Global change due to human activities causes an increasing number of disturbed ecosystems and poses a challenge to humanity because human life on earth depends on the stability of ecosystem services. Microorganisms are main drivers of element cycling, they contribute largely to the global organic carbon budget and are therefore fundamental for all biological processes and relevant for ecosystem services. They furthermore represent model organisms to test ecological theory, as they are small and have short generation times what facilitates the generation of comprehensive datasets for statistical evaluation.

Earlier research points to the following mechanisms that support the stability of community functioning in fluctuating environments: first, the tolerance of individual community members against environmental change impacts also the community-level robustness and the classification of individual community members along the generalist-specialist continuum may be a valuable tool to evaluate the vulnerability of a whole community in a disturbed environment. Secondly, there has been a long debate about the link of community structures, such as diversity patterns but also the architecture of species interactions to the vulnerability of these communities to environmental change.

The genomic material of microbial communities represents a blueprint of their functioning and should contain the information about the tolerance of individual community members as well as information concerning the above mentioned community structures. However, apart from ongoing advances, there is still a lack in the ability to fully exploit the large amount of information stored in meta-omic data.

The overall aim of my project is to use metagenomic and metatranscriptomic data of aquatic microbial communities to delineate and understand mechanisms that enable functional stability of these communities. To address this aim I have been working during the last 3 years with a strong emphasize on experimental work to better understand and interpret genomic signatures that are decisive for the functional stability of microbial communities. My focus in the coming three years will be to link findings from these experiments to the dynamics of microbial communities in natural environments. For this purpose, I want to apply machine learning algorithms to (i) use metagenome data for predicting functional attributes, such as resilience and resistance of aquatic microbial communities and (ii) evaluate the environmental heterogeneity of natural aquatic environments from microbial metagenomic signatures.