Pathways to carbon liberation: A systems approach to understanding carbon transformations and losses from thawing permafrost

Background/Question/Methods: The IsoGenie Project is investigating in situ changes in carbon cycling and microbiology across a thawing permafrost peatland mosaic landscape, at Stordalen Mire (68°21′N, 19°02′E) in Arctic Sweden. The Mire hosts palsas, bogs, fens, and post-glacial ponds and lakes, with total landscape methane emissions split approximately evenly between the wetlands and lakes. Over the last half-century of climate warming, the spatial extent of palsas has been decreasing while bogs and fens are increasing, with vegetation shifting from ericaceous shrubs to mosses and sedges. Our team uses a multi-disciplinary systems approach to understand carbon transformations and losses across this thawing habitat series.

Results/Conclusions: High-resolution characterization of organic matter using a range of methods (including Fourier transform ion cyclotron resonance mass spectrometry, UV/Vis absorbance, and excitation-emission matrix spectroscopy) has revealed that soil carbon becomes increasingly reduced and labile across the gradient, with evidence of greater humification rates and faster decomposition in the fen habitat. The changes in substrate quality translate to shifting methane emissions, which peak in the fens, and which increase relative to carbon dioxide across the gradient. These emissions shift isotopically from ratios characteristic of hydrogenotrophic production in the bog habitat to a mixture with acetoclastic production in the fen, consistent with the more reactive nature of the fen organic material. Complex microbial communities mediate carbon transformations and losses across this landscape, transforming substrates to emissions. Roughly 200 samples have been metagenomically characterized allowing for the assembly of ~1500 microbial population genomes and ~1900 viral population genomes. Both traditional and novel analytical approaches are connecting these microbiota to specific carbon transformations, and targeted metatranscriptomics and metaproteomics are helping identify active lineages for the processes of interest, while a variety of incubation experiments are addressing specific hypotheses. The ecosystem process model DNDC has been adapted to the site and not only updated to include isotopes, but also to include more explicit representation of microbes to more accurately represent, in a predictive framework, the processes giving rise to greenhouse gas emissions. Unraveling the interconnections of microbial lineages and carbon transformations is ongoing. We are housing all project data, from autochamber isofluxes to metaproteomes, in a spatially-explicit database with both internal and public web-access.