Fire decreases soil respiration and its components in terrestrial ecosystems

The impact of fire on aboveground biomass has significant consequences on soil carbon (C) dynamics, which is essential in predicting the global C budget during the Anthropocene. However, there is considerable spatiotemporal variability in the directions and magnitudes of fire effects on soil respiration, and the drivers associated with these effects are not well understood.
Here, we conducted a global meta-analysis of 1327 individual observations from 170 studies to determine the extent to which fire influenced soil total respiration (Rs), heterotrophic respiration (Rh), and autotrophic respiration (Ra).
We found fires reduced Rs, Rh, and Ra, with an average effect of -11.0%, -17.5%, and -40.6%, compared to unburnt sites. Specifically, wildfires significantly reduced Rsand Rh (-20.4% and -25.0%, respectively), and prescribed fire significantly decreased Ra (-74.8%). The influences of fire on Rs and its components were moderated by fire severity, season, type, climate zones, and biomes. After several years from the time of the fire, the negative effects of fire on Rs diminished and then recovered to a state not significantly different from unburnt sites; Rh exhibited a similar but decayed temporal response. Similarly, the negative effects on Ra disappeared after 3 years following the latest fire. The magnitude of the effect on Rs was strongly associated with soil temperature, cation exchange capacity, total nitrogen (N) content, and N-acquiring enzyme activity. In contrast, the magnitude of the effect on Rh significantly changed with pH, bulk density, texture, soil C and nutrient contents, and C- acquiring enzyme activity.
Our findings advance the understanding of the inhibition and associated mechanisms of fire on Rs and its components, highlighting the need for new research efforts to predict the spatial-temporal shifts in underground C cycling induced by fire. 

Methods
We extracted data directly from text, tables, figures, and appendices from the primary articles using the Getdata 2.2.5 software. In cases where site coordinates were not provided, we estimated them by referring to site descriptions and cited sources. Additionally, for unavailable climate data, we used a global database (http://www.worldclim.org/) and used spatial coordinates to fill in the missing climate data.