Refine
Has Fulltext
- yes (46) (remove)
Is part of the Bibliography
- yes (46)
Year of publication
- 2010 (46) (remove)
Document Type
- Doctoral Thesis (46) (remove)
Keywords
- Taufliege (9)
- Drosophila melanogaster (4)
- Signaltransduktion (4)
- Assoziatives Gedächtnis (3)
- Bioinformatik (3)
- Bruchpilot (3)
- Drosophila (3)
- Genexpression (3)
- Genmutation (3)
- Knochen-Morphogenese-Proteine (3)
Institute
- Theodor-Boveri-Institut für Biowissenschaften (46) (remove)
Foraging behavior is a particularly fascinating topic within the studies of social insects. Decisions made by individuals have effects not only on the individual level, but on the colony level as well. Social information available through foraging in a group modulates individual preferences and shapes the foraging pattern of a colony. Identifying parameters influencing foraging behavior in leaf-cutting ants is especially intriguing because they do not harvest for themselves, but for their symbiotic fungus which in turn influences their plant preferences after the incorporation of the substrate. To learn about the substrates’ unsuitability for the fungus, ants need to be able to identify the incorporated substrate and associate it with detrimental effects on the fungus. Odor is an important plant characteristic known to be used as recognition key outside the nest in the context of foraging. Chapter 1 shows that foragers are able to recall information about the unsuitability of a substrate through odor alone and consequently reject the substrate, which leads to the conclusion that inside the nest, odor might be enough to indentify incorporated substrate. Identification of plant species is a key factor in the foraging success of leaf-cutting ants as they harvest a multitude of different plant species in a diverse environment and host plant availability and suitability changes throughout the year. Fixed plant preferences of individuals through innate tendencies are therefore only one factor influencing foraging decisions. On the individual as well as the colony level, foraging patterns are flexible and a result of an intricate interplay between the different members involved in the harvesting process: foragers, gardeners and the symbiotic fungus. In chapter 2 I identified several conditions necessary for naïve foragers to learn about the unsuitability of substrate inside the nest. In order to exchange of information about the unsuitability of a substrate, the plant in question must be present in the fungus garden. Foragers can learn without own foraging experience and even without experiencing the effects of the substrate on the fungus, solely through the presence of experienced gardeners. The presence of experienced foragers alone on the other hand is not enough to lower the acceptance of substrate by naïve foragers in the presence of naïve gardeners, even if experienced foragers make up the majority of the workforce inside the nest. Experienced foragers are also able to reverse their previous negative experience and start accepting the substrate again. The individual behavior of foragers and gardeners with different experiential backgrounds in the presence of suitable or unsuitable substrate inside the fungus chamber was investigated in chapter 3 to shed some light on possible mechanisms involved in the flow of information about substrate suitability from the fungus to the ants. Gardeners as well as foragers are involved in the leaf processing and treatment of the applied leaf patches on the fungus. If the plant material is unsuitable, significantly more ants treat the plant patches, but foragers are less active overall. Contacts between workers initiated by either gardeners or foragers occur significantly more frequent and last longer if the substrate is unsuitable. Even though experienced gardeners increase naïve foragers’ contact rates and duration with other workers in the presence of suitable plant patches, naïve foragers show no differences in the handling of the plant patches. This suggests that foragers gain information about plant suitability not only indirectly through the gardening workers, but might also be able to directly evaluate the effects of the substrate on the fungus themselves. Outside the nest, foragers influence each other the trail (chapter 4). Foraging in a group and the presence of social information is a decisive factor in the substrate choice of the individual and leads to a distinct and consentaneous colony response when encountering unfamiliar or unsuitable substrates. As leaf-cutting ants harvest different plant species simultaneously on several trails, foragers gain individual experiences concerning potential host plants. Preferences might vary among individuals of the same colony to the degree that foragers on the same trail perceive a certain substrate as either suitable or unsuitable. If the majority of foragers on the trail perceives one of the currently harvested substrates as unsuitable, naïve foragers lower their acceptance within 4 hours. In the absence of a cue in the fungus, naïve foragers harvesting by themselves still eventually (within 6 hours) reject the substrate as they encounter experienced gardeners during visits to the nest within foraging bouts. As foraging trails can be up to 100 m long and foragers spend a considerable amount of time away from the nest, learning indirectly from experienced