Dependência entre o tamanho da unidade básica e as estimativas das dimensões da parcela experimental de batata

Abstract

A uniformity trail with potato was carried out at potato seed production area of FEPAGRO, Julio de Castilhos city, in 2003 agricultural year. Tuber yield of 3456 hills was evaluated with a set of basic units (BU) compound of 1; 2; 3; 4; 6; 8 and 12 hills per BU, in order to study the effect of size of BU about experimental plot dimensions, with width (W) and length (L) restricted to (W ≤ L). Soil heterogeneity index estimated to each BU and a heterogeneity contour map constructed to six hills per BU in the same row. Relationships between numbers of hills per BU and optimum plot size (X0) and correspondent area (S) were through three algebric methods: Meier & Lessman (1971), MO, and two methods related by Thomas (1974), MV and MC. Correlation analysis and path analysis allowed verify direct and indirect dependence among parameters that compound functions that represent each method analyzed and m2. The results showed that same number of hills per BU, MO demanded more area and presented results coherent with practices applications than MV and MC that presented values of S minor that equivalent to one hill, not feasible to application of treatments. As the magnitude of origin as magnitude of curve were determinative of X0. The area demand occurs more intensely as from four hills per BU. The MC method was least responsive to changes. The soil presented more homogeneity within rows, indication of plot orientation in area such as plot shape, narrow and length in the same row.