The aim of this work was to evaluate the selectivity of herbicides applied in post-emergence on Zoysia japonica Steud (Poaceae) and determine associations with the leaf anatomy of this grass. The experimental design was randomized blocks with four replications. The treatments were the application of the herbicides bentazon (720 g ha -1 ), nicosulfuron (50 g ha -1 ), halosulfuron (112.5 g ha -1 ), oxadiazon (875 g ha -1 ) and 2.4-D (698 g ha -1 ), plus a control treatment without herbicide application. Phytotoxicity was assessed every seven days after application (DAA) of the herbicides until the symptoms disappeared. Foliar anatomical analyses of the leaves in the collected grass were conducted until the 35 th DAA. The quantitative characters of the keel and wing region of the blade of Z. japonica were assessed, as well as the biometric characters, which were submitted to an analysis of variance F test, and the averages were compared by Tukey's test at a probability of 5%. The values of the anatomical characters of the foliar blade were tested by cluster analysis. The application of herbicides did not negatively influence the height of the plants but did reduce their dry mass. Toxic symptoms disappeared after 21 DAA, with the only symptoms of injury observed in plants treated with the herbicides oxadiazon and nicosulfuron. In addition, the cluster analysis indicated the formation of a unique discriminatory group. Thus, the results show that the herbicides applied to Z. japonica were selective for the species. Key words: Plant anatomy. Lawns. Selectivity. Zoysia japonica.
ResumoObjetivou-se com esse trabalho avaliar a seletividade de herbicidas aplicados em pós-emergência sobre grama Esmeralda -Zoysia japonica Steud (Poaceae), em associação com estudos da anatomia foliar dessa grama. O delineamento experimental foi em blocos ao acaso, com quatro repetições. Os tratamentos constituíram da aplicação dos herbicidas: bentazon (720 g ha -1 ), nicosulfuron (50 g ha -1 ), halosulfuron (112,5 g ha -1 ), oxadiazon (875 g ha -1 ) e 2,4-D (698 g ha -1 ), além de um tratamento controle, sem aplicação de herbicidas. As avaliações de fitointoxicação foram realizadas a cada sete dias após a aplicação dos herbicidas (DAA) até o desaparecimento dos sintomas. Realizaram-se análises anatômicas foliares da grama coletando-se folhas no 35 o DAA. Avaliaram-se os caracteres quantitativos da região da quilha e da asa do limbo de Z. japonica, bem como os caracteres biométricos da grama, os quais foram