wir monatliche hydrologische und meteorologische Daten sowie Daten zur Wasserqualität von drei Stauseen aus dem Gebiet des unteren Jordans, die zur Trinkwassergewinnung und zur Bewässerung genutzt werden, und schätzten damit deren Emissionsrate an CO2 ab. Wir untersuchten den Effekt von Kläranlagen auf die umliegenden Gewässer im Hinblick auf CH4 und CO2-Emissionen indem wir saisonal aufgelöste Daten der Konzentration der beiden gelösten Gase in Kläranlagenauslässen und in Vorflutern von neun Kläranlagen in Deutschland analysierten. Mithilfe von Low-Cost-Methoden die die CO2-Transportrate und die Ausgasungsrate über Gasblasen messen, untersuchten wir die räumliche und zeitliche Variabilität der CH4 und CO2-Emissionen von aquatischen Süßwasser-Ökosystemen.
Unsere Schätzungen zeigen, dass Stauseen in semi-ariden Regionen CO2 übersättigt sind und somit CO2 an die Atmosphäre abgeben, also eine Netto-Quelle sind.
Die Größenordnung der beobachteten Transportraten der drei jordanischen Stauseen ist vergleichbar mit denen von tropischen Stauseen (3,3 g CO2 m-2 Tag-1). Die CO2-Emissionsrate ist abhängig von Änderungen der Wasseroberfläche, welche durch den Betrieb der Stauseen verursacht sind. Kläranlagen entlassen eine beachtlichen Menge an CH4 (30.9±40.7 kg Jahr-1) und CO2 (0.06±0.05 Gg Jahr-1) in ihre umgebenden Flüsse und Bäche. Deren Emissionsraten sind durch diese Einleitung der Kläranlagen um 1,2-fach für CH4 oder 8,6-fach für CO2 erhöht. Unsere Ergebnisse zeigen, dass sowohl die diffusive als auch die Gasblasenemissionsrate räumlich und zeitlich variabel ist, weshalb beide Emissionsraten bei zukünftigen Studien auch in der nötigen Auflösung gemessen werden sollten.
Wir schlussfolgern, dass bei zukünftigen Emissionsmessungen und –schätzungen von Binnengewässern auch die Gewässerbewirtschaftung, die Kohlenstoffquelle von Kläranlagen und die räumliche und zeitliche Variabilität der Emissionen beachtet werden sollten.
Agriculture covers one third of the world land area and has become a major source of water pollution due to its heavy reliance on chemical inputs, namely fertilisers and pesticides. Several thousands of tonnes of these chemicals are applied worldwide annually and partly reach freshwaters. Despite their widespread use and relatively unspecific modes of action, fungicides are the least studied group of pesticides. It remains unclear whether the taxonomic groups used in pesticide risk assessment are protective for non-target freshwater fungi. Fungi and bacteria are the main microbial decomposers converting allochthonous organic matter (litter) into a more nutritious food resource for leaf-shredding macroinvertebrates. This process of litter decomposition (LD) is central for aquatic ecosystem because it fuels local and downstream food webs with energy and nutrients. Effects of fungicides on decomposer communities and LD have been mainly analysed under laboratory conditions with limited representation of the multiple factors that may moderate effects in the field.
In this thesis a field study was conducted in a German vineyard area to characterise recurrent episodic exposure to fungicides in agricultural streams (chapter 2) and its effects on decomposer communities and LD (chapter 3). Additionally, potential interaction effects of nutrient enrichment and fungicides on decomposer communities and LD were analysed in a mesocosm experiment (chapter 4).
In the field study event-driven water sampling (EDS) and passive sampling with EmporeTM styrene-divinylbenzene reverse phase sulfonated disks (SDB disks) were used to assess exposure to 15 fungicides and 4 insecticides. A total of 17 streams were monitored during 4 rainfall events within the local application period of fungicides in 2012. EDS exceeded the time-weighted average concentrations provided by the SDB disks by a factor of 3, though high variability among compounds was observed. Most compounds were detected in more than half of the sites and mean and maximum peak (EDS) concentrations were under 1 and 3 µg/l, respectively. Besides, SDB disk-sampling rates and a free-software solution to derive sampling rates under time-variable exposure were provided.
