EVALUATION OF NO SAMPLING ERRORS OF THE VARIABLES LOCATION (x_i, y_i ) AND DIAMETER (cm) IN COMMERCIAL INVENTORIES AT 100% IN THE NATIONAL FOREST OF TAPAJÓS, PARÁ, BRAZIL

ABSTRACT

This research originated from a base of data of 2 forest inventories at 100% carried out in 1984 and 2000 of a same area of 576 hectares of primary tropical forest, located in the National Forest of Tapajós (FLONA TAPAJÓS), Belterra, Pará, Brazil. The objective was to evaluate the levels of errors in obtaining the tree location (x_i, y_i ) and diameter at breast height (DBH)-cm, basing on the commercial inventories at 100% accomplished in the two ocasions. Circumference at breast height (CBH)-cm, were measured with metric tape in the two inventories. The commercial heights in 1984, were measured with the hipsometer of Weise and, in 2000, estimated visually with the use of sticks; minimum DBH in 1984 were of 55cm for all the species and, in 2000, minimum DBH of 35cm, was adopted. The global medium error of measurement of DBH (cm) for the 6 analyzed species ranged from 2 to 5%, and in terms of diametric class it varied from 8 to 17%. This research showed also that the global level of the average error of estimated of the distance until 5 meters was approximately 14%, being that 41% of the error was refered to distances that ranged from 5 to 10 meters, and that about 60% of the error concentrated in the distance over 10 meters. The evaluation of the average error of estimated of the coordinates (x_i, y_i ) showed that the used method in 1984 and the adopted method by TREVISO in 2000, produced considered satisfactory result for the planning of the forest management.

Key words:
Systematic error; complete enumeration; amazonian forest