Refine
Year of publication
- 2012 (36) (remove)
Document Type
- Doctoral Thesis (11)
- Diploma Thesis (8)
- Part of Periodical (8)
- Bachelor Thesis (4)
- Master's Thesis (4)
- Conference Proceedings (1)
Language
- English (36) (remove)
Keywords
- Petri-Netze (2)
- probability propagation nets (2)
- 101companies (1)
- API Analysis (1)
- API Migratiom (1)
- Africa (1)
- Afrika (1)
- Archivierung (1)
- Auditing (1)
- Ausbreitung (1)
Institute
- Fachbereich 4 (25)
- Institut für Informatik (4)
- Fachbereich 7 (3)
- Fachbereich 8 (3)
- Institut für Integrierte Naturwissenschaften, Abt. Biologie (3)
- Institute for Web Science and Technologies (3)
- Institut für Management (1)
- Institut für Wirtschafts- und Verwaltungsinformatik (1)
- Mathematisches Institut (1)
A taxonomic revision of the genus Pteris in tropical Africa revealed 26 species. An identification key to the species is provided. Morphological characters were used to prepare a cladistic analysis of the relevant taxa. Each species was evaluated concerning the IUCN red list status. Only Pteris mkomaziensis was considered as Near Threatened, and all other species only as Least Concern. An inventory of the ferns of Kakamega Forest / Kenya and Budongo Forest / Uganda revealed 85 species in Kakamega and 66 species in Budongo. Life form spectra were analysed and the ferns were studied for their value for bioindication.
Software projects typically rely on several, external libraries. The interface provided by such a library is called API (application programming interface). APIs often evolve over time, thereby implying the need to adapt applications that use them. There are also reasons which may call for the replacement of one library by another one, what also results in a need to adapt the applications where the library is replaced. The process of adapting applications to use a different API is called API migration. Doing API migration manually is a cumbersome task. Automated API migration is an active research field. A related field of research is API analysis which can also provide data for developing API migration tools.
The following thesis investigates techniques and technologies for API analysis and API migration frameworks. To this end, design patterns are leveraged. These patterns are based on experience with API analysis and migration within the Software Languages Team.
In summary, this study revealed the widespread occurrence of antiviral drugs in the aquatic environment. Furthermore, it could be shown that the elimination of pharmaceuticals in both biological and oxidative treatment do not necessarily result in their mineralization but rather leads to the formation of a variety of transformation and oxidation products.
This is one of the first studies in which the fate and in particular the transformation of pharmaceuticals has been comprehensively investigated in almost the complete water cycle, from biological wastewater treatment to advanced oxidation processes via ozone. It was shown that the transformation of pharmaceuticals in the urban water cycle can ultimately result in the formation of toxic transformation products.
Wild boars belong to the most wide spread ungulates in the world. They are characterized by a well performed adaption to their environment mainly due to their omnivorous dietary. The wild boar population in Germany increased during the past three decades. Nowadays their high density leads to problems in agricultural areas due to damage of crops and plays a significant role as disease vector as the classical swine fever. For an effective population management population size information is of crucial importance. Different traditional methods exist to estimate population sizes as direct sightnings, faecal drop counts or hunting harvest which provide only relative estimates and population trends. Absolute population sizes could be yielded by a Capture-Mark-Recapture (CMR) approach. However, capturing of wild boars is difficult to realize and costly in terms of personnel and field effort.
Furthermore the capture probabilities are heterogeneous due to the variable behaviour of individuals influenced by age, sex, and experience of the animals. Non-invasive genetic methods are a promising complement to the traditional methods for population size estimation particularly for wild boar. These methods reduce stress and capture bias and increase the number of re-captures. Faeces proved to be a suitable DNA source for wild boar genotyping, due to almost equal capture probability. However working with faeces implicates difficulties such as low DNA rnquality and quantity, genotyping errors as dropout and false alleles.