Several biotic endpoints related to decomposers and LD were measured in the same sampling sites as the fungicide monitoring, coinciding with the major litter input period. Our results suggest that polar organic fungicides in streams change the structure of the fungal community. Causality of this finding was supported by a subsequent microcosm experiment. Whether other effects observed in the field study, such as reduced fungal biomass, increased bacterial density or reduced microbial LD can be attributed to fungicides remains speculative and requires further investigation. By contrast, neither the invertebrate LD nor in-situ measured gammarid feeding rates correlated with water-borne fungicide toxicity, but both were negatively associated with sediment copper concentrations. The mesocosm experiment showed that fungicides and nutrients affect microbial decomposers differently and that they can alter community structure, though longer experiments are needed to determine whether these changes may propagate to invertebrate communities and LD. Overall, further studies should include representative field surveys in terms of fungicide pollution and physical, chemical and biological conditions. This should be combined with experiments under controlled conditions to test for the causality of field observations.
This study had two main aims. The first one was to investigate the quality of lesson plans. Two important features of lesson plans were used as a basis to determine the quality of lesson plans. These are adaptability to preconditions and cognitive activation of students. The former refers to how the planning teacher considers the diversity of students pre-existing knowledge and skills. The latter refers to how the planning teacher sequences deep learning tasks and laboratory activities to promote the cognitive activation of students.
The second aim of the study was to explore teachers thinking about and explanation of externally generated feedback data on their students’ performance. The emphasis here was to understand how the teachers anticipate planning differentiated lessons to accommodate the variations in students learning outcomes revealed by the feedback data.
The study followed a qualitative approach with multiple sources of data. Concept maps, questionnaires, an online lesson planning tool, standardized tests, and semi-structured interviews were the main data collection instruments used in the study. Participants of this study were four physics teachers teaching different grade levels. For the purpose of generating feedback for the participant teachers, a test was administered to 215 students. Teachers were asked to plan five lessons for their ongoing practices. The analysis showed that the planned lessons were not adapted to the diversity in students pre-existing knowledge and skills. The analysis also indicated that the lessons planned had limitations with regard to cognitive activation of students. The analysis of the interview data also revealed that the participant teachers do not normally consider differentiating lessons to accommodate the differences in students learning, and place less emphasis on the cognitive activation of students. The analysis of the planned lessons showed a variation in teachers approach in integrating laboratory activities in the sequence of the lessons ranging from a complete absence through a demonstrative to an investigative approach. Moreover, the findings from the interviews indicated differences between the participant teachers espoused theory (i.e. what they said during interview) and their theory- in –use (i.e. what is evident from the planned lessons). The analysis of the interview data demonstrated that teachers did not interpret the data, identify learning needs, draw meaningful information from the data for adapting (or differentiating) instruction. They attributed their students’ poor performance to task difficulty, students’ ability, students’ motivation and interest. The teachers attempted to use the item level and subscale data only to compare the relative position of their class with the reference group. However, they did not read beyond the data, like identifying students learning needs and planning for differentiated instruction based on individual student’s performance.
The largest population of the anadromous Allis shad (A. alosa) of the 19th century was found in River Rhine and has to be considered extinct today. To facilitate the return of A. alosa into River Rhine an EU LIFE-project was initiated in 2007. The overall objective of this thesis was to assist aquaculture and stocking-measures at River Rhine, as well as to support restoration and conservation of populations of Allis shad in Europe.
By culturing the free-swimming nematode T. aceti in a solution of cider vinegar we developed a cost-effective live food organism for the larviculture of fish. As indicated by experiments with C. maraena, T. aceti cannot be regarded as an alternative to Artemia nauplii. However it has to be considered a suitable supplemental feed in the early rearing of C. maraena by providing essential fatty acids, thereby optimizing growth.
Also mass-marking practices with Oxytetracycline, as they are applied in the restocking of Allis shad have been evaluated. In experiments with D. rerio we demonstrated that water hardness can detrimentally affect mortality during marking and has to be considered crucial in the development of marking protocols for freshwater fish.
In order to get independent from wild spawners an ex-situ Broodstock-facility for Allis shad was established in 2011. Upon examination of two complete year classes of this broodstock, we found a high prevalence of various malformations, which could be traced back to distinct cysts developing one month post hatch. Despite applying a variety of clinical tests we could not identify any infectious agents causing these malformations. The observed malformations are probably a consequence of suboptimal feeding practices or the properties of the physio-chemical rearing environment.
The decline of stocks of A. alosa in Europe has been largely explained with the increase of river temperatures as a consequence of global warming. By investigating the temperature physiology of larval Allis shad we demonstrated that A. alosa ranges among the most thermo-tolerant species in Europe and that correlations between rising temperatures and the disappearance of this species have to be understood in a synecological context and by integrating a variety of stressors other than temperature. By capturing and examining juvenile and adult Allis shad from River Rhine, we demonstrated the first natural reproduction of A. alosa in River Rhine since nearly 100 years and the success of stocking measures within the framework of the LIFE project.
