With the rapid development of industry and agriculture, the soil heavy metal pollution presents an increasing threat to the environment and agriculture products quality. Understanding the spatial distribution of the heavy metal contents is critical for agricultural production and decision-making. Feixiang county is one of the predominant wheat production areas in the low Hai River Plain and its croplands are facing a potentially serious problem of the soil heavy metal pollution. Therefore 100 sampling sites in the topsoil (020 cm) were investigated in this county in July 2004 and the contents of eight heavy metals (Cu, Zn, Ni, Pb, Cr, Hg, As and Cd) in the soil samples were measured. As a first step, the conventional statistical methods were used to explore the normality of the heavy metal levels measured and eliminate possible outliers from the data set. Geostatistics combined with GIS technique were then employed to analyze the spatial variability of the eight heavy metal contents in soils.
The conventional statistical analysis showed that the eight heavy metal contents were normally distributed after excluding the outliers. Their arithmetical means were below the national second-grading standard of environmental quality assessment for soils. By performing experimental semivariograms analysis, it was observed that Pb and Cr exhibited pure nugget effect; Cu and Zn were best described by exponential models; Ni and Cd were best fitted to spherical models; and Hg and As were suitable to linear models with sill. The 2.0 km distance sampling distance was found to be insufficient in this region and it is recommended that sampling density should increase in future studies.
The spatial variability of the contents of Zn and Cd was mainly affected by intrinsic factors (soil formation factors, such as soil parent materials). Soil management practices, such as the use of pesticide and chemical fertilization, appear to be the primary factors influencing the spatial variability of the contents of Cu, Ni, Hg and As. Pb and Cr had a random variability. The difference in the ranges of Cu, Zn, Ni, Hg, As and Cd is large between 2.5 and 13.7 km. The maps of Cu, Zn, Ni, As and Hg were interpolated using the method of ordinary kriging. The contents of Cu, Zn, Ni and As gradually ascended from the southwest to the northeast. But Hg represented an opposing trend in this region. This research will provide theoretical basis for the regional layout of predominant crop products in the local region.