Interactive effects of hydrological conditions on soil respiration in China's Horqin sandy land: An example of dune-meadow cascade ecosystem.

Soil moisture (Ms) strongly influences dynamic changes in soil respiration (Rs) and is thus an important factor when predicting soil carbon emissions. However, the various sources of Ms (rainfall, groundwater, and condensation) exert complicated and uncertain effects on Rs. This study examined the growth seasonal variation (from April to October) of Rs and the diurnal variation in a cascade ecosystem consisting of sandy bare ground, a transitional artificial Populus forest, and a meadow Phragmites communis community in China's Horqin sandy land. Simultaneous measurements of the 0-10 cm depth soil temperature (Ts) and Ms, rainfall, the surface air relative humidity and the groundwater depth were collected. The results revealed that in sandy bare ground with Ms below field capacity, Ms had a greater impact on Rs than Ts, and rainfall could increase Rs. The effect of condensation on Rs during periods of continuous drought could not be ignored. In the meadowlands with Ms above field capacity, the groundwater affected Rs indirectly by regulating Ms and the relationship with Ts, and rainfall had an adverse effect on Rs. The effects of rainfall, Ms and Ts on Rs were minimum as Ms approached the saturation water content. In the transitional forest, Ms and Ts were the main factors controlling Rs. The most favorable Ms for Rs was close to the field capacity. The results emphasize that field capacity and saturation water content are the demarcation points of a soil carbon emissions prediction model, and the effect of different hydrological conditions and Ts on Rs at each segment are reconsidered accordingly. Ultimately, the carbon emission patterns of the cascade ecosystems in arid and semi-arid areas are extremely complicated and have to be considered specially for estimating terrestrial carbon emissions.