Microorganisms have colonized virtually all soils on earth, yet our understanding of how soil microbial processes influence terrestrial carbon cycling and carbon-climate feedbacks remains rudimentary. Because of environmental and/or nutritional stress, most (>80-90%) microorganisms in soil are in a state of minimum metabolic activity (i.e. dormant) under which they are incapable of driving soil biochemical processes. Only a few of the most recent soil carbon models incorporate microbial processes, and even fewer explicitly represent the disproportionate differences in the pools of active and dormant microbial biomass in soil. However, because of the lack of experimental/observational data, validation of microbial dormancy (and most microbial processes) in models remains unresolved and challenging. Here we quantified the relationships between soil respiration and the pools of total (TMB) and active (AMB) microbial biomass in a variety of soils (i.e. forest and grassland soils from two climatic regions) and environmental conditions (i.e. rewetting pulses on heated and unheated soils). We estimated microbial biomass and its active/dormant fractions using the Substrate Induced Growth Response (SIGR) method. We compared predictions of cumulative soil CO2 from a decomposition model parameterized with observations of TMB or with only the active fraction (based on our findings) of microbial biomass.
Results/Conclusions
Rewetting of dry soils (i.e. from 10 to 60% water holding capacity) generated respiration pulses (up to a 30-fold increase in 2-3h) that were linked to increases in AMB rather than to changes in TMB. These responses did not vary across ecosystems or climatic regions, but were different among soils acclimated to different temperatures. Rewetting-induced increases in soil respiration rates were explained by increases in the fraction of active biomass (up to 7.3±2.6%) in heated soils, but not in unheated soils. Predictions of cumulative CO2 based on TMB observations were up to almost twice as great as predictions that assumed a 90-95% dormant fraction. Altogether, our findings suggest that: 1) in a variety of soils TMB is unresponsive to short-term changes in environmental conditions, 2) CO2 pulses such as those that follow rewetting events are more dependent on re-activation of dormant microbes than on microbial growth, 3) warming increases soil CO2 emissions by increasing the capacity of soils to support larger fractions of AMB (when moisture constraints are released), and 4) caution should be used when defining and parameterizing microbial biomass in models.