Scale Characteristics and Effects on Spatial Variability of Soil Available Nutrients

Abstract

Abstract.The spatial variability of farmland soil nutrients on different scales is important for farming as it forms the basis for the efficient utilization of soil nutrients and precision fertilization. Survey points were distributed throughout the study area on three different scales (county, field, and block). Research on the scale effect of the spatial variability of available nitrogen (AN), available phosphorus (AP), and available potassium (AK) involved a combination of classical statistics, geostatistics, and Geographic Information System (GIS) techniques. Results indicated that the three kinds of nutrients presented moderate variation intensity on the three scales. All of the nutrients tested exhibited strong spatial autocorrelation, indicating that spatial variability was primarily affected by structural factors, including climate, soil type and topography. As the sampling scale decreased, the nutrients showing weak variation at the large scale exhibited great variation at the small scale; the spatial autocorrelation of these three nutrients first became greater and then weakened; the distance of the spatial autocorrelation shortened gradually. Furthermore, the patch density value of the soil nutrient map increased, which indicated that the distribution of nutrients tended to be more fragile. When combined, sampling methods on the multi-scale allowed us to obtain real and systematic soil information. This study explored scale characteristics and the effects of spatial variability with regards to the primary nutrients available on farmland and provided a theoretical basis to effectively understand the nutrient status of regional farmland and improve the efficacy of soil sampling. Keywords: Multi-scale, Geostatistics, Patch density, Fractal dimension, Kriging interpolation.