Chlorsulfuron and triasulfuron, applied at 15-35 g ha-1, are a challenge to analytical methods because residues are phytotoxic at concentrations of mu-g kg-1 soil. A bioassay based on the suppression of root growth of lentils (Lens culinaris cv. Laird) can detect a minimum concentration of 0.4-mu-g kg-1 soil. The sandy loam, 20 g, is extracted with methanol and extracts are made up in 5 mM CaCl2. The extracts, 20 mL, are added to 250 g of coarse white sand in plastic cups and sown with seven pre-germinated lentil seeds. The cups are enclosed in plastic bags to prevent evaporation, and incubated for 6 days at 20-degrees-C. Herbicide concentrations are brought within the range of sensitivity of the bioassay by dilutions of the extracts. The assay was easier, and more sensitive, precise and reproducible than a direct assay of soil. Herbicides in the field were lost due to photolysis (38-49% in 8 h) and in laboratory manipulations by hydrolysis with high pH extractants, such as saturated Ca(OH)2 (pH 11.2) where the half-life was about 4 h. Hydrolysis at the pH of the sandy loam and the bioassay in coarse sand, pH 5.8, gave a half-life of 176 and 198 days for chlorsulfuron and triasulfuron, respectively, at 20-degrees-C; at 30-degrees-C, the corresponding half-lives were 47 and 27 days. At concentrations for 50% suppressions of lentil roots (ID50 = 0.6-mu-g L-1 in coarse sand), adsorption in the sandy loam, estimated by the Langmuir equation, was 76% of the total residues of both herbicides. Adsorption could account for the differences in the ID50 of lentils grown in sandy loam and coarse sand for triasulfuron, but it was overcorrected for chlorsulfuron. The herbicides should be applied as near to seeding the cereal crop as possible and should be incorporated with the seeding operation so that losses from photolysis and hydrolysis in warm acidic soils are minimized before crop protection is needed.
