A cooperative trial to test the use of vegetation to enhance treatment of soils contaminated with weathered petroleum hydrocarbons (PHCs) was initiated in 2002. Experimental protocols were adapted from those developed by the Phytoremediation Action Team of the USEPA Remediation Technologies Development Forum. Four treatments were compared in a randomized complete block experimental design with four replications. Treatments included (1) an unplanted, unfertilized control; (2) an unplanted, fertilized control; (3) a standard cool-season grass/legume mixture composed of a combination of fescue, ryegrass, and a legume; and (4) a locally optimized treatment that included slender wheatgrass, tall wheatgrass, Altai wild rye, red clover,
and Nuttall’s salt-meadow grass. Each trial was monitored for a three growing seasons, with soil sampling conducted at planting and at the end of each growing season. Soils were sampled at two depths (0–15 cm and 15–45 cm). Soil samples analyzed for total petroleum hydrocarbons (TPH), polynuclear aromatic hydrocarbons (PAHs), and CCME PHC-fractions using standard (CCME) methods; plant assessments (above- and belowground biomass) were conducted at the end of each growing season.
Sampling at the time the site was established (t0) yielded a mean TPH concentration of 5551 mg kg-1 (0–45 cm). At the end of the fourth growing season (October 2005) TPH concentrations averaged across the site had been reduced to 1101 mg kg-1 in the surface (0–15 cm) soil and 1942 mg kg-1 in the sub-surface (15–45 cm) soil. Reductions in PHC concentrations in the planted plots and unplanted plots were generally similar, though reductions in the F3 (C16–C34) fraction generally occurred more rapidly in the planted treatments. Likewise, reductions in PHC concentrations were generally greater in the plots amended with fertilizer and compost.
There were no significant differences between the site-specific (USASK) and standard (RTDF) plant mixes; likewise, there were no significant differences between the planted and unplanted treatments. [Note: differences between treatment means were assessed using the least significant difference test at a probability level of P ≤ 0.20.]