Ajuste de modelo logístico ao acúmulo de massa seca de cultivares de feijoeiro via regressão isotônica

Abstract

The nonlinear models, among them the logistic model, are widely used in the study of plant growth analysis. In the case of bean, which is of great nutritional and economic importance for the country, knowing its way of growth can allow to detect problems in the growth and to increase the production of the grain. However, in some situations, the non -convergence of the non-linear model can occur, making it impossible to know about such growth. Thus, applying techniques that soften the problem of non-convergence are necessary. Among these techniques, we highlight isotonic regression, which is characterized as a data transformation. Thus, the objective of this work is to analyze the growth of two bean cultivars (Ouro Vermelho and Bolinha) at different sowing densities (75, 145, 215, 285 and 355 thousand plants ha -1 ) by adjusting the model nonlinear logistic analysis, to the dry mass data of the different parts of the common bean plant submitted to the application of isotonic regression. The objective of this study is to compute the growth rate of the crop that allows a more comprehensive interpretation of bean growth and to estimate the maximum accumulation of dry mass and age of the plant with maximum growth. The results show that the application of isotonic regression (equal weights) to bean growth data, both in the Ouro Vermelho and Bolinha cultivars, and in the different sowing densities, allows to visualize a good quality in the adjustment of the logistic nonlinear model. The growth rate curve allows to analyze the moments of greatest accumulation of dry mass in the different parts of the bean. The cultivar Ouro Vermelho presents greater asymptotic accumulation of dry mass than the cultivar Bolinha.