Quality not quantity: Organic matter composition controls of CO2 and CH4 fluxes in neotropical peat profiles

Tropical peatlands represent an important source of carbon dioxide (CO2) and methane (CH4) to the atmosphere.However, we do not know where in the peat profile these gases are produced and how controlling factors, such as substrate quality, which can vary substantially with peat age, and anoxic-oxic conditions, interact to determine production rates.To address this knowledge gap, this study investigated if substrate limitation of CO2 and CH4 production differs under anoxic-oxic peat conditions using entire peat profiles, from tropical peatlands in Panama.We determined the variation in peat organic chemistry through stratigraphic profiles using tetramethylammonium-pyrolysis-gas chromatography-mass spectrometry (TMAH-Py-GC/MS).To explore how variation in peat organic chemistry through the depth profile impacted on CO2 and CH4 production rates under anoxic-oxic conditions we carried out a series of incubation experiments.The TMAH-Py-GC/MS analysis showed high concentrations of long chain fatty acids (> C20) in surface peat, and variation in the distribution of the lignin monomers through the peat profile.Both anoxic CH4 and CO2 production was greatest from the surface of the peat profile with surface peat accounting for 92 ± 1.7 and 54 ± 2.9 % of the cumulative CH4 and CO2 production, respectively.The high CO2 and CH4 production rate under anoxic conditions, in surface peat, was strongly related to greater concentrations of lignin, but also long chain fatty acids and polysaccharides, in this section of the peat profile.As expected, CH4 production decreased, and became decoupled from peat organic chemistry, following peat aeration.In contrast, aeration dramatically increased CO2 emissions throughout the entire peat profile.This demonstrates that the recalcitrance of buried peat does not protect C stocks in tropical peatlands, if their water tables are lowered in response to drainage or prolonged drought.In conclusion, our work highlight that information on both labile substrate availability and water table fluctuation are needed to predict CO2 and CH4 fluxes from tropical peatlands.