|dc.description.abstract||The effect of temperature on the decomposition of labile and stable carbon (C) in soil is a critical factor in understanding soil C dynamics, particularly under a global warming scenario. Despite the temperature response of soil respiration being well-studied, the temperature sensitivity of the labile and stable C pools in soil remains unclear. Here, I conducted a laboratory-based incubation experiment to separate the temperature response of these two C pools. Soil was incubated at 18 discrete temperatures (~8-52°C) for five hours with and without added C substrates. After the incubation period, respiration rates were determined by measuring carbon dioxide (CO₂) production from the soils. Previous studies have attributed differences in the temperature sensitivities between C pools to substrate availability. Therefore, the labile C pool was measured as the CO₂ derived from the added C substrates (450 mM C) to the amended soil (i.e. high C availability), and the stable C pool was represented by the decomposition of soil organic matter (SOM) measured directly from soil without added C (i.e. low C availability).
I applied the macromolecular rate theory (MMRT) to respiration rates (RS) derived from labile and stable C decomposition, to describe their relationship with temperature. I found in all cases, the temperature response of the decomposition of labile C substrates (labile pool) was well fitted with MMRT (i.e. Δ𝐶𝑃‡≠ 0), resulting in clear temperature optima (Topt) and inflection (Tinf) points. While previous studies have been able to fit SOM-RS (stable pool) with MMRT, the equation in this study collapsed to the exponential-like Arrhenius equation (i.e. Δ𝐶𝑃‡ = 0), with no calculable Topt or Tinf points.
Representing the labile C pool, five simple C compounds, a substrate comprising a wide variety of C compounds (yeast extract), and a complex, long-chained glucan (dextran) were added separately to soil and incubated. The temperature response of the simple C compounds (glucose, glutamine, arginine, maltose, and lysine) and yeast extract were largely the same, with average Topt and Tinf points of 37°C and 22°C, respectively. Dextran behaved similarly to the recalcitrant SOM as it could not be fitted with MMRT; thus, no Topt and Tinf parameters were derived. This behaviour was attributed to the high complexity and molecular weight of the compound, reducing its metabolic compatibility with the soil microbes and increasing its physiochemical protection by adhering to soil surfaces. Additionally, I also determined whether the temperature response of these two pools varied with different soil properties by incubating three soils (allophanic, gley, and organic) with and without added glucose. Results suggested that between three very different soils, the temperature response of the decomposition of glucose (labile C) and SOM (stable C) was remarkedly similar, only varying by 3°C (Topt=35°C, Tinf=18-21°C).
A final preliminary experiment determining the soil priming effect was undertaken using a newly developed isotopic analysis method, where 13C labelled glucose was added to soil (allophanic). The results showed that within this study, priming enhanced the native SOM decomposition by 30% and had a similar temperature response to the labile C pool. Priming-RS was well fitted to MMRT suggesting that the C made available by priming was a labile C source for decomposers. Topt and Tinf points for priming-RS (29-31°C and 14-15°C, respectively) were slightly lower than the Topt and Tinf observed for labile C compounds. Additionally, in both relative and absolute terms, the temperature sensitivity of C decomposition induced by priming was, generally, lower than for the added C substrates and SOM. These results suggest that the C made available by priming must be more labile and physically accessible to microbes compared to the labile C compounds and SOM.
Overall, this study found that the temperature sensitivity of the stable C pool was higher in relative terms (i.e. Q10) and lower in absolute terms (i.e. first derivative) compared to the labile C pool. The temperature responses of these two pools were not the same. MMRT accurately described the temperature-respiration relationship for labile C decomposition, whereas the Arrhenius function better described this relationship for stable C. MMRT describes typical reaction rate responses from systems where biological processes dominate (in this case the biological degradation of soil C by microbes), and is a typical response of the temperature-respiration relationship when substrate supply is high in soil. The Arrhenius function, on the other hand, accurately describes chemical reaction rates, which in this case would be physiochemical processes such as, sorption/desorption and diffusion that transport the protected C to the microbes for decomposition. An Arrhenius-like behaviour is commonly exhibited when substrate supply is low for the soil microorganisms. Since the temperature response of the labile and stable C pools did not significantly differ between labile substrate types and soil properties, it could be suggested that a two-pool soil C model might be sufficient for the prediction of soil C dynamics. This two-pool model could also be extrapolated to larger ecosystem models, potentially leading to improved accuracy of climate and soil C storage projections. To determine the adequacy and appropriateness of a simple two-pool C model, further measurements of the temperature response of these two pools from a wide distribution of soils from different locations and under various conditions needs to be undertaken.||