Confidentiality, integrity, and availability are often listed as the three major requirements for achieving data security and are collectively referred to as the C-I-A triad. Confidentiality of data restricts the data access to authorized parties only, integrity means that the data can only be modified by authorized parties, and availability states that the data must always be accessible when requested. Although these requirements are relevant for any computer system, they are especially important in open and distributed networks. Such networks are able to store large amounts of data without having a single entity in control of ensuring the data's security. The Semantic Web applies to these characteristics as well as it aims at creating a global and decentralized network of machine-readable data. Ensuring the confidentiality, integrity, and availability of this data is therefore also important and must be achieved by corresponding security mechanisms. However, the current reference architecture of the Semantic Web does not define any particular security mechanism yet which implements these requirements. Instead, it only contains a rather abstract representation of security.
This thesis fills this gap by introducing three different security mechanisms for each of the identified security requirements confidentiality, integrity, and availability of Semantic Web data. The mechanisms are not restricted to the very basics of implementing each of the requirements and provide additional features as well. Confidentiality is usually achieved with data encryption. This thesis not only provides an approach for encrypting Semantic Web data, it also allows to search in the resulting ciphertext data without decrypting it first. Integrity of data is typically implemented with digital signatures. Instead of defining a single signature algorithm, this thesis defines a formal framework for signing arbitrary Semantic Web graphs which can be configured with various algorithms to achieve different features. Availability is generally supported by redundant data storage. This thesis expands the classical definition of availability to compliant availability which means that data must only be available as long as the access request complies with a set of predefined policies. This requirement is implemented with a modular and extensible policy language for regulating information flow control. This thesis presents each of these three security mechanisms in detail, evaluates them against a set of requirements, and compares them with the state of the art and related work.
This thesis presents novel approaches for integrating context information into probabilistic models. Data from social media is typically associated with metadata, which includes context information such as timestamps, geographical coordinates or links to user profiles. Previous studies showed the benefits of using such context information in probabilistic models, e.g.\ improved predictive performance. In practice, probabilistic models which account for context information still play a minor role in data analysis. There are multiple reasons for this. Existing probabilistic models often are complex, the implementation is difficult, implementations are not publicly available, or the parameter estimation is computationally too expensive for large datasets. Additionally, existing models are typically created for a specific type of content and context and lack the flexibility to be applied to other data.
This thesis addresses these problems by introducing a general approach for modelling multiple, arbitrary context variables in probabilistic models and by providing efficient inference schemes and implementations.
In the first half of this thesis, the importance of context and the potential of context information for probabilistic modelling is shown theoretically and in practical examples. In the second half, the example of topic models is employed for introducing a novel approach to context modelling based on document clusters and adjacency relations in the context space. They can cope with areas of sparse observations and These models allow for the first time the efficient, explicit modelling of arbitrary context variables including cyclic and spherical context (such as temporal cycles or geographical coordinates). Using the novel three-level hierarchical multi-Dirichlet process presented in this thesis, the adjacency of ontext clusters can be exploited and multiple contexts can be modelled and weighted at the same time. Efficient inference schemes are derived which yield interpretable model parameters that allow analyse the relation between observations and context.
The application of pesticides to agricultural areas can result in transport to adjacent non-target environments. In particular, surface water systems are likely to receive agricultural pesticide input. When pesticides enter aquatic environments, they may pose a substantial threat to the ecological integrity of surface water systems. To minimize the risk to non-target ecosystems the European Union prescribes an ecotoxicological risk assessment within the registration procedure of pesticides, which consists of an effect and an exposure assessment.
This thesis focuses on the evaluation of the exposure assessment and the implications to the complete regulatory risk assessment, and is based on four scientific publications. The main part of the thesis focuses on evaluation of the FOCUS modelling approach, which is used in regulatory risk assessment to predict pesticide surface water concentrations. This was done by comparing measured field concentrations (MFC) of agricultural insecticides (n = 466) and fungicides (n = 417) in surface water to respective predicted environmental concentrations (PEC) calculated with FOCUS step 1 to step 4 at two different levels of field relevance. MFCs were extracted from the scientific literature and were measured in field studies conducted primarily in Europe (publications 1 and 3).
In addition, an alternative fugacity-based multimedia mass-balance model, which needs fewer input parameters and less computing effort, was used to calculate PECs for the same insecticide MFC dataset and compared to the FOCUS predictions (publication 3). Furthermore, FOCUS predictions were also conducted for veterinary pharmaceuticals in runoff from an experimental plot study, to assess the FOCUS predictions for a different class of chemicals with a different relevant entry pathway (publication 2).
