4.1. SWC variability in different landcover types
In general, amounts of mean annual precipitation determine the type and spatial pattern of dominant landcover in an ecosystem (Liu et al., 2020; Sun et al., 2014); however, human activities shift the rate and the magnitude of this relationship, particularly with large-scale programs such as “Grain for Green” in China (Cao, 2008; Cao et al., 2011). New landcover types change the water cycle from the land surface to the atmosphere by abiotic evaporation and biotic transpiration; with a complex hydraulic interaction between landcover and the environment, a new trade-off forms gradually and this change produces irreversible impacts to the local people’s livelihood (Liu et al., 2020).
SWC in this study decreased in the following order: natural forest, brushland, grassland, planted forest, and mixed forest, with SWC in natural forest differing significantly from that in the other vegetation types (p<0.05) (Fig. 4). Furthermore, SWC in planted and mixed forests years decreased with age in this study, likely leading to a negative feedback to the water circulation processes in the ecosystem; these results were consistent with those of previous studies (Wang et al., 2011; Zhu et al., 2017), and may be illustrative of the repetitive degradation of the planted area within the past 20 years (A et al., 2019; He et al., 2007; Jia et al., 2017; Muneepeerakul et al., 2011; Sun et al., 2014). In the early stages of plantation establishment, soil moisture formed a new trade-off between the new vegetation cover and precipitation, and then a relatively stable water cycle was gradually established (Fu et al., 2003; Gómez-Plaza et al., 2001; Sun et al., 2014). It appears that forestland in arid regions can accumulate soil moisture when vegetation pattern is fully mature many years after planting. These findings highlight ecological efficiency of afforestation with hysteresis over long-time scales (Guan et al., 2020; Jia et al., 2017).
Topography is another notable source of SWC variability (Famiglietti et al., 1998; Fu et al., 2003). However, in this study spatial pattern of SWC showed along slope position SWC did not exhibit a significant declining trend from the top to the toe of slope in this study (plot #9 in Fig. 3 and Fig. 5). These characteristics could generally be explained by -14.3% runoff depth (forests decreased net water yield) and a mean annual precipitation of 374.1 mm in the Pailugou catchment (adjacent to this study area) in the Qilian mountains (He et al., 2012) indicating that interflow along the subsurface in this region cannot determine a self-organized vegetation pattern, as reported in the loess hilly region of China by Sun et al. (2014). The terrain-related processes may be the main environmental factor affecting the spatial pattern of vegetation during wet periods (Grayson et al., 1997; Western et al., 1999).