Spatial variability of soil nutrients in a hill-pond landscape

Understanding the spatial variability of soil nutrients and its relationship with ecosystem biogeochemical cycles is important for managing and utilizing land resources. The objective of this study was to characterize the spatial variation of soil nutrients in a hill-pond landscape. This landscape is composed of four patch types: forest in the upper slope, a fish pond at the bottom of the watershed, a fruit garden and grassland in between the forest and fish pond. Soil samples at 0～10 cm and 10～20 cm depths were collected from each of the four patch types. Total carbon content (TC), total nitrogen (TN), and total phosphorus (TP) were analyzed. The spatial variability of the nutrients at the two soil layers was analyzed using geostatistical methods and Geographical Information System (GIS). Semivariance analysis showed that the effective range of TN (164.1 m) was larger than TC (102.4 m) and TP (89.2 m) at 0～10 cm depth. The effective range of TN increased from 164.1 m at 0～10cm to 1242.7m at 10～20cm, whereas those of TC and TP remained almost unchanged. The ratios of nugget to sill for TC, TN, and TP were 5%, 6%, and 49%, respectively. In terms of the ratio of nugget to sill, the distribution pattern of TC was similar to TN. The content of each of the three types of soil nutrients were classified into 5 levels, each of which occupied a certain amount of landscape area. TC showed less variation (10.54%～23.15%) than TN (5.79%～32.73%) and TP (1.80%～42.06%). In general, the semi-variogram for the models three soil nutrients fit better at the landscape than at the patch type scales. TC and TP were better fitted by spherical model, or exponential model was better for TN. The spatially structured variance (the variance due to the location of sampling sites) accounted for a large proportion of the sample variance for OC(95%), TN(94%) and TP(53%). According the effective range, the efficiency of the sampling can be improved. The scale of the sampling is 165m(TN), 103m(OC) and 90m(TP).To enhance analyzing precision we need increase the sample number, especially for the orchard.