Characterization of the soil CO2 production and its carbon isotope composition in forest soil layers using the flux-gradient approach

• We estimate soil CO2 & δ13CO2 source profiles with the Flux-Gradient Approach (FGA).
• FGA gave consistent results except in Ol which is influenced by turbulent transport.
• Local temperature is the main driver of intra-day & inter-day variability of CO2 sources.
• Humidity was the main driver of surface δ13CO2 source inter-day variability.
• Variability of δ13CO2 sources was not detected from the δ13C of surface fluxes (δ13Fs).

The Flux-Gradient Approach (FGA) allows the vertical distribution of gas turnover and production in the soil to be calculated. This approach has been used successfully for greenhouse gases, such as CO2, CH4 and N2O but not yet for the stable isotope composition of soil CO2 (δ13CO2), although this subject has become increasingly important. In this study, an innovative experimental set-up, based on the MEmbrane Tube Technique (METT) and adapted to C stable isotope recording, was used to carry out in situ measurements in a Scots pine forest soil in Hartheim, Germany. Continuous measurements of soil CO2 and δ13CO2 efflux and of soil air CO2 and δ13CO2 concentration were combined with the FGA to investigate the vertical distribution and temporal variability in CO2 production (P) and its isotopic signature (δ13P). The FGA gave consistent values for P and δ13P in each soil horizon, except for the Ol horizon where diffusive transport seemed to be affected by atmospheric turbulence. The results showed that P was subjected to a significant vertical stratification, whereas no significant δ13P vertical variation was found. The surface soil water content (SWC) was reported to be critical to the accurate partitioning of P between the topsoil layers. In addition, significant temporal P and δ13P variations were found in the most productive horizon, the first being best explained by the soil temperature, the second depending on the moisture conditions. No visible δ13P climatic dependence was visible in the measured surface flux isotopic composition (δ13Fs). Finally, in the litter layer (Ol), a significant correlation was found between P and friction velocity. This was attributed to the predominance of non-diffusive transport in the litter layer.