Avaliação de metodologias de adaptabilidade e estabilidade de cultivares milho

Abstract

The aim of this study was to compare three methodologies which analyze adaptability and stability of corn cultivars. It evaluated the productivity of corn cultivars of the national experiment of Embrapa, center, regarding the crops 2004/2005, 2005/2006 and 2006/2007, in order to determine the adaptability and stability, by the methods of Lin and Binns with decomposition (CRUZ; CARNEIRO, 2006), Annicchiarico with decomposition (SCHMILDT; CRUZ, 2005) and by another method proposed based on the Annicchiarico method with decomposition. The use of the average standard error instead of the standard deviation in the calculation of the variability of environments for each cultivar, as well as for the increasing of the confidence level of 75 to 95% was proposed. It was noticed that the results of indication of cultivars by the three methods showed Spearman's high correlation (p < 0.01). The new method presents advantages: it values the deviations between the environments by taking into account the number of environments and presenting the lowest level of significance (allowing lowest type I error). The proposed methodology presents the numerical results of the confidence index equal to the results of the Annicchiarico method with decomposition when the compared number of environments is equal to six.