O plano experimental e a variação da heterogeneidade da área experimental na cultura do milho.

Abstract

Public and private companies are working on genetic improvement of maize in Brazil. This generates new cultivars to be recommended to rural workers yearly. In the state of Rio Grande do Sul, southern Brazil, before being recommended to rural workers, cultivars are evaluated by means of preliminary and state assays with low experimental precision. It is evaluated according to the magnitude of experimental error, measure influenced by many aspects, including the heterogeneity of the area where experiments are installed. With this in mind, the aim of the present research was analyzing some characteristics of the experimental plan of competition assays in maize crop, in order to improve its quality. For this purpose, the class of higher frequency of the heterogeneity index for the experimental area of the assays and its effects on the experimental precision were analyzed. Moreover, this research aims at defining the number of replications and the plot size for the assays. Ninety-seven competition assays of maize crop were carried out, from 2002 to 2005, in different locations of the state. Experimental plots consisted of two rows, each one with five meters in length, spaced 80 centimeters apart, following a randomized complete block design, with three replications, for all locations. The use of the design was correct in 40 (41.2%) out of ninety-seven assays. In 85% of the assays, b values for the considered years, locations and or both of them remained constant. In 14 (14.4%) out of ninety-seven assays, b values remained in class b ≤ 0.3; in 53 (54.6%) out of ninety-seven assays, they remained in class b ≥ 0.7; and in 30 (30.9%) out of ninetyseven assays, b values remained in class 0.3 < b < 0.7. Therefore, it is necessary to adjust the experimental plan with respect to the design, number of repetitions and size of plot to each evaluated location.