Characterization of soil organic matter fractions in fire-affected hillslopes using mid-infrared spectroscopy

Background/Question/Methods

Fire is a significant driver of numerous ecological processes, including carbon (C) cycling and erosion.  In many fire-affected upland ecosystems, post-fire erosion rates are significantly higher than ambient levels of lateral soil distribution.  Most of the soil material mobilized by erosion is deposited downslope and gets buried and stabilized within the soil after subsequent erosion events. In addition, fire-altered carbon, or pyrogenic carbon (PyC) is produced with fire and gets eroded and deposited along with bulk OM in depositional landform positions.  We used Mid-infrared (MIR) spectroscopy that was calibrated with ¹³C NMR (nuclear magnetic resonance) spectroscopy to determine how distribution of C in fire-affected soils varies spatially (along a hillslope) and temporally (up to 10 years after the fire event) at the Gondola Fire site in South Lake Tahoe.  Here the spectroscopic analysis of soil carbon was combined with a partial least squares regression (PLSR) to divide up the soil organic carbon (OC) into three fractions: resistant OC (charcoal), humic OC, and particulate OC.   

Results/Conclusions

In the post-burn hillslope locations, we found no significant loss of carbon, likely because of inputs from charred aboveground biomass.  The hillslopes had no significant losses in any of the carbon fractions immediately after the fire, however the soil structure exhibited important changes in the long-term.  The depositional area increased in carbon, nitrogen, and in all three organic carbon fractions post-burn.  This increase is consistent with the lateral movement of material downslope post-burn into the depositional area.  Understanding the temporal fluxes in PyC and OC fractions post-wildfire is crucial for our understanding of how soils respond to fires and erosion.  The ability of soil to recover from perturbations is important for overall ecosystem health and functioning.  In addition, understanding where and in what form of C is stored in a landscape is crucial to generating more accurate C budgets.