Determinação estatística do número de amostras simples de solo por área para avaliação de sua fertilidade

Resumo

This study reports the determination of the number of single samples used to constitute a compgsed sample/ha, through the statistical formula n = (t a .CV/f)2.

The number of single samples was computed for several percentages of variation around the mean at t = 5%.

The soil samples were collected form level ground, red yellow podzolic-latosol,located in Viçosa country, in the State of Minas Gerais. One hundred single samples were drawn from each area.

The sampling sites were established by using the intersec- tions of a 10 x 10 meter coordinate system. Each sample was assayed for pH, P, Ca + Mg, Al and K.

The results obtained have shown that

1. The variation for the components Ca + Mg and K was hi- gher in the slope, and for P and Al, it was higher in the flat ground as for the pH no significant variation was found;

2. The samples drawn form the lower sections of the slope yielded higher values for pH, Ca+Mg and P while for K and Al variation was not related to slope.

3. The CV values for the grups studied were as follows: level ground: P > K > Al > Ca+Mg > pH; slope: K > P > Ca*Mg > AI > pH and

4. The number of 30 single samples/ha for these soil types in the region, in relation to the allowed variation around the mean, gives a precision of 1,7; 51,5; 29,9; 6,2 and 13,5 for the pH, P,K, Ca Mg and Al, on the slope and 1,4; 22,3; 43,1; 15,7 and 13,0% on the level ground respectively