foragers on the trail accelerates the distribution of information about substrate suitability. The level of rejection of a formerly unsuitable substrate after eight hours of foraging by naïve foragers correlates with the average percentage of unladen experienced foragers active on the trail. This suggests that unladen experienced foragers might actively contact laden naïve workers transmitting information about the unsuitability of the load they carry. Results from experiments were I observed individual laden foragers on their way back to the nest backed up this assumption as individuals were antennated and received bites into the leaf disk they carried. Individuals were contacted significantly more often by nestmates that perceived the carried leaf disk as unsuitable due to previous experience than by nestmates without this experience (chapter 6). Leaf-cutting ants constantly evaluate, learn and re-evaluate the suitability of harvested substrate and adjust their foraging activity accordingly. The importance of the different sources of information within the colony and their effect on the foraging pattern of the colony depend on the presence or absence of each of them as e.g. experienced foragers have a bigger influence on the plant preferences of naïve foragers in the absence of a cue in the fungus garden.
Applying microarray‐based techniques to study gene expression patterns: a bio‐computational approach
(2010)
The regulation and maintenance of iron homeostasis is critical to human health. As a constituent of hemoglobin, iron is essential for oxygen transport and significant iron deficiency leads to anemia. Eukaryotic cells require iron for survival and proliferation. Iron is part of hemoproteins, iron-sulfur (Fe-S) proteins, and other proteins with functional groups that require iron as a cofactor. At the cellular level, iron uptake, utilization, storage, and export are regulated at different molecular levels (transcriptional, mRNA stability, translational, and posttranslational). Iron regulatory proteins (IRPs) 1 and 2 post-transcriptionally control mammalian iron homeostasis by binding to iron-responsive elements (IREs), conserved RNA stem-loop structures located in the 5’- or 3‘- untranslated regions of genes involved in iron metabolism (e.g. FTH1, FTL, and TFRC). To identify novel IRE-containing mRNAs, we integrated biochemical, biocomputational, and microarray-based experimental approaches. Gene expression studies greatly contribute to our understanding of complex relationships in gene regulatory networks. However, the complexity of array design, production and manipulations are limiting factors, affecting data quality. The use of customized DNA microarrays improves overall data quality in many situations, however, only if for these specifically designed microarrays analysis tools are available. Methods In this project response to the iron treatment was examined under different conditions using bioinformatical methods. This would improve our understanding of an iron regulatory network. For these purposes we used microarray gene expression data. To identify novel IRE-containing mRNAs biochemical, biocomputational, and microarray-based experimental approaches were integrated. IRP/IRE messenger ribonucleoproteins were immunoselected and their mRNA composition was analysed using an IronChip microarray enriched for genes predicted computationally to contain IRE-like motifs. Analysis of IronChip microarray data requires specialized tool which can use all advantages of a customized microarray platform. Novel decision-tree based algorithm was implemented using Perl in IronChip Evaluation Package (ICEP). Results IRE-like motifs were identified from genomic nucleic acid databases by an algorithm combining primary nucleic acid sequence and RNA structural criteria. Depending on the choice of constraining criteria, such computational screens tend to generate a large number of false positives. To refine the search and reduce the number of false positive hits, additional constraints were introduced. The refined screen yielded 15 IRE-like motifs. A second approach made use of a reported list of 230 IRE-like sequences obtained from screening UTR databases. We selected 6 out of these 230 entries based on the ability of the lower IRE stem to form at least 6 out of 7 bp. Corresponding ESTs were spotted onto the human or mouse versions of the IronChip and the results were analysed using ICEP. Our data show that the immunoselection/microarray strategy is a feasible approach for screening bioinformatically predicted IRE genes and the detection of novel IRE-containing mRNAs. In addition, we identified a novel IRE-containing gene CDC14A (Sanchez M, et al. 2006). The IronChip Evaluation Package (ICEP) is a collection of Perl utilities and an easy to use data evaluation pipeline for the analysis of microarray data with a focus on data quality of custom-designed microarrays. The package has been developed for the statistical and bioinformatical analysis of the custom cDNA microarray IronChip, but can be easily adapted for other cDNA or oligonucleotide-based designed microarray platforms. ICEP uses decision tree-based algorithms to assign quality flags and performs robust analysis based on chip design properties regarding multiple repetitions, ratio cut-off, background and negative controls (Vainshtein Y, et al., 2010).
