Increasing temperatures and changes in precipitation due to climate change are projected to increase the frequency, severity and duration of droughts, with potentially large and long-lasting ecological effects. Over the past two decades, passive rainfall removal shelters have emerged as the dominant method to experimentally impose drought in natural ecosystems. There are currently hundreds of published, ongoing and new drought experiments in a wide range of ecosystems globally. Given the importance of such experiments in advancing our understanding of the ecological effects of drought, a quantitative assessment of the methodology is urgently needed. In this study, we evaluated the methods of experimentally imposing droughts with rainout shelters, using 109 relevant published studies combined with 50-year precipitation records from nearby weather stations. Our goals were threefold: 1. examine where and how these studies have been conducted, 2. evaluate the extremity of the drought treatments by placing them into a historical context, and 3. assess the influence of interannual precipitation variability on the ability to experimentally impose drought and detect significant ecological effects.
In the literature review, we found that drought experiments tended to be short-term, extreme droughts that were imposed over short-statured vegetation in water-limited ecosystems. Such biases are not surprising given financial and logistical constraints, however it is important to note the important knowledge gaps that remain. When placed in a historical context, the majority of experimental droughts were extreme, as 61% were below the 5th percentile of annual precipitation and 43% were below the 1st percentile. We also found that the differences in annual precipitation inputs between control (ambient precipitation) and drought treatments depended on the ambient precipitation for a given year. Thus, while the imposed precipitation reduction remained constant (e.g. 50% reduction), the relationship between precipitation in control and drought treatments varied nonlinearly with interannual precipitation amount. Overall we found that interannual variability in precipitation had a large and potentially underappreciated influence on drought experiments. This may explain why despite the general trend towards extreme magnitude of experimentally imposed droughts, the ecological effects within and across experiments can vary drastically.