Proposal of methods for determining the optimal size of experimental plots

Abstract

Adequate planning of experiments is extremely important and for this to occur, the appropriate choice of plot size is essential. Thus, determining the plot size seeks to increase experimental precision, since precision decreases when a plot size smaller than the ideal is chosen; on the other hand, when opting for plot sizes larger than the ideal, the researcher may use more resources than necessary, as well as increasing the time to set up the experiment. Therefore, this work aimed to propose two new methods for determining the optimal size of experimental plots, which were applied in experiments with yellow passion fruit in the field, and compare them to the modified maximum curvature method. The maximum curvature method of the modified Vx function uses the equation proposed by Smith and the maximum curvature method of the modified CVx function uses the equation proposed by Thomas. The modified maximum curvature method of the CVx function proved to be suitable for estimating the optimal size of plots in experiments with yellow passion fruit. It is recommended to use the optimal plot size of 4 plants per plot for variables related to the fruit, 5 plants per plot for variables related to the pulp and for the production variables of 9 plants per plot.