Fish of the genus Xiphophorus belong to the oldest animal models in cancer research. The oncogene responsible for the generation of spontaneous aggressive melanoma encodes for a mutated epidermal growth factor receptor (Egfr) and is called xmrk for Xiphophorus melanoma receptor kinase. Xmrk constitutive activation mechanisms and subsequent signaling pathways have already been investigated and charaterized but it is still unknown if Egfr ligands may also play a role in Xmrk-driven melanoma formation. To investigate the potential role of Egfr ligands in Xmrk-driven melanoma, I firstly analyzed the evolution of teleost and tetrapod Egfr/Egfr ligand systems. I especially focused on the analysis on the medaka fish, a closely related species to Xiphophorus, for which the whole genome has been sequenced. I could identify all seven Egfr ligands in medaka and could show that the two teleost-specific Egfr copies of medaka display dissimilar expression patterns in adult tissues together with differential expression of Egfr ligand subsets, arguing for subfunctionalization of receptor functions in this fish. Our phylogenetic and synteny analyses supported the hypothesis that only one gene in the chordate ancestor gave rise to the diversity of Egfr ligands found in vertebrate genomes today. I also could show that the Egfr extracellular subdomains implicated in ligand binding are not evolutionary conserved between tetrapods and teleosts, making the use of heterologous ligands in experiments with fish cells debatable. Despite its well understood and straight-forward process, Xmrk-driven melanomagenesis in Xiphophorus is problematic to further investigate in vivo. Our laboratory recently established a new melanoma animal model by generating transgenic mitf::xmrk medaka fishes, a Xiphophorus closely related species offering many more advantages. These fishes express xmrk under the control of the pigment-cell specific Mitf promoter. During my PhD thesis, I participated in the molecular analysis of the stably transgenic medaka and could show that the Xmrk-induced signaling pathways are similar when comparing Xiphophorus with transgenic mitf::xmrk medaka. These data together with additional RNA expression, protein, and histology analyses showed that Xmrk expression under the control of a pigment cell-specific promoter is sufficient to induce melanoma in the transgenic medaka, which develop very stereotyped tumors, including uveal and extracutaneous melanoma, with early onset during larval stages. To further investigate the potential role of Egfr ligands in Xmrk-driven melanoma, I made use of two model systems. One of them was the above mentioned mitf::xmrk medaka, the other was an in-vitro cell culture system, where the EGF-inducible Xmrk chimera HERmrk is stably expressed in murine melanocytes. Here I could show that HERmrk activation strongly induced expression of amphiregulin (Areg) and heparin-binding EGF-like growth factor (Hbegf) in melanocytes. This regulation was dependent on the MAPK and SRC signaling pathways. Moreover, upregulation of Adam10 and Adam17, the two major sheddases of Egfr ligands, was observed. I also could demonstrate the functionality of the growth factors by invitro analyses. Using the mitf::xmrk medaka model I could also show the upregulation of a subset of ligand genes, namely egf, areg, betacellulin (btc) and epigen (epgn) as well as upregulation of medaka egfrb in tumors from fish with metastatic melanoma. All these results converge to support an Xmrk-induced autocrine Egfr ligand loop. Interestingly, my in-vitro experiments with conditioned supernatant from medaka Egf- and Hbegf-producing cells revealed that not only Xiphophorus Egfrb, but also the pre-activated Xmrk could be further stimulated by the ligands. Altogether, I could show with in-vitro and in-vivo experiments that Xmrk is capable of inducing a functional autocrine Egfr ligand loop. These data confirm the importance of autocrine loops in receptor tyrosine kinase (RTK)-dependent cancer development and show the possibility for a constitutively active RTK to strengthen its oncogenic signaling by ligand binding.
