The objective of this work was to evaluate the procedure adaptive sampling procedure used in forest inventories for species with low population density and aggregated spatial distribution pattern, in relation to procedures normally used in forest inventories. Thus, a population was simulated in a 90 ha forest, subdivided into 100 units (N = 100) with areas of 9,000 m² each, presenting 44 individuals of a given species, and submitted to three sampling procedures: simple random sampling; systematic sampling and adaptive cluster sampling, with initial samples selected through simple random sampling and systematic sampling. After analyses, it was verified that the systematic sampling was the best procedure to estimate the total number of individuals of the studied species. Besides, it was observed the need of further investigation on the effect of the size and form of plots, the aggregation scale and size of the population, and their combinations, on the efficiency of the estimators of the adaptive cluster sampling.
Systematic sampling; random sampling; adaptive cluster sampling
