Search for a command to run...
Abstract Community dynamics are complex and thus challenging to infer from observational data alone. Experiments, with their ability to control variables and isolate mechanisms, are a powerful tool for uncovering the causal processes that drive community dynamics. They therefore allow us to move beyond correlations and to directly test theoretical predictions. Yet, because experiments are often logistically demanding and resource-intensive, they are less frequently employed than observational approaches in community ecology. Here, we trace the past three decades of experimental research in community ecology through a systematic literature review. We focus on the motivation behind experiments, their links to ecological theory, the types of questions they address, their scale, and the methods used to do this. Our results corroborate the historically tight relationship between experiments and ecological theory and document a gradual increase in experimental complexity —particularly related to the use of molecular methods. However, persistent gaps remain in the taxa and ecosystems studied, with aquatic ecosystems, fungi, and microbes still underrepresented compared to terrestrial plants and animals. Moreover, experiments are still limited in their spatial and temporal scale; they are typically short-term, local, and reliant on manual methods. The integration of high-throughput technologies with experimental workflows is still in its infancy, even though they are increasingly common in biomonitoring. To illustrate the potential of such tools in experimental research, we present a proof-of-concept study. It shows how automated technologies can be incorporated at different stages of the experimental workflow to expand the scale of experiments while reducing the reliance on human labor and potentially lowering financial costs. We conclude that many of the long-lasting biases and challenges in experimental community ecology could be addressed by combining technological innovations with broader collaboration among research groups. Coordinated networks, standardized protocols, and the integration of long-term and large-scale experimental designs can substantially improve in situ replication as well as cross-site comparability. Such efforts are essential for developing a more comprehensive mechanistic understanding of community dynamics across diverse ecosystems.