In our analysis I was interested in the gene duplications, with focus on in-paralogs. In-paralogs are gene duplicates which arose after species split. Here I analysed the in-paralogs quantitatively, as well as qualitatively. For quantitative analysis genomes of 21 species were taken. Most of them have vastly different lifestyles with maximum evolutionary distance between them 1100 million years. Species included mammals, fish, insects and worm, plus some other chordates. All the species were pairwised analysed by the Inparanoid software, and in-paralogs matrix were built representing number of in-paralogs in all vs. all manner. Based on the in-paralogs matrix I tried to reconstruct the evolutionary tree using in-paralog numbers as evolutionary distance. If all 21 species were used the resulting tree was very far from real one: a lot of species were misplaced. However if the number was reduced to 12, all of the species were placed correctly with only difference being wrong insect and fish clusters switched. Then to in-paralogs matrix the neighbour-net algorithm was applied. The resulting "net" tree showed the species with fast or slow duplications rates compared to the others. We could identify species with very high or very low duplications frequencies and it correlates with known occurrences of the whole genome duplications. As the next step I built the graphs for every single species showing the correlation between their in-paralogs number and evolutionary distance. As we have 21 species, graph for every species is built using 20 points. Coordinates of the points are set using the evolutionary distance to that particular species and in-paralogs number. In mammals with increasing the distance from speciation the in-paralogs number also increased, however not in linear fashion. In fish and insects the graph close to zero is just the same in mammals' case. However, after reaching the evolutionary distances more than 800 million years the number of inparalogs is beginning to decrease. We also made a simulation of gene duplications for all 21 species and all the splits according to the fossil and molecular clock data from literature. In our simulation duplication frequency was minimal closer to the past and maximum in the near-present time. Resulting curves had the same shape the experimental data ones. In case of fish and insect for simulation the duplication rate coefficient even had to be set negative in order to repeat experimental curve shape. To the duplication rate coefficient in our simulation contribute 2 criteria: gene duplications and gene losses. As gene duplication is stochastical process it should always be a constant. So the changing in the coefficient should be solely explained by the increasing gene loss of old genes. The processes are explained by the evolution model with high gene duplication and loss ratio. The drop in number of in-paralogs is probably due to the BLAST algorithm. It is observed in comparing highly divergent species and BLAST cannot find the orthologs so precisely anymore. In the second part of my work I concentrated more on the specific function of inparalogs. Because such analysis is time-consuming it could be done on the limited number species. Here I used three insects: Drosophila melanogaster (fruit y), Anopheles gambiae (mosquito) and Apis mellifera (honeybee). After Inparnoid analyses and I listed the cluster of orthologs. Functional analyses of all listed genes were done using GO annotations and also KEGG PATHWAY database. We found, that the gene duplication pattern is unique for each species and that this uniqueness is rejected through the differences in functional classes of duplicated genes. The preferences for some classes reject the evolutionary trends of the last 350 million years and allow assumptions on the role of those genes duplications in the lifestyle of species. Furthermore, the observed gene duplications allowed me to find connections between genomic changes and their phenotypic manifestations. For example I found duplications within carbohydrate metabolism rejecting feed pattern adaptation, within photo- and olfactory-receptors indicating sensing adaptation and within troponin indicating adaptations in the development. Despite these species specific differences, found high correlations between the independently duplicated genes between the species. This might hint for a "pool" of genes preferentially duplicated. Taken together, the observed duplication patterns reject the adaptational process and provide us another link to the field of genomic zoology.
