The impoundment of dams can significantly alter shorelines, hydrological regimes, sediment settling, and the soil nutrient cycle. With the completion of the Chushandian Reservoir Dam, land-use types around the reservoir have been profoundly affected by inundation. As yet, the changes due to inundation in soil properties for different land-use types have not been fully explored. We investigated the variation of soil physicochemical properties of different land-use types (abandoned cropland, grassland, and woodland) near the Chushandian Reservoir at several distances after short-term inundation and found that soil pH varied in the range of 5-6.5. The soil texture of abandoned cropland and grassland was dominated by silt particles, but the soil texture of woodland was dominated by sand particles. The ranges of variation in soil TC and TN for different land-use types were 5.26-33.91 and 0.59-2.68 g/kg, respectively. Woodland soil was characterized by low NH₄-N and high NO₃-N, which may be related to denitrification. In addition, we observed that inundation affected the soil nutrient status of the different land-use types in the riparian zone in the short-term after reservoir impoundment, especially in abandoned cropland and grassland. However, woodland showed better tolerance to inundation. Furthermore, there are two different mechanisms of soil properties in response to inundation at the distance scales for the woodland land-use type. Soil properties are dominated by microbial activity in the far-away region and by chemical properties in the near-water region. By contrast, the soil properties in abandoned cropland and grassland are dominated by physical properties. Finally, we found that microbial biomass may be an essential indicator that can characterize the ability of riparian zone soil to respond to inundation.

We conducted our soil survey on December 11, 2020, which was two months after the first controlled flood of the reservoir area reached its predetermined normal storage level. This meant that all corresponding land-use types were under short-term inundation. Sample plots of three different land-use types (abandoned cropland, grassland, and woodland) that border the watershed were selected along the riparian zone on the west side of the reservoir (which has no dike protection), with three replicated for each type (Figure 1), sampling vertically according to the distance from the water (0m, 2m, 20m). For convenience of expression, we write land-use types and distance scales together as woodland: W0, W2, W20; grassland: G0, G2, G20; abandoned cropland: C0, C2, C20. We collected 27 soil samples in total.

Taking the boundary between water and land as a sampling point of 0 m, we took our samples using a Petersen grab, and three adjacent soil samples were selected for each location and mixed into one. A five-point sampling method was used to mix the soil into one portion at 2m and 20m, and an additional copy was kept with the soil core intact and brought back to the laboratory immediately for measurement of BD (soil bulk density) and SM (soil moisture). All soil samples were divided into two sections. One was kept in a self-sealed bag and placed in a refrigerator at 4° C for the determination of NH₄-N, NO₃-N, DOC (dissolved organic carbon), MBC (microbial biomass carbon), and MBN (microbial biomass nitrogen) content. And the other one was stored in a self-sealed bag and brought back to the laboratory for air drying. Stones and roots were picked out with forceps, and each soil sample was divided into two portions using a 2 mm sieve. One portion was used to determine the soil particle structure, and the other was screened through a 0.25 mm sieve and refrigerated at 4° C for pH, TC (total carbon), TN (total nitrogen), TP (total phosphorus) determination. All tests were conducted within two weeks after sampling.

The soil bulk density and soil moisture at 0m were missing due to immersion in water.