2015/07/31 by Hans J. De Boeck, Sara Vicca, Jacques Roy +11 · 146 citations
Agricultural and Biological Sciences · Environmental Science · Mathematics · Psychology · #Artificial intelligence #Attribution #Biology #Cartography #Climate change impacts on agriculture #Climate variability and models #Computer science #Control (management) #Data science #Ecology #Ecosystem dynamics and resilience #Epistemology #Extrapolation #Focus (optics) #Geography #Mathematics #Process (computing) #Psychology #Realism #Scale (ratio) #Set (abstract data type) #Social psychology
paper · pdf · doi:10.1093/biosci/biv099
published in BioScience 65(9), 922-931 (Oxford University Press)
openalex publication_date 2015/07/31 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/29
Manipulation experiments are invaluable tools in global change ecology because they enable causal and process-based understanding. However, artifacts and inherent limitations can lead to misinterpretations. Across the wide range of approaches to set up such studies, we distill the main challenges associated with the imposed treatment(s), the spatial and time scale, proposing solutions and outlining the limitations in interpreting and extrapolating results. The inherent trade-offs between experimental realism (facilitating extrapolation) and control (facilitating the attribution of observed responses) resonate throughout this review. The focus on realism or control determines which issues become more important and how they should be handled. For example, covarying factors such as temperature and moisture can be explicitly separated to attribute effects more precisely but could also be left uncontrolled to increase realism. Ultimately, combining results across gradients of scale and control, including the use of “natural laboratories,” stimulates fundamental understanding, enabling more confident predictions of responses to global change.