Synergistic effects of diffusion and microbial physiology reproduce the Birch effect in a micro-scale model

摘要

Large rainfall events following drought cause pulses of CO2 flux that are higher than models predict. This phenomenon, named the "Birch effect" after its discoverer, has been observed for decades, and will influence carbon-climate feedbacks as drying-rewetting (DRW) cycles become more common under intensified climates. Yet, the many interacting factors that determine how soil DRW cycles affect C balance have been difficult to separate empirically. Here we use a spatially explicit biogeochemical-microbial model to examine the mechanisms underlying CO2 dynamics under DRW. We independently model physiological activity and diffusion based on how they vary with (constant) moisture levels in nature, and subject the model to DRW to test the importance of different mechanisms in models with one or two microbial functional groups (cheaters and producers). Our model reproduces respiration patterns similar to empirical observations of the Birch effect when we include mechanisms that link water content to microbial growth and to diffusion rate, whereas inclusion of either mechanism alone produces significantly lower pulses upon rewetting. Diffusion limitation under drought increases substrate availability under rewetting, a process mediated by biogeochemical hotspots and continued enzyme activity under drought. At the same time, high microbial growth under rewetting is needed to replenish enzyme pools and to sustain the biomass required to generate respiration pulses under repeated DRW. Inclusion of cheaters in the model dampens the size of the rewetting pulse and the cumulative amount of CO2 release, as cheaters outcompete producers and reduce overall biomass. Our results provide several novel hypotheses regarding the microbial, biogeochemical, and spatial processes that mediate the Birch effect, which will contribute to a better mechanistic understanding of this important deviation from model predictions. (C) 2015 Elsevier Ltd. All rights reserved.