Plot size and number of replications for productive traits of cotton cultivars

Abstract

The objective was to determine the optimal plot size and
number of replications to evaluate productive traits of
cotton (Gossypium hirsutum L.). Fifteen uniformity trials
were conducted, each trial (48 m2) divided into 48 basic
experimental units (BEU) of 1 m × 1 m, totaling 720 BEU.
Harvesting was performed manually and in stages. The
optimal plot size (Xo) was determined using the
maximum curvature method of the coefficient of
variation model. The means of the statistics were
compared between cultivars using the Scott-Knott test
and between harvests using the Student's t-test, both via
bootstrap (10,000 resamplings and ? = 0.05). The number
of repetitions was calculated by iterative process for
completely randomized design and randomized blocks,
in combinations of i treatments (3 to 50) and minimum
differences between treatment means to be detected as
significant at 5% probability expressed as a percentage of
the experiment mean (10% to 50%). The optimal plot
size to evaluate cotton productivity traits is 8.62 m2.
Eight replications are sufficient to identify, at 5%
probability by Tukey's test, differences between
treatments means of 38.86% of the experiment mean in
trials with up to 50 treatments.