Temperature sensitivity of soil respiration across multiple time scales in a temperate plantation forest

Soil respiration (Rs) is the largest carbon (C) flux from terrestrial ecosystems to the atmosphere. Predictions of Rs and associated feedback to climate change remain largely uncertain, in part due to the high temporal heterogeneity of temperature sensitivity (apparent Q₁₀) of Rs under a changing climate. Therefore, it is of critical importance to provide better insight into how Q₁₀ varies across multiple temporal scales. We investigated the diurnal, seasonal, and annual variabilities in the Q₁₀ of Rs using continuous Rs measurements (at hourly intervals) over six growing seasons in a mature temperate larch plantation in North China. We found that night-time values of Q₁₀ were slightly lower than daytime values. Large seasonal and annual fluctuations of Q₁₀ were observed, as illustrated by high coefficients of variation of 15.0% and 21.8%, respectively. The higher Q₁₀ in spring and autumn were primarily regulated by fine root growth and higher soil moisture after snow melt in spring, and leaf senescence in autumn. Lower Q₁₀ in summer may have been caused by limitations in substrate availability and microbial activity resulting from drought, which also caused a decoupling of Rs from soil temperature in summer. Furthermore, a bivariate nonlinear model incorporating both soil temperature and soil moisture best explained Q₁₀ variability. Generally, lower soil temperature and higher soil moisture lead to higher values of Q₁₀, indicating that climate warming could exert a negative effect on Q₁₀, partially offsetting the warming-induced increase in soil C loss. We provide long-term field experimental evidence that it would be inappropriate to estimate Rs on a multiyear scale using a fixed Q₁₀ value or a value obtained from one season and/or one year. Thus, we emphasize the importance of incorporating the seasonal and annual heterogeneities of Q₁₀ into C cycle model simulations under future climate change scenarios.