Constant and diurnally-varying temperature regimes lead to different temperature sensitivities of soil organic carbon decomposition

In a 122-day incubation experiment with two soil types under four temperature treatments, we examined whether the temperature sensitivity of soil organic carbon (SOC) decomposition differed between constant and diurnally-varying soil temperature regimes. We calculated the Q10 values after accounting for changes in substrate availability and quality among treatments over time. The Q10 values under constant temperature regime were consistently and significantly higher than those under diurnally-varying temperature regime, particularly in the later stages of decomposition (by up to 30%). This result indicated that different temperature regime was one of the important factors causing the current controversy about the temperature sensitivity of SOC decomposition in published reports. 2011 Elsevier Ltd. All rights reserved. The temperature dependence of soil organic carbon (SOC) decomposition effectively controls whether and howmuch CO2will be released to the atmosphere under global warming, and to what extent this release of CO2 from SOC may lead to further climatic warming (Davidson and Janssens, 2006; Heimann and Reichstein, 2008). Despite the appreciable amount of experimental and modeling work in the past few decades, there is still no scientific consensus on the temperature sensitivity of SOC decomposition (Kirschbaum, 2006; von Lützow and Kögel-Knabner, 2009), largely due to the fact that different experimental settings have often been used in separate studies (Davidson et al., 2006; Subke and Bahn, 2010). Among these experimental settings, temperature control regimes, e.g. diurnally and seasonally varying temperatures vs. preset constant temperatures,may play a significant role in causing this controversy. In most lab incubation experiments (e.g. Reichstein et al., 2000; Conant et al., 2008), jars filled with soil samples were incubated in several incubators where temperature was maintained at a constant level throughout the entire experimental duration. The temperature sensitivity of SOC decomposition (Q10) was calculated based on CO2 efflux measurements at different temperatures. In contrast, in most field warming experiments (e.g. Luo et al., 2001; Hartley et al., 2007), soil CO2 efflux was measured under both ambient and warming treatments in which soil temperature varied diurnally and seasonally. The calculated Q10 values based on CO2 efflux measurements under both ambient and warming treatments were usually lower than those from lab incubation experiments (Kirschbaum, 2006). To what degree the difference in temperature regimes between these two groups of experiments has resulted in inconsistent Q10 values remains unexplored. In this study, we incubated two soils (a farm soil and a grassland soil) under two temperature regimes (constant vs. diurnallyvarying), each of which has two average temperatures. Soil CO2 efflux was measured five times during the 122-day incubation period and Q10 was calculated for the two soil temperature regimes. Our main objective was to determine whether Q10 differs significantly between the two soil temperature regimes. The experiment was conducted in three growth chambers. We filled 16 bottom-capped PVC pots (diameter 15 cm, height 40 cm) with either 8.3 kg air-dried “farm” soil or 7.5 kg air-dried “grassland” soil. The farm soil and the grassland soil were both sandy loam, collected from the top 0e30 cm of an organic farm (converted from the nearby grassland in 1974) and an annual grassland on University of California at Santa Cruz campus, respectively. After four decades of farming, the farm soil has less carbon (0.94% vs. 1.20%) and nitrogen (0.11% vs. 0.13%) compared to the grassland soil. To minimize the initial disturbance effect on CO2 efflux due to air-drying, sieving, packing, and rewetting the soils, all pots were pre-incubated at room temperature (22 C) for 30 days before the 122-day experiment started. There were four soil temperature * Corresponding author. Tel.: þ1 831 331 3627; fax: þ1 831 459 4015. E-mail address: [email protected] (B. Zhu).