Riverine ecosystems are among the most species-rich ecosystems on earth. Furthermore, river networks are important dispersal and invasion corridors for many organisms. For a long time, however, diversity and ecological process in riverine networks has been studied from a localized perspective, ignoring the inherent dendritic structure of riverine networks. We study dispersal processes and diversity patterns in dendritic networks, using a wider range of approaches, including simulation models, protist microcosm experiments and field studies. We are interested in understanding how dispersal along differently structured spatial networks affects species interactions and community composition. Thereby, we are using the metapopulation and metacommunity concepts to identify the significance of spatial dynamics on the composition of local populations and communities. Recently, we also started to address the significance of spatial dynamics on ecosystem functioning. The theory of meta-ecosystems gives a conceptual and mathematical framework for studying flow of materials and organisms in spatially explicit systems and subsequent consequences on community dynamics. Meta-ecosystem models, for example, suggest that the flow of material can reverse source-sink community-dynamics and describe how community dynamics and ecosystem processes (e.g., productivity, decomposition) in one patch can outreach and influence other patches. We are using protist microcosm systems to test these predictions. Our broad approach, combining theory, laboratory and field studies, aims at a causal understanding of spatial dynamics in complex habitat networks.