The main aim of the present study was to develop a reliable, cost-efficient, reproducible and practicable method for wild boar genotyping. This method should provide a reliable dataset of genotypes obtained from the collected faeces samples. Individual identification forms the basis for an improved mark-recapture approach. As there is no sound method for absolute population counts in free living wild boar, reference values for the validation of this new approach are missing. Therefore, different routines to reduce and to assess genotyping errors were compared within this thesis. For maximum amplification rate, the storage, the extraction methods and the PCR-procedure were optimised. A step by step procedure was evaluated in order to determine the minimum required microsatellite (MS) number for reliable individual identification including a test with family groups (female and embryo tissue) to distinguish even between close relatives. A multiple-tubes approach, post-amplification checking and different correction procedures were applied to reduce genotyping errors. In order to quantify real genotyping error rates (GER) of datasets derived from sampling in the Palatinate Forest in western Germany, different methods for GER determination were compared with each other, obtaining GERs between 0% and 57.5%. As a consequence, more strict criteria for the multi-tube approach and increased repetition number of homozygous samples were used. An additional method validation was the implementation of a blind test to achieve the reliability of the genotyping and error checking procedure. Finally a strict and practicable proposal for the lab procedure was developed, by beginning with faecal sample collection and ending with a reliable dataset with genotypes of each sample.
The results of the presented method were derived from two sampling periods in a 4000 ha area in the Palatinate Forest in Rhineland-Palatinate in December 2006 and 2007. Both provided high confidence intervals (CI) applying inaccurate estimates (eg. for 2006 population size amounted to 215 with CI 95% of 156-314 and for 2007 population size amounted to 415 with CI 95% of 318-561) due to low sampling sizes (for 2006 n = 141 and for 2007 n = 326), successfully analysed samples (for 2006 n = 89 and for 2007 n = 156) and recapture numbers (for 2006 n = 12 and for 2007 n = 24). Furthermore, the population estimates even for the lowest values were considerably higher than previously assumed by hunting statistics, which implicates an ineffective hunting regime in the study area. For the future prospect, to obtain more precise population size estimations the increase of sampling sizes is inevitable, because absolute and reliable estimates are highly desirable for wildlife management and the control of diseases transmission. Nevertheless, the method for individual genotyping of wild boars evaluated in this thesis could be successfully established resulting in reliable datasets for population estimation modelling with sufficiently low GER.
Nandi forests (South and North Nandi forests) are situated in the Rift Valley Province of Kenya very close to Kakamega forest. From previous documents it has been seen that Kakamega and Nandi forests were connected to each other forming one big "U" shaped forest block till the beginnings of 1900s. Due to human pressures, currently there are three different forests form the previous one block forest. Although they were one forest, information on Nandi forests is very scanty when it is compared to that of Kakamega forest. The species composition and diversity as well as plant communities and population structure of Nandi forests have not been studied. Information is not available about the similarity status of South and North Nandi forests. Furthermore the natural regeneration potential (seedling bank) of these forests is not well studied and documented. Hence this study aims to fill these gaps.
In this study totally 76 quadrates (49 from South Nandi and 27 from North Nandi) were used to collect data. In the South Nandi forests 27 of the quadrates were laid in the better side of the forest (at Kobujoi) and the remaining 22 were in the heavily disturbed part of this forest (Bonjoge). The quadrates were arranged on transects that have one to one and half km which were parallel to the slope. The distance between the quadrates was 100 meter and transects are 500 m apart. The size of the main quadrate was 400 m2 (20 X 20 m) which also had five small plots (3 X 3 m) distributed on the four corners and in the center. Each woody plants (climbers, shrubs and trees) having more than one meter and greater than two centimeter diameter at breast height (dbh) were measured and recorded. Seedlings and herbaceous plants were sampled in the smaller plots. Individual plants were identified at species level and when it was not possible to identify in the field voucher specimen were prepared and latter identified at the East African Herbarium, National Museum of Kenya, and Nairobi. Clustering and ordination were performed using PC-ORD and CANOCO ecological softwares, respectively. For both clustering and ordination abundance data of the species was used. Shannon diversity index and evenness were computed using PC-ORD while similarity indices, Fisher alpha, rarefaction, species richness estimation (nonparametric species richness estimators) were conducted using EstimateS. Indicator species analysis was undertaken using PC-ORD. Basal area and height class distribution at forests level or site level (Bonjoge and Kobujoi) and diameter (dbh) class distribution for selected trees species were performed to evaluate population structure.
