Search

Global climate change models predict an increase in the frequency and intensity of extreme droughts, with uncertain ecological impacts across ecosystems. In particular, it is not clear when extreme droughts will elicit extreme ecological responses.For this study, we employed three complementary approaches to explore the relationships between extreme drought and ecosystem responses. First, we used global data mining to evaluate the relationship between extreme gross primary productivity and extreme precipitation from 1980 to 2013. Second, we conducted a meta‐analysis using 132 drought experiments across the globe to assess the response ratios (RRs) of above‐ground net primary productivity (ANPP) to extreme versus non‐extreme drought treatments. Third, we examined community and ecosystem responses in an alpine meadow to a drought gradient experiment, which included five precipitation treatment levels (1/12 P, 1/4 P, 1/2 P, 3/4 P, and P, where P is the growing season precipitation).This study had three key results. In our historical data mining, we found that extreme droughts elicited extreme ecological responses only 15.1% of the time. The meta‐analysis results indicated that there were no significant differences in the RRs of ANPP between the extreme versus non‐extreme drought treatments. The drought gradient experiment results revealed that although the four drought treatments were statistically extreme, only the most extreme drought treatment (1/12 P) significantly reduced ANPP over the 3 years. Meanwhile, species richness and asynchrony were significantly reduced under the 1/12 P treatment, which led to a significant reduction in productivity.Synthesis. These results suggest that extreme ecological responses to extreme drought may be less frequent than previously thought. But when they do occur, extreme ecological responses may be driven by plant community changes such as species asynchrony, species loss or species reordering. Our experimental results highlight the key role of community dynamics in determining the resistance of ecosystem productivity to extreme drought, which should be assessed when predicting ecological responses to climate change. [ABSTRACT FROM AUTHOR]