Sensitivity of soil organic carbon to the change in climate on the Tibetan Plateau

Abstract

Soil organic carbon, as the main terrestrial component in the Earth’s carbon cycle, has a profound effect on the accumulation of CO2 in atmosphere and consequently on global warming. In the alpine grasslands of the Tibetan Plateau, the decompo- sition rate of soil organic carbon is controlled by several biotic and abiotic factors, which mostly change simultaneously and often leads to freezing and thawing cycles. However, it is highly uncertain whether the temperature sensitivity of decomposition around the freezing point of water is similar as in higher temperature ranges. The main objective of this dissertation is to evaluate the effects of simultaneous changes in three climate factors on soil organic carbon decomposition rates using an incubation experiment and a biogeochemical model. Due to the large divergence between empirical and model-based approaches in predicting the effects of abiotic factors on soil carbon dynamics, this dissertation first provides an approach to un- cover some sources of uncertainty in estimated SOC processes in alpine grasslands. In this study, I evaluated the complexity of the model required to represent the dynamics of carbon observed in incubation studies. Information theory metrics including AIC and BIC, as well as carbon particle mean transit time, were used as criteria to select models that better predict the data without making additional assumptions about model structure. These analyses showed that during the limited course of an incubation experiment, the amount of transfer between the different SOC pools is negligible and adding these parameters to the model could lead to over-parameterization. These findings suggest that carbon models with less parameterized structures, such as the one-pool model or the two-pool model with parallel structure, that does not account for transfers between pools, indeed have better predictive power in describing the decomposition of carbon fractions while following the principle of parsimony. The aforementioned information was later used to evaluate the sensitivity of SOC degradation rates to changes in soil temperature, soil moisture, and oxygen availabil- ity, especially at low temperatures. Functions from the Dual Arrhenius-Michaelis- Menten model (DAMM) were implemented in a one-pool model of SOC represented as first-order differential equation with time-dependent coefficients. A manipulative freeze-thaw cycle was imposed on a soil from Tibetan grasslands, in addition to soil moisture treatments that ranged from extremely dry to fully saturated, under both oxic and anoxic conditions. The intrinsic sensitivities indicated that temperature (energy) is the main factor limiting decomposition in cold environments, provided moisture and oxygen are sufficiently available. However, the intrinsic sensitivities related to soil moisture and oxygen concentration are only relevant in very narrow ranges when soils are nearly dry or partially anoxic, and small changes within these narrow ranges can result in large changes in decomposition rates.