Seasonal variations in the availability of labile substrate confound the temperature dependence of organic matter decomposition

In empirically deriving the temperature dependence of organic matter decomposition, changing substrate availability can confound the derivation of any inferred intrinsic temperature dependence. In essence, when conditions are favourable for rapid decomposition, that fast rate can deplete the pool of available substrate leading to reduced subsequent decomposition rates. This is a potential problem under any experimental or observational setting. Its potential effect for measurements under seasonally varying temperatures is investigated here in a modelling study. Soil organic matter continuously loses carbon through decomposition which is generally replenished through new litter influx from senescing plant leaves, roots or other carbon sources. The CenW/CENTURY model was used to investigate to what extent inclusion of varying substrate supply within a realistic modelling framework modified the derived temperature dependence of organic matter decomposition. The model was run with different lignin to nitrogen ratios of fresh litter, and with litter either being generated continuously at a constant rate, or with litter fall being restricted to autumn. In systems with recalcitrant litter, as might be produced by conifers or eucalypts, the confounding effect of changing substrate supply was only slight. In systems with more labile litter, however, such as that produced by nutrient-rich grasslands, the confounding effect of varying substrate availability substantially weakened the derived temperature dependence. This effect was even more pronounced in systems with litter fall restricted to the autumn months. Reported temperature dependencies inferred from measurements with seasonally varying temperatures have shown weaker temperature dependencies than those inferred from laboratory incubation. The direction and magnitude of the confounding effect of changing substrate supply modelled here was consistent with the difference in temperature response observed in these different systems. It thus helps to reconcile these different reported temperature dependencies. (C) 2012 Elsevier Ltd. All rights reserved.