Furthermore importance value (IV) of woody plant species was calculated. SPSS version 16 was used to undertake both parametric (when data assume normal distribution) and nonparametric (when data are not assuming normal distribution) comparison of means, correlation and regression analysis.
In this study totally 321 vascular plant species comprising 92 families and 243 genera were identified in Nandi forests (both South and North Nandi forests). In South Nandi forest 253 plant species form 82 families and 201 genera were recorded while in North Nandi 181 species comprising 67 families and 155 genera were recorded. Jackknife second order estimators gave the highest species richness estimate for both South and North Nandi forests i.e. 284 and 209, respectively. In the case of highly disturbed and less disturbed parts of South Nandi forest 138 and 172 vascular plant species were recorded, respectively. Asteraceae, Rubiaceae and Euphorbiaceae are the top three species rich families of Nandi forests. In terms of different diversity measures (i.e. alpha and beta diversity, Fisher alpha, Shannon diversity and evenness indices) South Nandi is more diverse than North Nandi forest. Sörensen and Jaccard (classic) as well as their respective abundance based similarities showed that there is a low species similarity between South and Nandi forests. The cluster analysis resulted in three different plant communities and this result is supported by the ordination result.
South and North Nandi forest has inverted "J" height class distribution showing that larger proportion of woody plant individuals are found in the lower height classes. Similar pattern is observed when the diameters of all woody plants were considered together. However, different diameter class distributions (seven types) were identified when selected tree species were analyzed separately. It has been observed that the basal area of South Nandi forest is significantly lower than that of North Nandi forest (Mann-Whitney U =358, p < 0.001). Similarly Bonjoge has significantly lower basal area (t-value=3.77, p<0.01) than that of Kobujoi. Number of woody plat seedlings in South Nandi forest is significantly higher than that of North Nandi (Mann-Whitney U = 362.5, p<0.001). In the same way Bonjoge has significantly smaller number of ssedlings than Kobujoi (t-value 4.24, p<0.001). Most of species in both forests are able to resprout from stumps after physical damage; hence this helps the regeneration of the forests in addition to seedling banks. This study enables to fill some of the information gaps about Nandi forests especially of floristic composition, population structure, natural regeneration and human impacts on this ecosystem.
Based on dual process models of information processing, the present research addressed how explicit disgust sensitivity is re-adapted according to implicit disgust sensitivity via self-perception of automatic behavioral cues. Contrary to preceding studies (Hofmann, Gschwendner, & Schmitt, 2009) that concluded that there was a "blind spot" for self- but not for observer perception of automatic behavioral cues, in the present research, a re-adaption process was found for self-perceivers and observers. In Study 1 (N = 75), the predictive validity of an indirect disgust sensitivity measure was tested with a double-dissociation strategy. Study 2 (N = 117) reinvestigated the hypothesis that self-perception of automatic behavioral cues, predicted by an indirect disgust sensitivity measure, led to a re-adaption of explicit disgust sensitivity measures. Using a different approach from Hofmann et al. (2009), the self-perception procedure was modified by (a) feeding back the behavior several times while a small number of cues had to be rated for each feedback condition, (b) using disgust sensitivity as a domain with clearly unequivocal cues of automatic behavior (facial expression, body movements) and describing these cues unambiguously, and (c) using a specific explicit disgust sensitivity measure in addition to a general explicit disgust sensitivity measure. In Study 3 (N = 130), the findings of Study 2 were replicated and display rules and need for closure as moderator effects of predictive validity and cue utilization were additionally investigated. The moderator effects give hints that both displaying a disgusted facial expression and self-perception of one- own disgusted facial expression are subject to a self-serving bias, indicating that facial expression may not be an automatic behavior. Practical implications and implications for future research are discussed.
