Drought-associated changes in climate and their relevance for ecosystem experiments and models
- 1Research Group of Plant and Vegetation Ecology, Department of Biology, Universiteit Antwerpen (Campus Drie Eiken), Universiteitsplein 1, 2610 Wilrijk, Belgium
- 2Laboratory of Plant Ecology, Department of Applied Ecology and Environmental Biology, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, 9000 Ghent, Belgium
Abstract. Drought periods can have important impacts on plant productivity and ecosystem functioning, but climatic conditions other than the lack of precipitation during droughts have never been quantified and have therefore not been considered explicitly in both experimental and modeling studies. Here, we identify which climatic characteristics deviate from normal during droughts and how these deviations could affect plant responses. Analysis of 609 years of daily data from nine Western European meteorological stations reveals that droughts in the studied region are consistently associated with more sunshine (+45 %), increased mean (+1.6 °C) and maximum (+2.8 °C) air temperatures and vapour pressure deficits that were 51 % higher than under normal conditions. These deviations from normal increase significantly as droughts progress. Using the process-model ORCHIDEE, we simulated droughts consistent with the results of the dataset analysis and compared water and carbon exchange of three different vegetation types during such natural droughts and droughts in which only the precipitation was affected. The comparison revealed contrasting responses: carbon loss was higher under natural drought in grasslands, while increased carbon uptake was found especially in decidious forests. This difference was attributed to better access to water reserves in forest ecosystems which prevented drought stress. This demonstrates that the warmer and sunnier conditions naturally associated with droughts can either improve growth or aggravate drought-related stress, depending on water reserves. As the impacts of including or excluding climatic parameters that correlate with drought are substantial, we propose that both experimental and modeling efforts should take into account other environmental factors than merely precipitation.