novel approach for identifying the true temperature sensitivity from soil respiration measurements

• Published in: Global biogeochemical cycles 2008-12, Vol.22 (4), p.n/a
• Main Authors: Gu, Lianhong, Hanson, Paul J, Post, W. Mac, Liu, Qing
• Format: Article
• Language: English
• Subjects: acclimation; accuracy; air temperature; biogeochemical cycles; CARBON; CARBON CYCLE; climate change; CLIMATES; data analysis; environmental factors; ENVIRONMENTAL SCIENCES; heterotrophs; INCUBATION; localized ratio fitting; localized ratio fitting (LRF); measurement; methodology; model validation; MOISTURE; new methods; photosynthesis; plant litter; regression analysis; RESPIRATION; seasonal variation; SENSITIVITY; SIMULATION; soil microorganisms; soil organic carbon; soil respiration; soil temperature; soil water content; SOILS; temperature sensitivity; temporal variation; time series analysis; VALIDATION
• ISSN: 0886-6236; 1944-9224
• DOI: 10.1029/2007GB003164

We propose a novel approach, called the “localized ratio fitting” (LRF), to estimating the true temperature sensitivity from soil respiration measurements, a task crucial to modeling terrestrial carbon cycle and climate but so far hindered by the inadequate conventional regression approach. LRF takes advantage of the different timescales of the pool dynamics–induced and environmental variation–induced changes in soil CO2 efflux. It first transforms the expression for soil respiration into a form suppressing the influence of soil carbon pool dynamics and then uses the transformed expression to infer the parameters of environmental sensitivities. LRF works best for high‐frequency soil respiration measurements and thus is particularly suitable for analyzing time series produced by automated soil chambers and from soil incubation experiments. We evaluated the validity of LRF with both simulated (with a multipool soil organic carbon model driven by realistic plant litter input scenarios) and measured (with automated soil chambers) time series of soil respiration. LRF accurately retrieved the true temperature sensitivity from the simulated heterotrophic soil respiration while the conventional approach failed to do so. The simulation also revealed that LRF performed better than the conventional approach when a direct photosynthetic signal existed in the time series of soil respiration although even LRF could not completely eliminate the interference of photosynthetic contribution for estimating the true temperature sensitivity. Importantly, the simulation on the photosynthetic influence reproduced a typical seasonal pattern of apparent temperature sensitivity reported in the literature: higher sensitivity in winter (dormant season) and lower sensitivity in summer (growing season). Such pattern has been interpreted as an indication of temperature acclimation of soil respiration by previous studies. Our simulation now indicated that that interpretation may be incorrect. The validation with actual soil chamber data showed that the use of LRF led to more consistent estimates of temperature and moisture sensitivities from observations, indicating its better robustness against compounding effects of parallel processes on soil respiration. It was demonstrated that once the true environmental controls were properly accounted for, soil respiration measurements could be used to infer effects of biological processes on soil respiration.