This study aimed to evaluate the herbicidal activity of picloram on the biomass of the remediation plants Eleusine coracana and Panicum maximum after cultivation in a soil contaminated with this herbicide. These species were grown in three soils, differentiated based on texture (clayish, middle, and sandy, with 460, 250, and 40 g kg(-1) of the clay, respectively), previously contaminated with picloram (0, 80, and 160 g ha(-1)). After 90 days, the plants were harvested and an extract was produced by maceration of leaves and stems of these plants. It was applied to pots containing washed sand, comprising a bioassay in a growth chamber using soybean as a bioindicator for picloram. Soil and plant samples were analyzed by HPLC. The results showed the presence of picloram or metabolites with herbicidal activity in the shoots of E. coracana and P. maximum at phytotoxic levels with regard to soybean plants, indicating that they work only as phytoextractors and that the presence of straw on the soil surface can promote recontamination within the area. It is not recommended to cultivate species susceptible to picloram in areas where it was reported remediation by E. indica and P. maximum and still present residues of these species.
Keywords: African finger millet; Glycine max; Tanzania; bioremediation; carryover; crop-livestock integration system.