Is resistant soil organic matter more sensitive to temperature than the labile organic matter?

A recent paper by Knorr et al. (2005a) suggested that the decomposition of resistant soil organic matter is more temperature sensitive than labile organic matter. In Knorr et al.’s (2005a) model, the reference decay rate was presumed to be same for all pools of soil carbon. We refit Knorr et al.’s (2005a) model but allow both the activation energy and the reference decay rate to vary among soil C pools. Under these conditions, a similar fit to measured data can be obtained without invoking the assumption that the resistant C pool is more temperature sensitive than the labile pool. Other published evidence does not unequivocally support Knorr et al.’s (2005a) hypothesis of increased temperature sensitivity of resistant pools of soil carbon. Because of the lack of experimental data, Knorr et al.’s (2005a) conclusion that the decomposition of the resistant SOM is more temperature sensitive than the labile pool is premature. 1 Temperature sensitivity of resistant soil organic matter The response of soil organic carbon (SOC) to temperature change or global warming is important for predicting feedbacks between SOC and climate change. Because of the difficulties and large uncertainties in estimating the temperature sensitivities of the decomposition of soil organic matter (SOM) pools, the relationship between the temperature sensitivity of decomposition and SOM pools is of paramount interest (Davidson et al., 2000; Ågren, 2000; Reichstein et al., 2005). The resistant C pool (often referred to as stable, recalcitrant or “old” SOM in literature) was thought to be less sensitive to temperature change than the labile pool (Liski et al., 1999; Giardina and Ryan, 2000), which was not supported by analysis of SOM pools and their contributions to Correspondence to: C. Fang ([email protected]) the overall turnover of soil C stocks (Davidson et al., 2000), simulation of SOM dynamics in soil warming experiments (Kirschbaum, 2004; Eliasson et al., 2005) and soil incubation (Townsend et al., 1997). A similar temperature sensitivity was applied to both labile and resistant pools in many multi-pool SOM models (e.g. the CENTURY model, Parton et al., 1987; the DNDC model, Li et al., 1993; the Roth-C model, Coleman and Jenkinson, 1996). Two recent papers highlight current debate in this field. Based on a laboratory incubation of soil samples, Fang et al. (2005) concluded that the decomposition of resistant SOM pool is not less sensitive to temperature than the labile pool. Knorr et al. (2005a) used a multi-pool model to fit data from Holland et al. (2000) and suggested that the model can simulate the long-term temperature sensitivity of SOC decomposition, and that the resistant carbon pool is more sensitive to temperature than the labile pool. As the future response of soil stored C to global warming is mainly dependent on the temperature sensitivity of the resistant C pool (Fang et al., 2005), Knorr et al.’s (2005a) finding may have important implications for future studies. Here, we argue that a presumption of a fixed reference decay rate for all pools used in the model of Knorr et al. (2005a) necessarily leads to the conclusion that the resistant pool is more temperature sensitive than the labile pool. We show that if this assumption is incorrect, the finding that resistant C is more sensitive to temperature is not supported. In Knorr et al. (2005a), the decomposition of SOM was simulated with a multiple pool model: