The temperature sensitivity of soil organic matter (SOM) decomposition has gained increasing interest because of its potential importance to soil carbon (C) cycling in response to future global warming. SOM includes a complex continuum of organic components with varying chemical structures, diverse associations with minerals, and a wide range of turnover times. Relatively recalcitrant C makes up the bulk of SOM and has a longer turnover time relative to labile C, so its response to temperature may be particularly influential on predictions of future C concentrations in soils and in the atmosphere. Two temperate grassland soils were incubated at four temperatures, and two mathematic models were fit to observed SOM decomposition rates. Both models showed that temperature sensitivities of the relatively labile C were significantly larger than those of relatively recalcitrant C.

The stable C isotope composition of soil-respired CO₂ (the flux out of soil) is an important parameter that has been used to evaluate soil C dynamics. Results from incubation experiments of the same temperate grassland soil indicated that an isotope effect occurred during SOM decomposition. Microorganisms in the studied soils discriminated against ¹³C at the beginning of the incubation experiments and at lower temperatures. Soil microbes discriminated more against ¹³C at low temperatures (14^oC) relative to the isotope discrimination obtained at higher temperatures (24 and 34^oC). The natural abundance of ¹³C should be used with caution to estimate the soil C dynamics.

To experimentally assess how soil characteristics affect the temperature sensitivity of SOM decomposition, we developed an artificial soil with controlled composition. In this artificial soil, two soil parameters, SOM chemical recalcitrance (cellulose vs. lignin) and clay-mineral composition (montmorillonite vs. kaolinite), were varied, while other essential soil characteristics were controlled. The incubation results showed that the presence of cellulose enhanced the decomposition rate of lignin. Decomposition of high-cellulose organic matter was sensitive to temperature only at 2-12^oC, while decomposition of high-lignin organic matter had similar temperature sensitivities over the entire temperature range of 2-32^oC. SOM decomposition rate was greater in treatments containing both kaolinite and montmorillonite than in the treatment containing pure kaolinite at temperatures of 12^oC and above. Decomposition of organic matter associated with high-montmorillonite soils had high temperature sensitivities at 2-12^oC, whereas decomposition in pure-kaolinite soil was sensitive to temperature at 12-22^oC. The temperature sensitivities at 22-32^oC were all low, regardless of chemical recalcitrance or soil clay-mineral composition.

The interactive effect of substrate and clay-mineral compositions on the temperature sensitivity of SOM decomposition was then tested in an artificial soil designed to vary both lignin/cellulose content and montmorillonite/kaolinite content. The results showed a significant interactive effect at 2-12^oC: the temperature sensitivities of montmorillonite treatments were higher in the pure-cellulose treatment than in the pure-lignin treatment; in contrast, the temperature sensitivities of kaolinite treatments did not vary with substrate composition.