Integration of genome-informed microbial traits in ecosystem models reveals divergent methane productions in thawing permafrost
Date:
Permafrost contains a substantial amount of carbon which is vulnerable to microbial decomposition in a warming climate, releasing greenhouse gases (e.g., CH4, CO2) into the atmosphere. The magnitude of this positive climate feedback is uncertain, and much depends on microbial responses to climate change. Although many efforts have been made to develop earth system models (ESMs), most ecosystem models are generally deficient in explicitly representing microbial processes, weakening the predictive power of these models in a changing climate. This underrepresentation comes from the lack of a mechanistic framework to represent the complex interactions of microbial processes with environmental factors, the interaction between microbial functional groups, and microbial parametrization. In this study, microbial physiological traits (e.g., maximum respiration rate and affinity constants) were parameterized from genome sequences based on soil samples collected from Stordalen Mire. Those microbial traits were integrated into an ecosystemscale biogeochemical model (ecosys), which explicitly considers microbial processes, plant dynamics, and other thermal and hydrological processes. Morris sensitivity analyses were conducted to elucidate the dominant microbial functional groups and associated microbial traits on greenhouse gas emissions along a permafrost thaw gradient. We found microbial traits inferred from genome information had large impacts on CH4 flux. For example, fermenters strongly control annual CH4 fluxes at our fen site, while aerobic heterotrophs substantially regulate annual CH4 fluxes at our bog site. Our analyses show large CH4 cycling sensitivity to microbial functional groups and environmental conditions and indicate the need to represent microbial traits at the ecosystem scale. Our findings demonstrate the need to mechanistically incorporate microbial dynamics in biogeochemical models and highlight the importance of linking metagenomics with ecosystem-scale processes.
https://ui.adsabs.harvard.edu/abs/2021AGUFM.B12D..08L/abstract