4.1. Seasonal and vertical variations of soil CO2 flux
Because of the temporal and spatial heterogeneity in plant and animal species, activities and organic carbon matrix, soil CO2fluxes exist strong distribution pattern of vertical heterogeneity and seasonal variation.The CO2 flux in soil vertical profile reflects the production and transport of CO2 in the soil and also provides information on the dynamics of soil CO2 (Fontaine et al., 2007; Franzluebbers et al., 2001). The soil profile of this study extends from the surface to the depth of 200cm. With the deepening of the soil depth, the soil CO2 flux of crop land and Robinia pseudoacacia of different stand ages is decreasing, and the CO2 flux inRobinia pseudoacacia (80-200 cm) at deep layers only changed slightly (Fig. 2A), which was affected by the interaction of land-use types and profiles (P < 0.01) (Table 2). However, the CO2 flux at shallow layers (5-80 cm) has a large change, and the trend is similar to that in previous studies (Davidson, Savage, et al., 2006; Fang C M, 2005; Pumpanen et al., 2008). The difference of CO2 flux at different depth is mainly attributed to the relatively stable soil environment, lower content of soil organic carbon (Fig. 3), and small changes in soil temperature and moisture (Fig2B, Fig2C) at deep layers than shallow layers.
Due to the difference in temperature and moisture caused by seasonal changes, the characteristics of the CO2 soil flux at different profiles also presents certain easonal. Pumpanen (Pumpanen et al., 2008) and Hashimoto (Hashimoto et al., 2007) found that soil CO2 fluxes at different depths represented significant seasonal, high in summer ,low in winter, and soil CO2fluxes are relatively high in shallow layers than deep layers. However, Risk et al (Risk, Kellman, & Beltrami, 2002) found the opposite result, the CO2 flux in deep layers show stronger seasonality than it in shallow layers, which may be caused by land-use types and vegetation patterns in the experimental sites. In our study, there was a two peaks trend in the soil CO2 flux. The trend in shallow layers was more obvious than that in deep layer, and the peak value always appeared in Dec.2012 and Jun.-Aug.,2013 (Fig. 2A). The occurrence of double peaks indicates that the seasonal variation of CO2 flux could not be explained simply by changes of temperature and moisture. The double peaks appeared in the growing season and non-growing season, respectively. All of which can further indicate that restrictive factors affecting soil CO2flux by vegetation root activity may be converted, and further studies on root and microbial activity are needed to explain double-peaks changes. The slight change range of CO2 flux at deep layers may be attributed to the relatively stable temperature, moisture, and the lower content of soil organic carbon.