An inverse quadratic model to determine the size and shape of plots for maize and cotton intercropped

Abstract

One trial with maize and cotton intercropped was performed. To measure the relationship between the coefficient of variation of the plot and its size, two models (quadratic and inverse quadratic) were considered. The inverse quadratic model showed to be better, with coefficients of determination (R²) equal to 0.988 for maize and 0.990 for cotton. The nature of the fitted surfaces were studied through contour plots, and the estimated plot size, for a coefficient of variation about 15%, is smaller when we use the inverse quadratic model.