Field experiments with perennial arboreal trees usually occupy large areas due to the big plant size and the required spacings. In such situations the use of big plots and the low number of replicates are very common, in an attempt to reduce the experimental area, manpower and costs. This procedure, however, decreases the efficiency of results and difficults the application of statistical tests. This work was conducted in order to demonstrate that using small plots and higher number of replicates increases the trial precision and decreases the experimental area. The association of plot size and number of replicates, minimizing the mean variance of each treatment, allows the use of high number of replicates, increasing the precision of statistical tests, besides a better trial uniformity. The use of smaller plots benefits the increase of replicates number, allows a better estimation of the experimental error, of the treatment and parameters effects and a higher efficiency of statistical tests. Besides, it reduces about 18% the number of plants and the size of experimental area.
perennial plants; experimentation; methods; statistical analysis
