This dissertation investigates the biomechanical properties of the diatoms, a large group of single celled algae which, due to their high species diversity and percentage of global production, have a key position in marine food-webs. A unique characteristic of these autotrophic eukaryotes is their siliceous cell wall, the frustule. These frustules im-pact marine processes in multiple ways: the biogeochemical silicon cycle on a global scale, co-evolutionary processes in geological time scales and ecological processes in various spatial and temporal scales, from the individual level in predator-prey interac-tions and by bloom formation and mass sedimentation events on larger scales. The biomechanical properties of the frustules and the interaction between their me-chanical strength and the species’ mortality in grazing experiments - as shown in publica-tion III - is the central topic in this thesis. The results confirm a defensive function of diatom frustules, as the results from micromanipulator-based stability analysis and feed-ing experiments including North-Sea diatoms and copepods showed for the first time an inverse correlation between mechanical strength of diatom frustules and mortality due to grazing. In addition, it shows that diatoms with a principally similar geometry differ widely in their mechanical properties. Another finding is that large size can further in-crease the defensive potential of the diatom frustule. While the overall construction principles of diatom frustules are species-specific, re-cent work on predator-induced reinforcement of the frustules of small diatoms provided the idea for an experiment that investigated the heavily silicified frustules of large and complex diatoms in the presence of predator emitted chemical cues. Manuscript IV hy-pothesized a possible variation in the silica content and a resulting plasticity in the frus-tules of Coscinodiscus wailesii and Actinoptychus senarius under such circumstances. The experimental setup made use of novel 3D techniques to analyze the structure of the sampled diatoms and computer based algorithms for automated comparison of micro-scopic structures. Although the initial hypothesis could not be supported, the results contribute a valuable facet to the discussion of induced defense systems in diatoms. The ecological aspects addressed in this thesis were based on investigations on func-tional morphology, especially the biomechanical properties, of the diatom frustules. Since classical biological methods were not sufficient to address these hypotheses, I developed a technique for high resolution 3d-reconstruction of microstructures, which can be applied to biological basic research such as morphological and taxonomical ques-tions, but is also attractive for applied sciences like nanotechnology and lightweight construction. The result of this new method was a very precise model of Actinoptychus senarius that contained volumetric information from confocal laser scanning microscopy (CLSM) and a detailed surface reconstruction from photogrammetric edited scanning electron microscopic (SEM) images. The model was post-processed to be suitable for a finite element analysis (FEA), a versatile tool widely used in engineering and construc-tion that allows the diatom models to be investigated and compared in computer simula-tions regarding stability, stress propagation or various other parameters like the specific lightweight value (publication I). The availability of these tools enables a systematic de-scription and cataloguing of biomechanical properties of silicified protists on a larger scale. The prerequisites for a systematic census of the diatoms biomechanical properties with the methods described above are cleaned and fluorescent stained diatom frustules. Publication II introduces facilitated cleaning procedures of diatoms and introduced three novel fluorescent stains for silica structures. The methods introduced payed atten-tion especially to fossil frustules as the study of recent and fossil diatoms reveal insights into current biogeochemical processes and, by using them as proxies, into past ones. Be-sides biological and paleontological applications, diatoms have attracted interest in the fields of nanotechnology, architecture and lightweight constructions in general. For all these questions, the three dimensional information of the frustules acquired with the methods developed in this thesis are either essential or bear an additional benefit. This thesis has shown that the investigation of planktonic ecosystems from a biome-chanical perspective requires novel methodological approaches, but has a very high po-tential for a mechanistic understanding of the interaction between individual species. This can help to improve ecosystem models, e.g. to predict shifts in the composition of key organisms. The new methods described here have a considerable potential for appli-cation in other fields as well, such as the understanding and development of new tech-nical lightweight constructions.