The use of optimal plot size can reduce random error and increase experimental precision. The objective was to identify the best relationship between length and width and optimal experimental plot size in potato with the multiple regression method. Twelve homogeny trials of 12 rows and 24 hills of the potato cultivar Macaca were carried out at the experimental area of the Fundação Estadual de Pesquisa Agropecuária, Júlio de Castilhos, Rio Grande do Sul State. The multiple regression critical point and function type were estimated for the variation coefficient of each designed plot length and width. Cause and effect analysis was done considering optimal plot size as dependent and estimation parameters of the multiple regression function as independent variables. Hill production of independent rows can be used to estimate optimal plot size. The optimal plot size in potato depends upon the linear effect of plot length. The optimal experimental plot size in potato is one row with 24 hills.
Solanum tuberosum L.; experimental precision; response surfaces; path analysis
