Sand-fixing vegetation regulates deep percolation and soil water dynamics in semiarid sandy land: Evidence from a deep percolation recorder and in-situ test

Deep percolation (DP) determines the sustainable utilization of regional groundwater resources and provides large hydrological benefits to ecosystems in semiarid sandy regions. However, studies involving DP variability have rarely conducted in-situ tests with different vegetation coverages (VC) and vegetation types using direct measurements. This study measured soil moisture and DP for four years in five plots, with DP measured using a deep percolation recorder: mobile sand (MS) as the control group; Artemisia ordosica semifixed (AOSF) with VC of 30%-35% and Artemisia ordosica fixed (AOF) with VC of 70%-75% (AOSF and AOF stand ages: 30-40 years), representing different vegetation coverages; and Salix psammophila fixed (SPF) and Pinus sylvestrix var. Mongolica fixed (PSMF) with the same VC of 45%-50% (SPF and PSMF stand ages: 30 years), representing different vegetation types. The effects of sand-fixing vegetation on DP and soil water were evaluated using the relative extractable soil water (RESW) and soil water storage deficit metrics. DP occurrence in MS was closely related to rainfall amounts of > 10 mm and high-intensity rainfall rather than the total annual rainfall. Increases in vegetation coverages and forest selection significantly reduced DP and RESW compared with MS. DP in the shrub plot (SPF) was significantly higher than that in the forest plot (PSMF). Plants utilized water from different sources among the sandy land types: the shallow (0-80 cm) and deep layers (80-200 cm) in PSMF and AOF, and only from the shallow layer in AOSF and SPF. However, compared with MS, soil water storage deficit occurred to varying degrees in each plot. This study can provide a scientific reference for the restoration of the ecology and water cycle in sandy regions.