Precise control of mitotic progression is vital for the maintenance of genomic integrity. Since the loss of genomic integrity is known to promote tumorigenesis, the identification of knew G2/M regulatory genes attracts great attention. LINC, a human multiprotein complex, is a transcriptional activator of a set of G2/M specific genes. By depleting LIN9 in MEFs, a core subunit of LINC, Gas2l3 was identified as a novel LINC target gene. The so far uncharacterized Gas2l3 gene encodes for a member of the family of growth arrest specific 2 (GAS2) proteins, which share a highly conserved putative actin binding CH and a putative microtubule binding GAS2 domain. In the present study GAS2L3 was identified as a LINC target gene also in human cells. Gene expression analysis revealed that GAS2L3 transcription, in contrast to all other GAS2 family members, is highly regulated during the cell cycle with highest expression in G2/M. The GAS2L3 protein showed a specific localization pattern during the M phase: In metaphase, GAS2L3 localized to the mitotic spindle, relocated to the spindle midzone microtubules in late anaphase and concentrated at the midbody in telophase where it persisted until the end of cytokinesis. Overexpression of a set of different GAS2L3 deletion mutants demonstrated that the localization to the mitotic microtubule network is dependent on the C-terminus, whereas the midbody localization is dependent on full length GAS2L3 protein. Additionally, exclusive overexpression of the CH domain induced the formation of actin stress fibers, suggesting that the CH domain is an actin binding domain. In contrast, the GAS2 domain was neither needed nor sufficient for microtubule binding, indicating that there must be an additional so far unknown microtubule binding domain in the C-terminus. Interestingly, immunoblot analysis also identified the C-terminus as the domain responsible for GAS2L3 protein instability, partially dependent on proteasomal degradation. Consistent with its specific localization pattern, GAS2L3 depletion by RNAi demonstrated its responsibility for proper mitosis and cytokinesis. GAS2L3 depletion in HeLa cells resulted in the accumulation of multinucleated cells, an indicator for chromosome mis-segregation during mitosis. Also the amount of cells in cytokinesis was enriched, indicating failures in completing the last step of cytokinesis, the abscission. Strikingly, treatment with microtubule poisons that lead to the activation of the spindle assembly checkpoint (SAC) indicated that the SAC was weakened in GAS2L3 depleted cells. Although the exact molecular mechanism is still unknown, fist experiments support the hypothesis that GAS2L3 might be a regulator of the SAC master kinase BUBR1. In conclusion, this study provides first evidence for GAS2L3 as a novel regulator of mitosis and cytokinesis and it might therefore be an important guardian against tumorigenesis.
Protein phosphatases can be classified into at least three major families based on amino acid sequences at their active sites. A newly emerging phosphatase family contains the active site sequence DXDX(T/V), and belongs to the haloacid dehalogenase (HAD) superfamily of hydrolases, a ubiquitous and evolutionarily conserved enzyme family. Although the existence of 58 human HAD enzymes has been predicted by database analysis, our understanding of their biological functions remains rudimentary.By database mining amd phylogenetic analysis of human HAD phosphatases, we have found a marked increase in cell area of spreading cells, as well as accelerated cell spreading onfibronectin. Taken together, we have identified and characterized AUM as a novel member of the emerging family of aspartate-dependent protein tyrosine phosphatases. Our findings implicate AUM as an important regulator of Src-dependent cytoskeletal dynamics during cell adhesion and migration. a previously unidentified enzyme with homology to Chronophin, a cytoskeletal regulatory HAD phosphatase. We have cloned and characterized this novel enzyme and named it AUM,for actin remodeling, ubiquitously expressed, magnesium-dependent HAD phosphatase. By Northern blot, real-time PCR and Western blot analysis, we show that AUM is broadly expressed in all major human and mouse tissues with highest levels found in testis. Using immunohistochemistry, we can show that AUM is specifically expressed in maturing germ cells and that its expression peaks during spermiogenesis. To characterize the substrate preference of AUM, we have conducted an in vitro phosphatase substrate screen with 720 phosphopeptides derived from human phosphorylation sites. AUM exclusively dephosphorylates phosphotyrosine (pTyr)-containing peptides. Furthermore, only 17 pTyr peptides (~2% of all pTyr peptides investigated) acted as AUM substrates, indicating a high degree of substrate specificity. Putative AUM substrates include proteins involved in cytoskeletal dynamics and tyrosine kinase signaling.In accordance with the phosphopeptide screen, phosphatase overlay assays employing whole-cell extracts of pervanadate-treated HeLa cells show that AUM dephosphorylates only a limited number of tyrosyl-phosphorylated proteins.The role of AUM for cellular signaling was investigated in response to epidermal growth factor (EGF) stimulation in a spermatogonial cell line (GC-1 spg). The overexpression of AUM reduces, whereas the RNAi-mediated depletion of endogenous AUM increases EGF inducedtyrosine phosphorylation, including changes in the phosphorylation of the EGF receptor itself. Interestingly, in vitro kinase/phosphatase assays with purified Src and AUM indicate that AUM can activate Src, which in turn phosphorylates and inactivates AUM. Although it is at present unclear how Src and AUM regulate each other, our initial findings suggests that AUM enhances Src kinase activity independently of its phosphatase activity, whereas Src diminishes AUM phosphatase activity in a kinase dependent manner. On a cellular level, AUM-depleted cells are characterized by altered actin cytoskeletal dynamics and adhesion, as indicated by stabilized actin filaments, enlarged focal adhesions,a marked increase in cell area of spreading cells, as well as accelerated cell spreading on fibronectin. Taken together, we have identified and characterized AUM as a novel member of the emerging family of aspartate-dependent protein tyrosine phosphatases. Our findings implicate AUM as an important regulator of Src-dependent cytoskeletal dynamics during cell adhesion and migration.
Biodiversity may be investigated and explored by the means of genetic sequence information and molecular phylogenetics. Yet, with ribosomal genes, information for phylogenetic studies may not only be retained from the primary sequence, but also from the secondary structure. Software that is able to cope with two dimensional data and designed to answer taxonomic questions has been recently developed and published as a new scientific pipeline. This thesis is concerned with expanding this pipeline by a tool that facialiates the annotation of a ribosomal region, namely the ITS2. We were also able to show that this states a crucial step for secondary structure phylogenetics and for data allocation of the ITS2-database. This resulting freely available tool determines high quality annotations. In a further study, the complete phylogenetic pipeline has been evaluated on a theoretical basis in a comprehensive simulation study. We were able to show that both, the accuracy and the robustness of phylogenetic trees are largely improved by the approach. The second major part of this thesis concentrates on case studies that applied this pipeline to resolve questions in taxonomy and ecology. We were able to determine several independent phylogenies within the green algae that further corroborate the idea that secondary structures improve the obtainable phylogenetic signal, but now from a biological perspective. This approach was applicable in studies on the species and genus level, but due to the conservation of the secondary structure also for investigations on the deeper level of taxonomy. An additional case study with blue butterflies indicates that this approach is not restricted to plants, but may also be used for metazoan phylogenies. The importance of high quality phylogenetic trees is indicated by two ecological studies that have been conducted. By integrating secondary structure phylogenetics, we were able to answer questions about the evolution of ant-plant interactions and of communities of bacteria residing on different plant tissues. Finally, we speculate how phylogenetic methods with RNA may be further enhanced by integration of the third dimension. This has been a speculative idea that was supplemented with a small phylogenetic example, however it shows that the great potential of structural phylogenetics has not been fully exploited yet. Altogether, this thesis comprises aspects of several different biological disciplines, which are evolutionary biology and biodiversity research, community and invasion ecology as well as molecular and structural biology. Further, it is complemented by statistical approaches and development of informatical software. All these different research areas are combined by the means of bioinformatics as the central connective link into one comprehensive thesis.
