2.2 Soil sampling and analysis
In each county, all of the sites with the natural vegetation isLeymus chinensis (Trin.) Tzvel. dominated steppe, which originally covered the entire study area. After examining the land-use history, 2-4 locations with the same grassland and cropland characteristics and similar topography were chosen at each site. Guyuan, Huade and Kangbao were assigned four, three and two locations, respectively, and pairs of sampling points (grassland vs. cropland) were identified at each location. The number of locations differed among the sites because some sites were not large enough to accommodate four pairs of sampling points. The distance was ≥ 2 km between every two locations in the same county. The distances were more than 100 m between adjacent pairs and less than 50 m between grassland and adjacent cropland points. Ten soil cores were sampled randomly using a 3.5-cm diameter stainless steel auger and were combined to create replicates for the 0–10-, 10–20-, and 20–30-cm layers from each sampling point. The soil samples were sealed in plastic bags and transported to the laboratory.
After the visible plant residues and stones were removed, the sampled soil was passed through a 2-mm mesh and divided into two subsamples. One subsample was stored at -20°C and was used to measure the soil microbial communities and soil moisture, soil NH4+-N and NO3--N, while the other subsample was air-dried to analyze pH, soil aggregate, SOC, soil total nitrogen (STN) and soil total phosphorus (STP) concentrations. A 10 g soil from each composite subsample was dried at 105°C for soil moisture measurement. Soil pH was measured in a 1:5 soil:water suspension (v/v) with a digital pH electrode (Shanghai, China). Soil aggregates were obtained by wet sieving and further fractionated five size classes (0.053–2mm) following the method described by Elliott (1986). The 100g air-dried soil samples was submerged in deionized water for a duration of 5 minutes on a set of sieves with 1-, 0.5-, 0.25-, 0.106- and 0.053-mm meshes and then shaken with an amplitude of 3cm and a frequency of 50 min-1. After wet sieving, all aggregates were oven-dried at 65 oC until weighted and then contained in a shallow pan. The SOC concentration was determined following the Walkley-Black dichromate oxidation procedure (Bao, 2000), and STN was measured by Kjeldahl digestion in a 2300 Auto Kjeltec Analyzer Unit (FOSS, Sweden). For STP concentration analysis, the dry samples were digested with H2SO4 and H2O2 at 380°C for approximately 3h, and then analyzed by the vanado-molybdate method with a spectrophotometer (UV-2550; Shimadzu, Kyoto, Japan). Soil NH4+-N and NO3--N content were analyzed by extracting inorganic nitrogen at 120 rpm for 2.5 h with 60 ml of 2 mol/L KCl from 10 g fresh subsamples, and determined by a Flow-Solution analyzer (Flowsys, Ecotech, Germany).