Traffic emission dominates the spatial variations of metal contamination and ecological-health risks in urban park soil.

Metals in urban park soil are closely related to traffic emissions, which adversely affect soil quality and human health. However, little is known about the quantitative impacts of traffic on the spatial variations of metals in park soil after the banning of leaded gasoline. Herein, concentrations of Cu, Pb and Zn in surface soil of four recreational parks of Sydney (Ashfield, Robson, Lamberts and Leichhardt) were measured to evaluate their spatial characteristics in contamination, ecological and health risks and relationships with traffic emissions. Contamination of metals are assessed by contamination factor (CF). Normalized metal concentrations (<63 μm) in the park soil were 24-614, 23-3520 and 99-3060 mg kg for Cu, Pb and Zn, respectively, and CFs ranged from 1.4 to 207, whose variations inter- and intra-parks were related to traffic volumes. Traffic emission accounted for 72-84% of metals contamination in soil of Ashfield, Robson and Lamberts by sites, whereas the values were 25-70% for Leichhardt due to the absence of a surrounding arterial road. In Ashfield and Robson Parks, metal concentrations from traffic decreased exponentially with distance from arterial roads. Metals in Lamberts Park and in areas near arterial roads in Ashfield and Robson Parks may raise ecological risk, and traffic sources contributed to 61-81% of the risk. The ranges of ecological risk zones away from arterial roads and average daily traffic volumes showed an exponential relationship. Copper and Zn in soil of the four parks have no non-carcinogenic health risk for children and adults, and Pb has negligible health risk for adults. Lead in Lamberts Park and in sites near arterial roads of Ashfield and Robson Parks may raise non-carcinogenic risk for children (HI > 1) due to traffic emissions. These results emphasize the remarkable influence of traffic emissions on urban soil metal, which can be predicated quantitatively by traffic volume.