The lakes that form via ice-rich permafrost thaw emit CH4 and CO2 to the atmosphere from previously frozen ancient permafrost sources. Despite this potential to positively feedback to climate change, lake carbon emission sources are not well understood on whole-lake scales, complicating upscaling. In this study, we used observations of radiocarbon (14C) and stable carbon (13C) isotopes in the summer and winter dissolved CH4 and CO2 pools, ebullition-CH4, and multiple independent mass balance approaches to characterize whole-lake emission sources and apportion annual emission pathways. Observations focused on five lakes with variable thermokarst in interior Alaska. The 14C age of discrete ebullition-CH4 seeps ranged from 395 ± 15 to 28,240 ± 150 YBP across all study lakes; however, dissolved 14CH4 was younger than 4,730 YBP. In the primary study lake, Goldstream L., the integrated whole-lake 14C age of ebullition-CH4, as determined by three different approaches, ranged from 3,290 to 6,740 YBP. A new dissolved-14C-CH4-based approach to estimating ebullition 14C age and flux showed close agreement to previous ice-bubble surveys and bubble-trap flux estimates. Differences in open water versus ice-covered dissolved gas concentrations and their 14C and 13C isotopes revealed the influence of winter ice trapping and forcing ebullition-CH4 into the underlying water column, where it comprised 50% of the total dissolved CH4 pool by the end of winter. Across the study lakes, we found a relationship between the whole-lake 14C age of dissolved CH4 and CO2 and the extent of active thermokarst, representing a positive feedback system that is sensitive to climate warming.

Methane is a potent greenhouse gas, and there are concerns that its natural emissions from the Arctic could act as a substantial positive feedback to anthropogenic global warming. Determining the sources of methane emissions and the biogeochemical processes controlling them is important for understanding present and future Arctic contributions to atmospheric methane budgets. Here we apply measurements of multiply-substituted isotopologues, or clumped isotopes, of methane as a new tool to identify the origins of ebullitive fluxes in Alaska, Sweden and the Arctic Ocean. When methane forms in isotopic equilibrium, clumped isotope measurements indicate the formation temperature. In some microbial methane, however, non-equilibrium isotope effects, probably related to the kinetics of methanogenesis, lead to low clumped isotope values. We identify four categories of emissions in the studied samples: thermogenic methane, deep subsurface or marine microbial methane formed in isotopic equilibrium, freshwater microbial methane with non-equilibrium clumped isotope values, and mixtures of deep and shallow methane (i.e., combinations of the first three end members). Mixing between deep and shallow methane sources produces a non-linear variation in clumped isotope values with mixing proportion that provides new constraints for the formation environment of the mixing end-members. Analyses of microbial methane emitted from lakes, as well as a methanol-consuming methanogen pure culture, support the hypothesis that non-equilibrium clumped isotope values are controlled, in part, by kinetic isotope effects induced during enzymatic reactions involved in methanogenesis. Our results indicate that these kinetic isotope effects vary widely in microbial methane produced in Arctic lake sediments, with non-equilibrium δ18 values spanning a range of more than 5‰.