Aim: Precipitation manipulation experiments have shown diverse terrestrial carbon (C) cycling responses when the ecosystem is subjected to different magnitudes of altered precipitation, various experimental durations, or heterogeneity in local climate. However, how these factors combine to affect C cycle responses to changes in precipitation remains unclear.

Location: Global.

Time period: 1990–2019.

Major taxa studied: Terrestrial ecosystems.

Methods: Using observations from 230 published studies in which precipitation was manipulated and terrestrial C cycling variables were measured, we conducted a global meta-analysis to investigate responses of diverse C cycle processes to altered precipitation, including gross ecosystem productivity, ecosystem respiration, net ecosystem productivity, ecosystem carbon use efficiency, net primary productivity, aboveground and belowground net primary productivity, aboveground and belowground biomass, shoot:root ratio, soil respiration, and soil microbial biomass C.

Results:  We found that C cycling responses correlated linearly and positively with the magnitude of precipitation treatments. We also detected that the responses of net primary productivity (NPP) and its aboveground component (ANPP) to altered precipitation weakened with experimental duration. Furthermore, gross ecosystem productivity, ecosystem respiration, and net ecosystem productivity had larger responses to precipitation treatments of greater magnitude over shorter time periods. The response of soil respiration, a key component of the C budget in most terrestrial ecosystems, particularly depended on the local climate. Local temperature and precipitation not only influenced the magnitude of the response of soil respiration to altered precipitation but also affected its sensitivity to the magnitude of the precipitation treatments, with higher sensitivities in the response of soil respiration to treatment magnitude at drier and colder sites.

Main conclusions:  Our findings highlight the importance of the interactions between the magnitude of precipitation treatments, their duration, and local climate in the response of ecosystem C cycling to altered precipitation, which is critical to better understanding and projecting ecosystem C processes and functioning under changing precipitation regimes.

We conducted a meta-analysis using 1775 paired observations from global precipitation manipulation experiments to synthesize the effects of changes in precipitation on a diverse array of C cycle processes.