Phytoremediation of Soils Contaminated with Wood Preservatives: Greenhouse and Field Evaluations

Abstract

ABSTRACT Phytoremediation was evaluated as a potential treatment for the creosote-contaminated surface soil at the McCormick and Baxter (M&B) Superfund Site in Portland, Oregon. Soil at the M&B site is contaminated with pentachlorophenol (PCP) and polyaromatic hydrocarbons (PAHs). Eight individual PAH compounds (containing four to six aromatic rings) were included in the investigation. Greenhouse and field studies were carried out using perennial ryegrass (Lolium perenne). The following three treatments were compared in both studies: T1 = planted-amended soil; T2 = unplanted-amended soil; and T3 = unplanted-unamended soil. The amendments were mineral nutrients and dolomite, which was used to raise the acidic pH of the soil. Contaminant concentrations in the soil were measured initially and at regular intervals for several months. In the greenhouse study, the concentrations of certain contaminants decreased as a function of time. Thus, PCP, fluoranthene, pyrene, chrysene, and benzo(k)fluoranthene appeared to undergo biodegradation in all of the treatments. On the other hand, certain larger molecular weight PAHs were relatively recalcitrant. These “recalcitrant PAHs” included benzo(b)fluoranthene, benzo(a)pyrene, benzo(g,h,i)perylene and indeno(1,2,3-cd)pyrene. Statistical methods were used to compare the concentrations of the more easily biodegraded contaminants in treatments T1, T2, and T3. The statistical analysis was facilitated by normalizing the contaminant concentrations relative to the sum of the recalcitrant PAHs in the same sample. Thus, ratios were created that could be compared directly to benchmark values indicative of the contaminant at the beginning of the study. In the greenhouse study, statistically significant differences between T1 and T2, across all treatment times, were obtained for fluoranthene, pyrene, and chrysene (p ≤0.05), suggesting enhanced rhizosphere biodegradation for these compounds. Significant differences between T2 and T3 were obtained for pyrene (p ≤0.03), indicating that nutrients stimulated the biodegradation of this contaminant. Although the greenhouse study was carried out with a well-mixed soil sample from the M&B site, an extremely uneven distribution of contaminants was encountered in the field study. The resulting scatter in the field data made comparisons difficult, and treatment-specific effects observed in the greenhouse study were not statistically significant in the field study. However, analysis of the normalized data from the field revealed the same time-dependent decreases in contaminant concentration as observed in the greenhouse study.