1. 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.2. 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).3. 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.

Disentangling the relative response sensitivity of soil autotrophic (Ra) and heterotrophic respiration (Rh) to nitrogen (N) enrichment is pivotal for evaluating soil carbon (C) storage and stability in the scenario of intensified N deposition. However, the mechanisms underlying differential sensitivities of Ra and Rh and relative contribution of Rh to soil respiration (Rs) with increasing N deposition remain elusive. A manipulative field experiment with multi‐level N addition rates was conducted over 3 years (2015–2017) in an alpine meadow to explore the relative impact of N enrichment on Ra and Rh and the response of Rh/Rs ratio to the gradient of N addition. Soil respiration components had different sensitivities to N enrichment, with Ra decreasing more than Rh, leading to a higher Rh/Rs ratio as a function of increasing N addition rates. Ra and Rh decreased nonlinearly as N addition rates increased, with a critical load of 8 g N m−2 year−1 above which N enrichment significantly inhibited them. Ra and Rh were controlled by different abiotic and biotic factors, and the regulation of controlling factors on soil respiration components varied over time. N‐induced reduction in the relative abundance of forb significantly affected Ra, and this effect was mainly evident in the second and third years. Nitrogen enrichment significantly changed Rh in the third year, and the decreased Rh under high doses of N addition could be attributed to the changes in microbial biomass C, soil substrate quality and microbial composition. Our study highlights the leading role of Ra in regulating Rs responses to N enrichment and the enhancement of Rh/Rs ratio with increasing N addition. We also emphasize that N‐induced shifts in plant community composition play a vital role in regulating Ra instead of Rh. The changing drivers of Ra and Rh with time suggests that long‐term experiments with multiple levels of N addition are further needed to test the nonlinear responses and underlying mechanisms of soil respiration components in face to aggravating N deposition. A free Plain Language Summary can be found within the Supporting Information of this article.