U.S. Environmental Protection Agency


Document Type


Date of this Version



Published in Mutation Research 567 (2004) 227–345. Doi:10.1016/j.mrrev.2004.09.003


The intentional and accidental discharges of toxic pollutants into the lithosphere results in soil contamination. In some cases (e.g., wood preserving wastes, coal-tar, airborne combustion by-products), the contaminated soil constitutes a genotoxic hazard. This work is a comprehensive review of published information on soil mutagenicity. In total, 1312 assessments of genotoxic activity from 118 works were examined. The majority of the assessments (37.6%) employed the Salmonella mutagenicity test with strains TA98 and/or TA100. An additional 37.6% of the assessments employed a variety of plant species (e.g., Tradescantia clone 4430, Vicia faba, Zea mays, Allium cepa) to assess mutagenic activity. The compiled data on Salmonella mutagenicity indicates significant differences (p < 0.0001) in mean potency (revertents per gram dry weight) between industrial, urban, and rural/agricultural sites. Additional analyses showed significant empirical relationships between S9-activated TA98 mutagenicity and soil polycyclic aromatic hydrocarbon (PAH) concentration (r2 = 0.19 to 0.25, p < 0.0001), and between direct-acting TA98 mutagenicity and soil dinitropyrene (DNP) concentration (r2 = 0.87, p < 0.0001). The plant assay data revealed excellent response ranges and significant differences between heavily contaminated, industrial, rural/agricultural, and reference sites, for the anaphase aberration in Allium cepa (direct soil contact) and the waxy locus mutation assay in Zea mays (direct soil contact). The Tradescantia assays appeared to be less responsive, particularly for exposures to aqueous soil leachates. Additional data analyses showed empirical relationships between anaphase aberrations in Allium, or mutations in Arabidopsis, and the 137Cs contamination of soils. Induction of micronuclei in Tradescantia is significantly related to the soil concentration of several metals (e.g., Sb, Cu, Cr, As, Pb, Cd, Ni, Zn). Review of published remediation exercises showed effective removal of genotoxic petrochemical wastes within one year. Remediation of more refractory genotoxic material (e.g., explosives, creosote) frequently showed increases in mutagenic hazard that remained for extended periods. Despite substantial contamination and mutagenic hazards, the risk of adverse effect (e.g., mutation, cancer) in humans or terrestrial biota is difficult to quantify.