The objective of this work was to estimate plot size for evaluation of the grain yields of wheat, under irrigated conditions, in no-tillage system and conventional tillage. Each experiment was formed by 30 rows of 30 m of length, with spacing of 20 cm among rows and density of 380 seeds/m². The harvest was made in basic units (bu) of a row meter (0.2 m²), being harvested, in each assay, the central 12-meter of the 24 central rows, totaling 288 bu. The plot size was estimated using the methods of maximum curvature, modified maximumcurvature, comparison of variances and method of Hatheway. The plot sizes presented great variation because of the method used in its estimation. Smaller plots, with larger number of replications, were more efficient in the use of the experimental area than larger plots with smaller number of replications. Plots size varying between 1.6 and 2.4 m² of useful area promoted appropriate evaluation of the grain yields in the different studied conditions.
Triticum aestivum; uniformity trial; field plot estimate; variability indexes