In publication 4, the FOCUS step-3 approach was used to determine relevant insecticide exposure patterns. These patterns were analysed for different monitoring strategies and the implications for the environmental risk assessment (publication 4).
The outcome of this thesis showed that the FOCUS modelling approach is neither protective nor appropriate in predicting insecticide and fungicide field concentrations. Up to one third of the MFCs were underpredicted by the model calculations, which means that the actual risk might be underestimated. Furthermore, the results show that a higher degree of realism even reduces the protectiveness of model results and that the model predictions are worse for highly hydrophobic and toxic pyrethroids.
In addition, the absence of any relationship between measured and predicted concentrations questions the general model performance quality (publication 1 and 3). Further analyses revealed that deficiencies in protectiveness and predictiveness of the environmental exposure assessment might even be higher than shown in this thesis, because actual short-term peak concentrations are only detectable with an event-related sampling strategy (publication 4). However, it was shown that the PECs of a much simpler modelling approach are much more appropriate for the prediction of insecticide MFC, especially for calculations with a higher field relevance (publication 3). The FOCUS approach also failed to predict concentrations of veterinary pharmaceuticals in runoff water (publication 2). In conclusion, the findings of this thesis showed that there is an urgent need for the improvement of exposure predictions conducted in the environmental risk assessment of pesticides as a group of highly relevant environmental chemicals, to ensure that the increasing use of those chemicals does not lead to further harmful effects in aquatic ecosystems.
The establishment of aquatic alien species can strongly affect community and food web structure of the invaded systems and thus represents a major threat to native biodiversity. One of the most important aquatic invasive species in European rivers is the Ponto-Caspian amphipod Dikerogammarus villosus. The species invaded most of the major European waterways within two decades, often associated with a decline of many other macroinvertebrate species, including other amphipods. Based on laboratory results predation by the so called ‘killer shrimp’ is often regarded as the key driver for observed displacement effects, but recent studies indicated a minor relevance of predation by D. villosus in the field. To allow the determination of exact predator-prey interactions from field samples, I established 22 group-specific rDNA primers for freshwater taxa suitable for prey species identification in dietary samples (Chapter II) and an approach for the screening of D. villosus gut contents using 16 of these primers. Combining genetic gut content analyses, with one of these primers, and stable isotope analyses, I examined the importance of intraguild predation (IGP) by D. villosus, which is often assumed the key driver for the displacement of native amphipod species, at an invasion front of the species in Switzerland (Chapter III). The results of this study revealed a low importance of IGP during this particular D. villosus invasion and indicated an overall sparsely predacious feeding behaviour of the species. As the feeding behaviour of D. villosus is supposed to differ between habitats and this study was only conducted at a few sampling sites of one river, I also investigated the role of predation by D. villosus at multiple sites of the River Rhine system, covering a broad range of microhabitats (Chapter IV). In keeping with the results from the invasion front results of this study strongly indicated a sparsely predacious feeding but rather a flexible feeding behaviour of D. villosus even within the same microhabitat.
However, established populations of D. villosus have changed aquatic food webs and can be expected to affect aquatic-terrestrial energy fluxes. In Chapter V of my thesis, I present a field study investigating the impact of D. villosus on the diet of two riparian spider taxa. The results of this study indicate an effect of D. villosus on the terrestrial food web via cross-ecosystem resource flow.
In conclusion, D. villosus influences terrestrial food webs by altering cross-ecosystem resource fluxes, but it is rather an opportunistic omnivore than a predator in the field.
Conversion of natural vegetation into cattle pastures and croplands results in altered emissions of greenhouse gases (GHG), such as carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O). Their atmospheric concentration increase is attributed the main driver of climate change. Despite of successful private initiatives, e.g. the Soy Moratorium and the Cattle Agreement, Brazil was ranked the worldwide second largest emitter of GHG from land use change and forestry, and the third largest emitter from agriculture in 2012. N2O is the major GHG, in particular for the agricultural sector, as its natural emissions are strongly enhanced by human activities (e.g. fertilization and land use changes). Given denitrification the main process for N2O production and its sensitivity to external changes (e.g. precipitation events) makes Brazil particularly predestined for high soil-derived N2O fluxes.
