Phytotoxicity of trace metals in spiked and field-contaminated soils: Linking soil-extractable metals with toxicity

Soil tests have been widely developed to predict trace metal uptake by plants. The prediction of metal toxicity, however, has rarely been tested. The present study was set up to compare 8 established soil tests for diagnosing phytotoxicity in contaminated soils. Nine soils contaminated with Zn or Cu by metal mining, smelting, or processing were collected. Uncontaminated reference soils with similar soil properties were sampled, and series of increasing contamination were created by mixing each with the corresponding soil. In addition, each reference soil was spiked with either ZnCl2 or CuCl2 at several concentrations. Total metal toxicity to barley seedling growth in the field-contaminated soils was up to 30 times lower than that in corresponding spiked soils. Total metal (aqua regia-soluble) toxicity thresholds of 50% effective concentrations (EC50) varied by factors up to 260 (Zn) or 6 (Cu) among soils. For Zn, variations in EC50 thresholds decreased as aqua regia > 0.43 M HNO3 > 0.05 M ethylenediamine tetraacetic acid (EDTA) > 1 M NH4 NO3 > cobaltihexamine > diffusive gradients in thin films (DGT) > 0.001 M CaCl2 , suggesting that the last extraction is the most robust phytotoxicity index for Zn. The EDTA extraction was the most robust for Cu-contaminated soils. The isotopically exchangeable fraction of the total soil metal in the field-contaminated soils markedly explained the lower toxicity compared with spiked soils. The isotope exchange method can be used to translate soil metal limits derived from soils spiked with metal salts to site-specific soil metal limits.