In this thesis the feasibility of a GPGPU (general-purpose computing on graphics processing units) approach to natural feature description on mobile phone GPUs is assessed. To this end, the SURF descriptor [4] has been implemented with OpenGL ES 2.0/GLSL ES 1.0 and evaluated across different mobile devices. The implementation is multiple times faster than a comparable CPU variant on the same device. The results proof the feasibility of modern mobile graphics accelerators for GPGPU tasks especially for the detection phase in natural feature tracking used in augmented reality applications. Extensive analysis and benchmarking of this approach in comparison to state of the art methods have been undertaken. Insights into the modifications necessary to adapt and modify the SURF algorithm to the limitations of a mobile GPU are presented. Further, an outlook for a GPGPU-based tracking pipeline on a mobile device is provided.
Population genetic structure in European Hyalodaphnia species: Monopolization versus gene flow
(2012)
Cyclic parthenogens displays an alternation of asexual and sexual reproduction which has consequences for the genetic structure of these organisms. The clonal diversity of cyclic parthenogenetic zooplankton populations is influenced by the size of the dormant egg bank, i.e., the amount of sexually produced dormant eggs that assembled in the sediment, as these dormant eggs contribute new genetic variants to the populations. Further, the clonal diversity is impacted by clonal erosion over time, which reduces the number of different clones through stochastic and selective processes. Although freshwater invertebrates are good dispersers through their dormant stages, the influence of gene flow is assumed to be negligible, as the local population successfully monopolizes the available resources. As these populations reach carrying capacity fast due to the asexual reproduction, the first colonizing individuals are able to successfully establish in the habitat, resulting in a priority effect which hinders the invasion of new genotypes. Due to clonal selection and sexual reproduction a population will locally adapt over time and will establish a dormant egg bank which facilitates the fast re-colonization after a hostile period. This thesis evaluates the processes altering the population genetic structure of cyclic parthenogenetic zooplankton with a special focus on the concepts of monopolization as well as the counteracting effects of gene flow, using large-lake Daphnia species. Thirty-two variable microsatellite DNA markers were developed and a subset of twelve markers was evaluated regarding their suitability for species assignment and hybrid class detection. With this marker set and an additional mitochondrial DNA marker forty-four natural European populations of the species D. cucullata, D. galeata and D. longispina were studied. In D. galeata, most populations were characterized by low clonal diversities which suggest high influence from clonal erosion over the growing season and a low contribution from the dormant egg bank. Further, recent expansions as well as gene flow were detected, probably caused by the anthropogenic alteration of freshwater habitats, in particular eutrophication of many European lakes. D. longispina and D. cucullata revealed a different genetic structure compared to D. galeata, with high genetic differentiation among populations. This indicates low levels of effective gene flow which is in line with the predictions of monopolization. Further, high clonal diversities were found in populations of the two taxa, suggesting a high contribution from the dormant egg bank while clonal erosion was often not detectable. In D. longispina, mitochondrial data revealed an ancient expansion which was probably initiated by the formation of glacial lakes after the last ice age.
In addition, in D. longispina not only clonal diversity but also genetic diversity was high, indicating that during the build-up of the studied populations the influence from gene flow was probably high. To better understand the processes that act on early populations the population build-up in regard to the temporal advantage of clones during invasion succession was experimentally studied and revealed that priority effects shape population structure of Daphnia species. However, in certain cases the highly superior clones resulted in the extinction of inferior clones independent of the temporal advantage the single clones had.
This clearly shows that not only the time of succession is important but also the competitive strength. rnIn conclusion, the results obtained show that the population genetic structure in cyclic parthenogenetic zooplankton species is impacted by various processes. In addition to earlier studies, which mainly focus on local adaptation, clonal erosion and the size of the dormant egg bank to understand population genetic structure, this thesis could show that gene flow may be effective as well. During population build-up the advantage of early arriving individuals does not necessarily predict the outcome of population assembly, as additional genotypes may contribute to the population. Finally, the genetic structure of established populations may be severely impacted by effective gene flow, if severe environmental changes alter the habitat of the locally adapted population.
Modern Internet and Intranet techniques, such as Web services and virtualization, facilitate the distributed processing of data providing improved flexibility. The gain in flexibility also incurs disadvantages. Integrated workflows forward and distribute data between departments and across organizations. The data may be affected by privacy laws, contracts, or intellectual property rights. Under such circumstances of flexible cooperations between organizations, accounting for the processing of data and restricting actions performed on the data may be legally and contractually required. In the Internet and Intranet, monitoring mechanisms provide means for observing and auditing the processing of data, while policy languages constitute a mechanism for specifying restrictions and obligations.