All animals learn in order to cope with challenges imposed on them by their environment. This is true also for both larval and adult fruit flies as exemplified in pavlovian conditioning. The focus of this Thesis is on various aspects of the fruit flies learning ability. My main project deals with two types of learning which we call punishment-learning and pain-relief learning. Punishment learning happens when fruit flies are exposed to an odour which is followed by electric shock. After such training, flies have learned that that odour signals pain and consequently will avoid it in the future. If the sequence of the two stimuli is reversed such that odour follows shock, flies learn the odour as a signal for relief and will later on approach it. I first report a series of experiments investigating qualitative and parametric features of relief-learning; I find that (i) relief learning does result from true associative conditioning, (ii) it requires a relatively high number of training trials, (iii) context-shock training is ineffective for subsequent shock-odour learning. A further question is whether punishment-learning and pain-relief learning share genetic determinants. In terms of genetics, I test a synapsin mutant strain, which lacks all Synapsin protein, in punishment and relief-learning. Punishment learning is significantly reduced, and relief-learning is abolished. Pan-neuronal RNAi-mediated knock-down of Synapsin results in mutant-like phenotypes, confirming the attribution of the phenotype to lack of Synapsin. Also, a rescue of Synapsin in the mushroom body of syn97 mutants restores both punishment- and relief-learning fully, suggesting the sufficiency of Synapsin in the mushroom body for both these kinds of learning. I also elucidate the relationship between perception and physiology in adult fruit flies. I use odour-shock conditioning experiments to identify degrees of similarity between odours; I find that those similarity measures are consistent across generalization and discrimination tasks of diverse difficulty. Then, as collaborator of T. Völler and A. Fiala, I investigate how such behavioural similarity/dissimilarity is reflected at the physiological level. I combine the behaviour data with calcium imaging data obtained by measuring the activity patterns of those odours in either the sensory neurons or the projection neurons at the antennal lobe. Our interpretation of the results is that the odours perceptual similarity is organized by antennal lobe interneurons. In another project I investigate the effect of gustatory stimuli on reflexive behaviour as well as their role as reinforcer in larval learning. Drosophila larvae greatly alter their behaviour in presence of sodium chloride. Increasing salt concentration modulates choice behaviour from weakly appetitive to strongly aversive. A similar concentration-behaviour function is also found for feeding: larval feeding is slightly enhanced in presence of low salt concentrations, and strongly decreased in the presence of high salt concentrations. Regarding learning, relatively weak salt concentrations function as appetitive reinforcer, whereas high salt concentrations function as aversive reinforcer. Interestingly, the behaviour-concentration curves are shifted towards higher concentrations from reflexive behaviour (choice behaviour, feeding) as compared to associative learning. This dissociation may reflect a different sensitivity in the respective sensory-motor circuitry.
Termites are the most important soil ecosystem engineers of semi‐arid and arid habitats. They enhance decomposition processes as well as the subsequent mineralisation of nutrients by bacteria and fungi. Through their construction of galleries, nests and mounds, they promote soil turnover and influence the distribution of nutrients and also alter texture and hydrological properties of soils, thereby affecting the heterogeneity of their ecosystem. The main aim of the present thesis was to define the impact of termites on ecosys‐tem functioning in a semi‐arid ecosystem. In a baseline study, I assessed the diversity of termite taxa in relation to the amount of precipitation, the vegetation patterns and the land use systems at several sites in Namibia. Subsequently, I focussed on a species that is highly abundant in many African savannas, the fungus growing and mound building species Macro‐termes michaelseni (Sjöstedt, 1914). I asked how this species influences the spatial hetero‐geneity of soil and vegetation patterns. From repeated samplings at 13 sites in Namibia, I obtained 17 termite taxa of 15 genera. While the type of land use seems to have a minor effect on the termite fauna, the mean annual precipitation explained 96% and the Simpson index of vascular plant diversity 81% of the variation in taxa diversity. The number of termite taxa increased with both of these explanation variables. In contrast to former studies on Macrotermes mounds in several regions of Africa that I reviewed, soil analyses from M. michaelseni mounds in the central Namibian savanna revealed that they contain much higher nitrogen contents when compared to their parent material. Further analyses revealed that nitrate