INTERPRETING STABLE CARBON ISOTOPE PROFILES OF ORGANIC MATTER IN SOILS (Invited Presentation)

The δ¹³C values of soil organic matter (SOM) typically increase with depth in well-drained soils. A widely accepted explanation for this trend is that isotopic fractionation during decomposition (ε_CO2-SOM) results in preferential loss of ¹²C as CO₂ and retention of ¹³C in the soil. Topsoil incubations, however, have shown that values of ε_CO2-SOM are too small or of the wrong sign to explain down-profile trends. These observations could be reconciled if 1) fractionation during respiration of recalcitrant carbon is larger than that for labile carbon observed in topsoil incubations, 2) microbial carbon use efficiency (CUE) is small allowing for microbes to preferentially concentrate ¹³C and simultaneously generate CO₂ with δ¹³C values similar to the carbon they consume, and/or 3) the effect of rising atmospheric PCO2 on C₃ plant δ¹³C values is accounted for. To explore these possibilities, we used a three-pool soil carbon model. We parameterize the model with a Markov Chain Monte Carlo Metropolis-Hastings approach using organic carbon concentrations and ¹⁴C activities from soils in Great Mountain Forest (GMF), CT. Prior distributions for parameters were defined by ranges of values from the literature. We then ran simulations with ε_CO2-SOM for the rapid pool prescribed to a small range consistent with the results from topsoil incubations (-0.9 to +0.6‰) to investigate possibilities 1 and 2. In another set of simulations, we prescribed ε_CO2-SOM for the slow and stable pools at 0‰ and imposed empirical CO₂ records in order to account for possibility 3. We find that the PCO2 effect explains the down-profile δ¹³C trend at GMF. Furthermore, magnitudes of ε_CO2-SOM for the three pools (calibrated while simultaneously considering the PCO2 effect) are small and of opposing signs (-0.02‰ to +0.08‰) suggesting that if GMF tree δ¹³C values responded primarily to PCO2 over the past 150 years then overall carbon isotope fractionation during decomposition is probably small. Alternatively, in the absence of the PCO2 effect, a larger value of ε_CO2-SOM during respiration of slow carbon (-0.6‰) or CUEs of 15-25% along with small ε_CO2-SOM (-0.2‰) for all pools explain the observations. Separation of these possibilities requires additional measurements such as microbial CUEs and δ¹³C values of tree rings and of microbially-respired CO₂.