Effects of soil erosion and land use on spatial distribution of soil total phosphorus in a small watershed on the Loess Plateau, China

Soil total phosphorus (STP) is a major determinant and indicator of soil fertility and quality. Information on the spatial distribution of STP is important in assessing current or potential soil productivity and estimating environmental pollution. In this study, a total of 1015 soil samples, taken at 20 cm intervals up to a depth of 100 cm, were collected from 202 sites representing five land-use types using grid sampling, this was an unprecedented high sampling density and sampling depth study in a small watershed on the Loess Plateau, China. Spatial variation in soil total phosphorus (STP) was studied using classical statistics and geo-statistics. Soil total phosphorus showed moderate variability across the soil profile and a marked dependence on depth, as follows: 0-20 cm (surface layer), 0.86-6.06 g/kg; 20-40 cm, 0.12-5.32 g/kg; 40-60 cm, 0.50-5.78 g/kg; 60-80 cm, 2.10-5.24 g/kg; and 80-100 cm (deepest layer), 2.00-5.60 g/kg. Of the five land-use types, the highest STP was recorded in dam field, followed, in that order, by terraced land, grassland, forest land, and sloped cropland. Redundancy analysis (RDA) showed that sand and SOC were the significant variables affecting STP in study area, which was also affected to some extent by the intensity of soil erosion, decreasing initially and then increasing slightly with depth. The minimum value of STP was affected by erosion intensities between 1000 t/km(2) and 2500 t/km(2) annually. Across the entire soil profile up to the depth of 1 m, STP varied with land use, as follows: grassland, 5.07 kg/m(2); forest land, 4.83 kg/m(2); dam, 4.73 kg/m(2); terraced land, 4.70 kg/m(2); and sloped cropland, 4.67 kg/m(2). STP up to a depth of 1 m was 24,908 tonnes in the study area. Thus, converting farmland to forest land or grassland would increase the levels of STP.