The objective of the present study was to evaluate the row hoe model CHOPSTAR®, the mechanical control of weeds in between the rows of soybean implanted in organic direct sowing system, associated with the camera-guided system. Two experiments were carried, being that in the first experiment an experimental design with sub-subdivided plots with four replicates. The plots corresponded to two soybean varieties ('Embrapa BRS 284' and 'Coodetec CD 216'), the subplots corresponded to the sowing densities of 329.2 and 574.6 thousand plants ha -1 ; and the subsubplots corresponded to four managements of weeds: one mechanized hoe (2 days after sowing -DAS), two mechanized hoes (22 and 47 DAS), one control manually hoed and other control without hoeing. In the second experiment a randomized block design in subdivided plots with three replicates was used. The plots corresponded to two soybean varieties ('BRS 284' and 'DF 2353'), Fey et al.; JEAI, 42(2): 25-36, 2020; Article no.JEAI.55215 26 the subplots constituted of different times when the hoes were made, being: one (14 DAS); two (7 and 21 DAS; two (14 and 28 DAS); three (7, 14 and 28 DAS); besides one control manually hoed up to 28 DAS. In the first experiment it was observed that the automatized hoe was efficient in controlling the weeds and it was necessary only one mechanized hoe (22 DAS) for the 'BRS 284' independent of the sowing density, while for the 'CD 216' the number of mechanized hoes depended on the sowing density. In the second experiment, it was necessary only one mechanized hoe (14 DAS) to avoid production losses in the varieties 'BRS 284' and 'DF 2353'. The automatized hoe is an alternative to control weeds in areas of organic soybean in direct sowing system, however, damages to the crop can occur depending on the sowing density, mainly in the late management of the mechanized hoe.
Original Research Article