O MANEJO DAS CULTURAS INTERFERE NO ERRO EXPERIMENTAL

The present study aimed to identify some common techniques of management which may affect the experimental error of competition assays in com, soybean, rice, wheat, bean and oat. A total of 1920 assays were developed in Rio Grande do Sul State, from 1987 to 1995. For each crop, the assays were grouped based on similarity of management. For each new group the value of the mean square of the average error was calculated, considering the error degrees of freedom. These values were used to apply F test to compare two variations. The effects of crop management on the experimental error were analyzed based on comparisons among specific groups. It can be concluded that seeds treatment in soybean and wheat, com trimming, use of pre-genninated seeds in rice, insect and weed control in soybean reduce the experimental error, generally. On the other hand, cover fertilizer application in soybean, and insect control in com and wheat usually increase the experimental error.