Four field experiments were conducted (two in Santo Antônio de Goiás and two in Santa Helena de Goiás - Brazil) in 1998/1999 and 1999/2000, to evaluate the selectivity of upland rice to herbicides at different growth stages. The cultivars tested were Primavera, Maravilha and Canastra and the herbicide treatments were metsulfuron-methyl (2,4 g a.i. ha-1) at 10 and 20 days after emergency (DAE); 2,4-D amine (335, 502 and 670 g a.i. ha-1), at 10, 20 and 30 DAE, fenoxaprop-p-ethyl (41,4 g a.i. ha-1) at 10, 20 and 30 DAE and clefoxydin (120 g a.i. ha-1) at 10, 20 and 30 DAE. The results showed that the selectivity of the herbicides depends on the cultivar and rice growth stage at application time. According to the results, for the cultivar Primavera, the applications of metsulfuron-methyl at 20 DAE and fenoxaprop-p-ehtyl and clefoxydin at 30 DAE are recommended. The herbicide 2,4-D amine is not recommended for this cultivar. For the cultivar Maravilha, metsulfuron-methyl and fenoxaprop-p-ethyl can be applied at any growth stage. The herbicide 2,4-D amine and clefoxydin are recommended for application at 30 DAE. The cultivar Canastra showed high tolerance to all herbicides at any growth stage.
rice; herbicides; Oryza sativa; selectivity; phytotoxicity
