Quantifying soil lead distribution by landscape variables on an urbanization gradient in Shanghai, China

Abstract:

How to quantify the spatial distribution of soil Pb in urban area is a pending question for urban soil Pb research. In this study, along with an urban-rural gradient, variables of landscape metrics, demographic and economic attributes, traffic volumes, and road density are included to model the soil Pb distribution. Methods of soil Pb and landscape variables spatial interpolation, moving windows zonal statistics, and linear regression analysis are used to quantify the soil Pb distribution. Results show that Pb is not only affected by traffic, but also by the urbanizing rate. The stepwise linear regression model reveals that landscape shape index (LSI) and road density (RD) could account for 69% of soil Pb spatial variation, in which the accounted percentage of LSI, and RD are 56% and 13% separately. This indicates that more fragmented and more complexity of the landscape, the higher the road density, and then the higher the Pb value. Our research demonstrates that the gradient analysis is workable to illustrate the spatial heterogeneity of urbanization and the associated soil Pb distribution.