On the available evidence for the temperature dependence of soil organic carbon

Abstract. Two recent papers by Knorr et al. (2005) and Fang et al. (2005) provide variations of model fitting conducted in the former study. Knorr et al. (2005) suggested that more recalcitrant fractions of soil organic carbon (SOC) could be more sensitive to temperature. Fang et al. (2005) argue that this is an implication of the choice of model used. Further, Reichstein et al. (2005) point out that the evidence for a stronger temperature sensitivity of recalcitrant soil carbon mainly rests on an analysis of data provided by Kätterer et al. (1998) and argue for a different selection criterion to exclude short-term incubations. Here, we explain why the model used by Knorr et al. (2005) is the simplest multi-pool model that can fit the available data and is at the same time fully consistent with the concept of "pools", as opposed to some of the model formulations proposed by Fang et al. (2005). It is also pointed out that the criterion proposed by Reichstein et al. (2005) uses posterior information to determine inclusion of experimental data, a practice that should be avoided. We conclude that the original analysis of Knorr et al. (2005) as well as the one added by Fang et al. (2005) indicate that there is a serious possibility that recalcitrant SOC reacts more to temperature changes than labile SOC.