Temperature and moisture sensitivities of soil respiration vary along elevation gradients: An analysis from long-term...

Based on long-term field observation data over 11 years at 23 sites in two mountainous areas (TS1 and TS2) at elevations from 829 to 2700 m, where the dominant vegetation type of TS1 and TS2 was temperate mixed broadleaf-coniferous forest and cold temperate coniferous forest, respectively, we analyzed the correlations between soil respiration (R_s) and abiotic and biotic factors to explore the response patterns of R_s to environmental factors within and between the sites along the elevation gradient. We found that soil moisture (θ) and its combinations (T_s × θ and θ/T_s) with soil temperature (T_s) increased significantly with increasing elevation, while T_s, soil bulk density (SBD), C/N ratio, and pH decreased significantly with increasing elevation. Within each site, both exponential- T_s (ET) and Gaussian-T_s (GT) models could be used for predicting the R_s seasonal variation, except for two sites in the area of TS1, where θ was a better predictor than T_s. The integrated ET-θ and GT-θ models could be applied to all sites except for 22S, and both were superior to the ET and GT models. The mean R_s of each site over the measurement period ranged from 3.07 to 6.94 μmol CO₂ m^-2 s^-1 and showed a quadratic increase along the elevation gradient. Among the 23 sites, Q₁₀ ranged from 1.15 to 3.79, and it increased with elevation, reaching a maximum at an elevation of 2366 m; the θ sensitivity parameter (d) decreased significantly with elevation and reached a minimum at an elevation of 1975 m. Both the d and T_s sensitivity parameter (b) of R_s were complementary to each other along the elevation gradient. Among the sites, T_s, θ, and combinations of the two were more important drivers for both R_s and Q₁₀ variations than microbial and physicochemical indicators.