In this study, we followed a bottom-up approach based on a country-wide literature research, own measurement campaigns, and modeling on the plot and regional scale, in order to quantify the scenario-specific development of GHG emissions from soils in the two Federal States Mato Grosso and Pará. In general, N2O fluxes from Brazilian soils were found to be low and not particularly dynamic. In addition to that, expected reactions to precipitation events stayed away. These findings emphasized elaborate model simulations in daily time steps too sophisticated for regional applications. Hence, an extrapolation approach was used to first estimate the influence of four different land use scenarios (alternative futures) on GHG emissions and then set up mitigation strategies for Southern Amazonia. The results suggested intensification of agricultural areas (mainly cattle pastures) and, consequently, avoided deforestation essential for GHG mitigation.
The outcomes of this study provide a very good basis for (a) further research on the understanding of underlying processes causing low N2O fluxes from Brazilian soils and (b) political attempts to avoid new deforestation and keep GHG emissions low.
Reaktiv lokale Algorithmen sind verteilte Algorithmen, die den Anforderungen großer, batteriebetriebener, Drahtloser Ad Hoc und Sensornetzwerke im besonderen Maße gerecht werden. Durch Vermeidung überflüssiger Nachrichtenübertragungen sowie Verzicht auf proaktive Ermittlung von Nachbarschaftstabellen (d.h. beaconing) minimieren solche Algorithmen den Kommunikationsaufwand und skalieren gut bei wachsender Netzgröße. Auf diese Weise werden Ressourcen wie Bandbreite und Energie geschont, es kommt seltener zu Nachrichtenkollisionen und dadurch zu einer Erhöhung der Paketempfangsrate, sowie einer Reduktion der Latenzen.
Derzeit wird diese Algorithmenklasse hauptsächlich für Geografisches Routing, sowie zur Topologiekontrolle, insbesondere zur Ermittlung der Adjazenzliste eines Knotens in zusammenhängenden, kantenschnittfreien (planaren) Repräsentationen des Netzgraphen, eingesetzt. Ersteres ermöglicht drahtlose multi-hop Kommunikation auf Grundlage von geografischen Knotenpositionen ohne Zuhilfenahme zusätzlicher Netzwerkinfrastruktur, wohingegen Letzteres eine hinreichende Grundlage für effiziente, lokale Lösungen einer Reihe algorithmischer Problemstellungen ist.
Die vorliegende Dissertation liefert neue Erkenntnisse zum Forschungsgebiet der reaktiven Algorithmen, zum Einen auf einer abstrakten Ebene und zum Anderen durch die Einführung neuer Algorithmen.
Erstens betrachtet diese Arbeit reaktive Algorithmen erstmalig im Ganzen und als eigenständiges Forschungsfeld. Es wird eine umfangreiche Literaturstudie zu dieser Thematik präsentiert, welche die aus der Literatur bekannten Algorithmen, Techniken und Anwendungsfelder systematisch auflistet, klassifiziert und einordnet. Weiterhin wird das mathematische Konzept der O- und Omega-reaktiv lokalen Topologiekontrolle eingeführt. Dieses Konzept ermöglicht erstmals die eindeutige Unterscheidung reaktiver von konventionellen, beacon-basierten, verteilten Topologiekontrollalgorithmen. Darüber hinaus dient es als Klassifikationsschema für existierende, sowie zukünftige Algorithmen dieser Art. Zu guter Letzt ermöglicht dieses Konzept grundlegende Aussagen über die Mächtigkeit des reaktiven Prinzips, welche über Entwurf und Analyse von Algorithmen hinaus reichen.
Zweitens werden in dieser Arbeit neue reaktiv lokale Algorithmen zur Topologiekontrolle und Geografischem Routing eingeführt, wobei drahtlose Netze durch Unit Disk bzw. Quasi Unit Disk Graphen modelliert werden. Diese Algorithmen berechnen für einen gegebenen Knoten die lokale Sicht auf zusammenhängende, planare, Euklidische bzw. Topologische Spanner mit konstanter Spannrate bzgl. des Netzgraphen und routen Nachrichten reaktiv entlang der Kanten dieser Spanner, wobei die Nachrichtenauslieferung garantiert wird. Alle bisher bekannten Verfahren sind entweder nicht reaktiv oder gewährleisten keine konstanten Euklidischen oder Topologischen Spannraten. Ein wesentliches Teilergebnis dieser Arbeit ist der Nachweis, dass die partielle Delaunay Triangulierung (PDT) ein Euklidischer Spanner mit konstanter Spannrate für Unit Disk Graphen ist.
Die in dieser Dissertation gewonnenen Erkenntnisse bilden die Basis für grundlegende und strukturierte Forschung auf diesem Gebiet und zeigen, dass das reaktive Prinzip ein wichtiges Werkzeug des Algorithmenentwurfs für Drahtlose Ad Hoc und Sensornetzwerke ist.