The aim of this research was to evaluate the effect of herbicide mixture (fomesafen + fluazifop) on the soil arthropod communities associated to the common bean in two cropping systems (conventional and no-tillage). The experiment was carried out on Red-yellow Podzol, in Coimbra, Minas Gerais, Brazil. The studied treatments in both cropping system were represented by the mixture of herbicide Robust® (fluazifop + fomesafen) (dosage of 0.8 L ha-1) applied two weeks after planting and control without herbicide application. A complete randomized block design with five replications was used in the experiment. The soil arthropod communities were sampled 1, 8, 21 and 42 days after herbicide application. The data were submitted to the canonical variate analysis (CVA). CVA is an indirect ordination technique that reduces the size of the original data set. The CVA techniques can be used to illustrate graphically the relative positions and the coordinates of the community's structure for each treatment. The differences between the treatments in the abundance of the selected arthropods had been determined by repeated measure analysis of variance considering the sampling data as repeated measure. The canonical variate analysis for cropping systems and herbicide application indicated differences between the arthropod densities in the different treatments considering the composition and the species abundance. In this study, differences in the set data of collected soil arthropods were observed through ordination diagrams under the two cropping systems. Herbicide application affected all arthropods associated to the common beans in both cropping systems, except Solenopsis sp. The cultivation system affected the densities of all species studied.
fluazifop-p-butyl; Fomesafen; Phaseolus vulgaris
