Changes in soil total organic carbon after an experimental fire in a cold temperate coniferous forest: A sequenced monitoring approach

- Spatiotemporal changes in total soil organic carbon (SOC) after fire were monitored.
- SOC content (SOCC) and SOC density (SOCD) differed after fire damage.
- Topsoil collapse, burning and erosion caused a shift-down of actual sampling depth.
- This resulted in a “dilution effect” for SOCC and “superadditive effect” for SOCD.
- Burnt organic matter was deposited in concave areas, increasing SOC markedly.

Although numerous studies have demonstrated the effects of fire on the total amount of soil organic carbon (SOC), much remains unknown. In Autumn 2009, a well-controlled fire experiment was conducted in the boreal forest of Daxing'an Mountains, China. Over the next three years, repeated samplings were performed at settled points, taking the pre-fire state as the reference base (control). All soil samples were collected at a fixed sampling depth (0-10 cm). Soil organic carbon content (SOCC) and soil organic carbon density (SOCD) were examined simultaneously. A paired-samples T-test was employed to ascertain any possible significant changes after the fire. The total amount of SOC in 0-10 cm soil was dynamic after fire disturbance, and the direction and magnitude of the changes largely depended on the burning severity and the post-fire period. The spatial patterns of SOC changes and their temporal succession were closely associated with the primitive pattern of fire severity. For low- and medium-severity fires, no significant immediate change in SOCC or SOCD was detected. After one or two leaching seasons, a statistically significant increase (less than + 5% on average for SOCC, less than + 10% on average for SOCD) was observed in low- and medium-severity burning areas due to incorporation of fire-derived forest debris and semi-decomposed dead roots, but this minor increase was diminished or offset by the third year. High-severity fire caused significant immediate C loss in topsoil through direct combustion, while the subsequent changes varied: either rapidly and continually decreased until the third year (from approximately − 10% to − 60% for SOCC, and from − 5% to − 40% for SOCD) as a result of subsequent erosion, or increased dramatically (from + 40% to more than + 100% for SOCC) due to deposition of burnt OM and/or organic soil, depending on the convex or concave microtopography. The sensitivities of SOCC and SOCD to fire disturbance were inconsistent due to changes in soil bulk density. SOCD was relatively sensitive to the temporary increase after medium-severity fire, while SOCC was more sensitive to the prolonged decrease following high-severity burning. Factors affecting changes in SOC were more complicated than previously reported. In particular, the collapse, burning and erosion of topsoil all contributed to the downward-shift of actual sampling depth, which resulted in a “dilution effect” on SOCC and a “superaddition effect” on SOCD. This effect was an underlying reason for the sensitivity differentiation of the two indices and further complicated the characterization of changes in SOC.