Contrasting temperature responses of soil respiration derived from soil organic matter and added plant litter

Accurate description of temperature response and sensitivity of different carbon pools within soil is critical for accurately modelling soil carbon stocks and changes. Inconsistent sampling, incubation and fractionation methods highlights the need for new approaches to this area of study. We developed and tested a new protocol which allowed measurement of the temperature response of two carbon pools within soil. A Horotiu silt loam soil, wet up to 60% maximum water holding capacity, was mixed with ¹³C-enriched plant litter and incubated for 5 or 20 h, at 30 discrete temperatures (~ 2-50 Celsius degree). A mixing model was used to separate respired CO₂ into litter and soil organic matter sourced carbon pools, which were then fitted using macromolecular rate theory. Overall, litter sourced respiration had a low T_opt (the temperature where respiration rate is maximal) and was less temperature sensitive (Q₁₀) than soil organic matter sourced respiration, which was more Arrhenius-like. We attribute these differences in temperature parameters to the factors that control the availability of carbon to microbes from the labile litter (enzyme kinetics with a clear temperature optimum) compared to the relatively stable soil organic matter (desorption and diffusion that exhibit Arrhenius behaviour). The developed method is rapid and reliable and may be suited to exploring temperature response of a variety of ¹³C-labelled pools in soil and more clearly demonstrates that labile carbon has very different temperature response than more stable carbon pools in soil.