Abstract: By using the moving windows method, the mean concentration and range of soil heavy metals including Cu, Cd and Pb in forest and builtup patches along an urbanrural gradient in Pudong New District, Shanghai were analyzed based on the sourcesink theory. The linear regression analysis and correlation analysis were applied to distinguish the major factors affecting the spatial distribution of soil heavy metals concentrations. With the selected factors affecting the spatial variation of heavy metals, the priority management areas associated with potential risks of heavy metals were identified by raster layer computation. The results showed that the mean and range values of heavy metals in the forest patches were higher than those in the builtup patches. The mean concentrations of forest soil Cu and Cd decreased significantly at the distance of 27 km from the urban center, while those of soil Pb had a peak value at the distance of 30 km from urban center. The linear regression analysis and the correlation analysis showed that road density, builtup area percentage, and population change index that reflects the industrial development process could significantly affect the spatial distribution of soil heavy metals. A multitarget forest management strategy should be applied to respond to the identified potential priority areas for risk management of heavy metals.

Key words: reproductive toxicity, signal pathway, inhibitor, heavy metal