Greenhouse gas (GHG) emissions from thawed permafrost are difficult to predict because they result from complex interactions between abiotic drivers and multiple, often competing, microbial metabolic processes. Our objective was to characterize mechanisms controlling methane (CH4) and carbon dioxide (CO2) production from permafrost. We simulated permafrost thaw for the length of one growing season (90 days) in oxic and anoxic treatments at 1 and 15 A degrees C to stimulate aerobic and anaerobic respiration. We measured headspace CH4 and CO2 concentrations, as well as soil chemical and biological parameters (e.g. dissolved organic carbon (DOC) chemistry, microbial enzyme activity, N2O production, bacterial community structure), and applied an information theoretic approach and the Akaike information criterion to find the best explanation for mechanisms controlling GHG flux. In addition to temperature and redox status, CH4 production was explained by the relative abundance of methanogens, activity of non-methanogenic anaerobes, and substrate chemistry. Carbon dioxide production was explained by microbial community structure and chemistry of the DOC pool. We suggest that models of permafrost CO2 production are refined by a holistic view of the system, where the prokaryote community structure and detailed chemistry are considered. In contrast, although CH4 production is the result of many syntrophic interactions, these actions can be aggregated into a linear approach, where there is a single path of organic matter degradation and multiple conditions must be satisfied in order for methanogenesis to occur. This concept advances our mechanistic understanding of the processes governing anaerobic GHG flux, which is critical to understanding the impact the release of permafrost C will have on the global C cycle.

Thawing permafrost supplies dissolved organic carbon (DOC) to aquatic systems; however, the magnitude, variability and fate of this DOC is not well constrained. Our objective was to examine DOC respiration from seasonally thawed and near-surface (<1.5 m) permafrost soils collected from five locations in the Kolyma River Basin, north-east Russia. We measured soil organic carbon (OC) content, water-soluble macronutrients (DOC, NH4, PO4) and the heterotrophic respiration potentials of soil extract DOC in five-day laboratory incubations. DOC concentrations ranged from 2.8 to 27.9 mg L-1 (n = 14). Carbon respiration was 0.03-0.47 mg C (n = 16) and 8.7-31.4%, total DOC (n = 14). While DOC concentration was a function of soil OC concentration, we did not find a relationship between C respiration and soil OC or DOC concentrations. Respiration was highest in the top active layer, but varied widely among sites, and lowest at the bottom of the active layer. Respiration from yedoma varied across sites (0.04-0.47 mg C respired, 8.7-31.4% total DOC). Despite the small sample size, our study indicates near-surface soils and permafrost are spatially variable in terms of both soil OC content and C respiration rates, and also that OC contents do not predict C respiration rates. While a larger sample size would be useful to confirm these results at broader geographic scales, these initial results suggest that soil OC heterogeneity should be considered in efforts to determine the fate of soil OC released from permafrost-dominated terrestrial ecosystems to aquatic ecosystems following permafrost thaw.