The increasing frequency and severity of extreme events is an inherent part of changing climate. Among extreme events, droughts are probably the most widespread and most drastic events, which can disrupt ecosystem functioning. Droughts are typically studied either by assessing the impacts of naturally occurring drought events or by excluding precipitation in drought experiments. The overall objective of the proposed research is to study ecosystem responses to drought by integrating experimental and observational studies. We plan to combine three interrelated research topics into a coherent framework. First, we will investigate the medium- and long-term ecosystem (productivity, decomposition, biomass allocation, soil moisture) and community (plant, nematode, root-colonising fungi) responses in a field experiment that combine a single extreme drought (in 2014) with recurring moderate manipulation of precipitation (drought and excess water; from 2015 onward) in a full factorial design. Second, we will asses, in the same ecosystem, how small-scale experimental disturbances affected the sensitivity to and recovery after an extreme natural drought (2003) and subsequent weather fluctuations in based on a 20-year dataset (1998-2017). Finally, we propose to conduct a meta-analysis and compare drought experiments with observational studies on natural droughts in terms of productivity responses. We believe that studying droughts in combination with other factors, such as chronic precipitation change and disturbances, as well as comparing experimental and natural drought, will provide new insight into current and potential future effects of droughts and climate change.