Response of soil respiration to temperature and soil moisture: Effects of different vegetation types on a small scale in the eastern Loess Plateau of China

Abstract

Soil respiration is affected by vegetation and environmental conditions. The purpose of this study was to investigate the effect of vegetation type on soil respiration, temperature and water content, and their correlations on a small scale. We measured soil respiration rate (R s ) over a 3-year period at biweekly intervals in three plots in the eastern Loess Plateau of China, with the same soil texture but different vegetation types: pine forest, grassland, and shrub land. Simultaneously, soil temperature (T s ) at 10 cm depth and soil water content (W s ) within 10 cm depth were measured. The seasonal course of R s and T s showed a similar temporal variation in the three plots, with higher values in summer and autumn and lower values in winter and spring. No significant differences (P . 0.05) were found between plots, except for W s . The mean cumulative release of CO 2 efflux from March to December was 962.5, 1027.5, and 1166.5 g C m 22 a 21 for plots 1, 2, and 3, respectively, with no significant difference between plots. The fitted exponential equations of R s versus T s from the 3-year data-set were significant (P , 0.05) with an R 2 of 0.72, 0.64, and 0.72 for plots 1, 2, and 3, respectively. The calculated Q 10 from the parameters of the fitted equation was 3.57, 3.52, and 3.61, and the R 10 was 2.36, 2.03, and 2.37 mmol CO 2 m 22 s 21 for plots 1, 2, and 3, respectively. Compared with the T s , the correlations between R s and W s were not significant for the three plots. However, if the T s was above 108C, then their correlation was significant, and W s had an impact on R s . Four combined regression equations including two variables of T s and W s could be well established to model correlations between R s and both T s and W s . Our study demonstrated that the exponential and power model fitted best and no significant different correlations of combined equations existed between the three plots. These results show that vegetation type had little impact on R s , T s , W s , and their correlations, as well as on related parameters such as Q 10 and R 10 . Therefore, while doing R s research in a horizontal patchy vegetation conditions on a small area, the sampling location of measurements should focus on vertical dominant vegetation and ignore patch vegetation so as to reduce field work load.