The experiment was carried out in columns filled with three different tropical soils, (very clayey soil, sand loam soil and sand clay loam soil) to evaluate ascending and descending movement of imazapyr. After applying 1 kg ha-1 of the herbicide on each column surface, columns were submitted to three conditions: 1) 14 min/35 min rainfall simulation followed by 48 hours rest; 2) 14 min/35 min rainfall simulation followed by 30 days rest; and column inversion after applying imazapyr on the surface, under sub irrigation during 20 days, followed by 10 days rest. After this stage, columns were sectioned every 5 cm of depth. Sorghum bicolor was sowed as bioindicator on each portion and after 20 days of sowing, dry matter of the above ground tissue was evaluated. The higher herbicide descending movement observed was in sandy loam soil, followed by sandy clay loam and clay soil. The ascending movement of this herbicide occurred together with water, causing herbicide distribution in all the column area in sandy clay loam and sandy loam soils, while clay soil showed a smaller ascending movement. Imazapyr presents high mobility on all three soils, together with water movement in the soil profile, both in ascending and descending motions. Such high mobility may cause ground water contamination, and also cycles of permanence of this product in the superficial layers of the soil, depending on humidity availability.
soil behavior; leaching; bioassay; herbicide