forms a major component of the nitrogen content in termite mounds. As nitrate solves easily in water, evaporation processes are most probably responsible for the transport of solved nitrates to the mound surface and their accumulation there. The analysed mounds in central Namibia contained higher sand propor‐tions compared to the mounds of the former studies. Through the higher percentage of coarse and middle sized pores, water moves more easily in sandy soils compared to more clayey soils. In consequence, evaporation‐driven nitrate accumulation can occur in the studied mounds at high rates. Hochgerechnet auf den Gesamtumfang der Hügel bedeckte das pro Jahr von einem bewohnten Hügel erodierte Material theoretisch einen 1 m breiten Kreisring um den Schwemmkegel des Hügels 2,4 mm hoch. Der entsprechende Wert für unbewohnte Hügel betrug 1,0 mm. To assess the amount of soil that erodes from termite mounds, I fastened four strong, 65 cm wide plastic bags at 14 mounds each and collected the soil that eroded during five rainfall events. Projected to the total mound circumference, the amount of soil eroded covers theoretically a 1 m wide circular ring around the pediment of an inhabited mound up to a height of 2.4 mm per year. For uninhabited mounds, the height of this soil layer would be 1.0 mm. Per hectare, roughly 245 kg eroded per year from the mounds. However, as the erosion rate depends on several factors such as rainfall intensity, soil texture and point of time within the rainy season, this is only a vague estimate. In order to determine up to which distance the soil erosion from the mounds still influences the chemical characteristics of the adjacent topsoil, I took samples from depth of 0–10 cm at 1, 5 and 25 m distances, respectively, from four different mounds and from the mounds themselves. The non‐metric multidimensional scaling of the soil properties showed strong differences between mound and off‐mound samples. Soil characteristics within the samples from the mounds did not differ largely. Similarly, I found no strong differences between the samples taken from the different distances from the mound. From these results I conclude that through the construction of foraging galleries and sheetings (soil constructions with which some termite species cover their food items), the soil eroding from termite mounds is quickly mixed with deeper soil layers. In consequence, mound material does not accumulate in the mound’s vicinity. In order to reveal how plant growth is influenced by termite mound material, we assessed the number of grass and herb individuals as well as the biomass of plants growing in situ on the base of mounds compared to adjacent sites. While the numbers of both grass and herb individuals were significantly lower compared to adjacent sites, the total biomass of plants growing on the base of mounds was significantly higher. Reverse results were obtained by pot experiments with radish (Raphanus sativus subsp. sativus) and sorghum (Sorghum sp.) growth. Both species grew significantly weaker on mound soil compared to adjacent soil. The contradictory results concerning the biomass of in situ and pot experi‐ments are most probably caused by the disturbance of the original soil structure during the potting process. The material was subsequently compacted through watering the plants. In contrast, Macrotermes mounds are pervaded by many macropores which seem to be essential for the plant roots to penetrate the soil. In the last part of this thesis, I posed the question how mounds of M. michaelseni are distributed and what factors might be responsible for this pattern. Former studies showed that mound size is correlated with the size of its inhabiting colony. With several multi‐scale analyses, I revealed that larger inhabited mounds were regularly distributed. Additionally, mounds which were closer together tended to be smaller than on average. This indicates that intraspecific competition controls the distribution and size of colonies and their mounds. Former studies concerning Odontotermes mounds substantiated that they are local hotspots of primary productivity and animal abundance. Based on these findings, simulations revealed that a regular distribution of these mounds leads to a greater ecosystem‐wide productivity compared to a random arrangement. As in the present study, plant biomass was higher at the mounds compared to off‐mound sites, this might hold true for M. michaelseni mounds. From the results of this thesis, I draw the conclusion that through their mound building activities, M. michaelseni strongly influences the distribution patterns of soil nutrients within the central Namibian savanna. These termites create sharp contrasts in nutrient levels and vegetation patterns between mound soils and off‐mound soils and enhance the heterogeneity of their habitats. Former studies revealed that habitat hetero‐geneity is important in generating species diversity and species richness in turn is correlated positively with biomass production and positively affects ecosystem services. In conclusion, the present thesis underlines the importance of M. michaelseni for ecosystem functioning of the central Namibian savanna.