In this thesis, we present our contributions to these fields by providing improvements for auditing and restricting the data processing in distributed environments. We define formal qualities of auditing methods used in distributed environments. Based on these qualities, we provide a novel monitoring solution supporting a data-centric view on the distributed data processing. We present a solution for provenance-aware policies and a formal specification of obligations offering a procedure to decide whether obligatory processing steps can be met in the future.
Technical products have become more than practical tools to us. Mobile phones, for example, are a constant companion in daily life. Besides purely pragmatic tasks, they fulfill psychological needs such as relatedness, stimulation, competence, popularity, or security. Their potential for the mediation of positive experience makes interactive products a rich source of pleasure. Research acknowledged this: in parallel to the hedonic/utilitarian model in consumer research, Human-Computer Interaction (HCI) researchers broadened their focus from mere task-fulfillment (i.e., the pragmatic) to a holistic view, encompassing a product's ability for need-fulfillment and positive experience (i.e., the hedonic). Accordingly, many theoretical models of User Experience (UX) acknowledge both dimensions as equally important determinants of a product's appeal: pragmatic attributes (e.g., usability) as well as hedonic attributes (e.g., beauty). In choice situations, however, people often overemphasize the pragmatic, and fail to acknowledge the hedonic. This phenomenon may be explained by justification. Due to their need for justification, people attend to the justifiability of hedonic and pragmatic attributes rather than to their impact on experience. Given that pragmatic attributes directly contribute to task-fulfillment, they are far easier to justify than hedonic attributes. People may then choose the pragmatic over the hedonic, despite a true preference for the hedonic. This can be considered a dilemma, since people choose what is easy to justify and not what they enjoy the most. The present thesis presents a systematic exploration of the notion of a hedonic dilemma in the context of interactive products.
A first set of four studies explored the assumed phenomenon. Study 1 (N = 422) revealed a reluctance to pay for a hedonic attribute compared to a pragmatic attribute. Study 2 (N = 134) demonstrated that people (secretly) prefer a more hedonic product, but justify their choice by spurious pragmatic advantages. Study 3 (N = 118) confronted participants with a trade-off between hedonic and pragmatic quality. Even though the prospect of receiving a hedonic product was related to more positive affect, participants predominantly chose the pragmatic, especially those with a high need for justification. This correlation between product choice and perceived need for justification lent further support to the notion that justification lies at the heart of the dilemma. Study 4 (N = 125) explored affective consequences and justifications provided for hedonic and pragmatic choice. Data on positive affect suggested a true preference for the hedonic - even among those who chose the pragmatic product.
A second set of three studies tested different ways to reduce the dilemma by manipulating justification. Manipulations referred to the justifiability of attributes as well as the general need for justification. Study 5 (N = 129) enhanced the respective justifiability of hedonic and pragmatic choice by ambiguous product information, which could be interpreted according to latent preferences. As expected, enhanced justifiability led to an increase in hedonic but not in pragmatic choice. Study 6 (N = 178) manipulated the justifiability of hedonic choice through product information provided by a "test report", which suggested hedonic attributes as legitimate. Again, hedonic choice increased with increased justifiability. Study 7 (N = 133) reduced the general need for justification by framing a purchase as gratification. A significant positive effect of the gratification frame on purchase rates occurred for a hedonic but not for a pragmatic product.
Altogether, the present studies revealed a desire for hedonic attributes, even in interactive products, which often are still understood as purely pragmatic "tools". But precisely because of this predominance of pragmatic quality, people may hesitate to give in to their desire for hedonic quality in interactive products - at least, as long as they feel a need for justification. The present findings provide an enhanced understanding of the complex consequences of hedonic and pragmatic attributes, and indicate a general necessity to expand the scope of User Experience research to the moment of product choice. Limitations of the present studies, implications for future research as well as practical implications for design and marketing are discussed.