Understanding soil and ecosystem respiration in a dune-meadow cascade ecosystem

Abstract Arid and semi-arid regions, which account for more than 30% of the Earth's land area, increasingly dominate the spatiotemporal trends in global carbon fluxes. The Horqin Sandy Land is a typical semi-arid fragile ecosystem in northern China. Understanding the components of the carbon budget in ecosystems under conditions of extreme soil moisture limitations provides a foundation for comprehending the carbon balance in semi-arid ecosystems. The seasonal and diurnal variations in soil respiration ( R s ) in semi-mobile dune (SD) and meadow wetland (MW) ecosystems of the Horqin Sandy Land were examined, and the sources of CO 2 emissions from R s were identified using stable carbon isotopes. The responses of R s and ecosystem respiration ( R eco ) to environmental temperature, moisture and leaf area index ( LAI ) were revealed. The results showed that on a seasonal scale, in SD with soil moisture content ( M s ) below field capacity (FC), M s had a greater influence on R s than soil temperature ( T s ) during the growing season. Changes in the LAI during the middle and late growth period affected R s by altering root carbon supply. In MW, the most favorable M s for R s was near FC. The increase in LAI before mowing could effectively promote root and soil microbial respiration, and the decomposition of litter driven by T s was the main form of R s at this time. After mowing, root respiration and soil microbial respiration were the main processes contributing to CO 2 emissions. On a daily scale, relative humidity ( RH ) dominated the R s variation under dry conditions, whereas in other conditions, the R s was adequately explained by temperature in SD and MW. The overall R eco was larger than R s , but occasionally R s was greater than R eco . The effects of temperature, moisture and LAI on R eco and R s varied with growing season. Adding factors, such as ecosystem type, vegetation growth, water, and heat, to the carbon cycle model can improve predictions of carbon emissions, and aid in further management decisions in arid and semi-arid areas.