Topsoil samples (0-10 cm) from four bioretention systems with different land-use types were collected, including parking lot, roadside, residential area, and industrial park systems. The accumulation contents of As, Hg, Cd, Cr, Pb, Cu, Ni, and Zn were analyzed and evaluated, as were the influencing factors, pollution level, potential ecological risk, and human health risk. The results showed that there were significant differences in the accumulated contents of eight heavy metals. The average ω(As), ω(Hg), ω(Cd), ω(Cr), ω(Pb), ω(Cu), ω(Ni), and ω(Zn) were 8.92, 0.25, 0.10, 31.56, 14.81, 21.27, 23.69, and 62.75 mg·kg-1, respectively, and the average contents of As and Hg were 1.26 and 5.21 times the soil background values, respectively. Correlation analysis showed that the contents of the eight heavy metals were positively correlated with soil organic matter, pH value (except Hg), and phosphorus content (except As). The results of the Nemerow Comprehensive Pollution Index and Hakanson Potential Ecological Index showed that the pollution level and ecological risk of the other seven heavy metals were relatively low, whereas the pollution level and ecological risk of Hg reached the level of severe pollution and strong ecological risk, respectively. Affected by Hg, the comprehensive pollution level and ecological risk were relatively high; thus, Hg was a potential threat to the soil environment. The non-carcinogenic risks of heavy metals in the four systems were acceptable, but the carcinogenic risks were all beyond 10-6 though lower than 10-4, which indicated that these four systems had a certain carcinogenic risk, in which As was the main risk factor. Among these four land-use types, the accumulated pollutant contents, pollution levels, ecological risk, and human health risk of parking lots and roadside bioretention systems were much higher than those of residential areas and industrial parks.
Keywords: bioretention systems; distribution characteristics; heavy metal accumulation; maintenance suggestions; risk assessment.