Plot size and replication number in potato trials

Abstract

A 'blank experiment' was designed at the Centro Nacional de Pesquisa de Hortaliças (CNPH), Brasília, DF, to estimate the optimum size of experimental plots in potato (Solanum tuberosum L.) trials. The cultivar 'Bintje' was used, in a 0,80 x 0,30 m spacing. Sixteen rows of 112 plants were harvested on an individual plant basis. The coefficient of soil heterogeneity proposed by Smith was estimated according to Hatheway & Williams, and the plot sizes were estimated by Hatheway's method. Results are presented in graphs as a function of the number of blocks, of the true differences to be detected, and of the types I & II errors, for experiments with various treatment numbers. The ideal plot size is chosen for each situation, as function of the parameters enumerated above.