Drought experiments need to incorporate atmospheric drying to better simulate climate change

Climate models predict more frequent, prolonged, and extreme droughts in the future. Therefore, drought experiments varying in amount and duration across a range of biogeographical scenarios provide a powerful tool for estimating how drought will affect future ecosystems. Past experimental work has been focused on the manipulation of meteorological drought: Rainout shelters are used to reduce precipitation inputs into the soil. This work has been instrumental in our ability to predict the expected effects of altered rainfall. But what about the nonrainfall components of drought? We review recent literature on the co-occurring and sometimes divergent impacts of atmospheric drying and meteorological drying. We discuss how manipulating meteorological drought or rainfall alone may not predict future changes in plant productivity, composition, or species interactions that result from climate change induced droughts. We make recommendations for how to improve these experiments using manipulations of relative humidity.