One contribution of 17 to a theme issue 'Biodiversity and ecosystem functioning in dynamic landscapes'. The relationship between biodiversity and ecosystem functioning (BEF) and its consequence for ecosystem services has predominantly been studied by controlled, short-term and small-scale experiments under standardized environmental conditions and constant community compositions. However, changes in biodiversity occur in real-world ecosystems with varying environments and a dynamic community composition. In this theme issue, we present novel research on BEF in such dynamic communities. The contributions are organized in three sections on BEF relationships in (i) multi-trophic diversity, (ii) non-equilibrium biodiversity under disturbance and varying environmental conditions, and (iii) large spatial and long temporal scales. The first section shows that multi-trophic BEF relationships often appear idiosyncratic, while accounting for species traits enables a predictive understanding. Future BEF research on complex communities needs to include ecological theory that is based on first principles of species-averaged body masses, stoichiometry and effects of environmental conditions such as temperature. The second section illustrates that disturbance and varying environments have direct as well as indirect (via changes in species richness, community composition and species' traits) effects on BEF relationships. Fluctuations in biodiversity (species richness, community composition and also trait dominance within species) can severely modify BEF relationships. The third section demonstrates that BEF at larger spatial scales is driven by different variables. While species richness per se and community biomass are most important, species identity effects and community composition are less important than at small scales. Across long temporal scales, mass extinctions represent severe changes in biodiversity with mixed effects on ecosystem functions. Together, the contributions of this theme issue identify new research frontiers and answer some open questions on BEF relationships in dynamic communities of real-world landscapes.