Another Inconvenient Truth: Evidence That Q10 Is Not a Static Parameter.

Predicting soil carbon dynamics requires accurately representing the temperature sensitivity of decomposition. Many ecosystem models represent this temperature sensitivity by using the temporally-constant lumped parameter Q₁₀, which attempts to describe the relative change of heterotrophic respiration in response to a 10 K temperature change. However, empirical observations have revealed very diverse behavior of Q₁₀, including dependencies on: (1) the temporal length of decomposition, (2) the season when measurements were taken, (3) the depth where data were collected, (4) the incubation methodology, etc. We therefore contend that it is unlikely the temperature sensitivity of decomposition can be represented by a single Q₁₀ value. We tested these dependencies using a thermodynamically based soil carbon decomposition model that dynamically represents interactions between polymeric organic compounds, dissolved organic compounds, a diverse microbial community, extra- and intra-cellular enzymes, and mineral surfaces. Even using the simplest formulation of the model (one representative from each of these classes), we found that Q₁₀ has hysteretic dynamics that depend on specific interactions between abiotic and biotic factors. Therefore, the temperature sensitivity of decomposition can neither be represented as a single parameter nor a single explicit function. We also investigated the impacts of using a static Q₁₀ and found large differences in predicted soil carbon stocks with realistic temporal variability in forcing compared to the temporally-dynamic thermodynamic model. In particular, we found the use of a static Q₁₀ significantly overestimates the response of soil carbon stocks to temperature changes. Finally, we will discuss experiments that can be used to test our theoretical prediction that a static Q₁₀, or static temperature sensitivity function in general, should not be used for soil carbon modeling.