Sampling intensity determination for continuous forest inventory, dense ombrophilous forest, Eastern Amazon, Brazil

The optimized sampling in forest inventories allows improving technologies of sustainable forest management. It was developed a methodology to determine the optimum sampling intensity for continuous forest inventories in order to meet the requirement of the Principle 08 of the Forest Stewardship Council (FSC). We used a census data, where trees equal or greater than 35 cm of dbh were measured and mapped. A cutting plan simulator that combines a genetic algorithm and a Geographic Information System (GIS) was developed. Ten plans were simulated, five with a cutting intensity of 22 m(3) ha(-1) and other five with cutting intensity of 30 m(3) ha(-1). The effective area of logging was divided into 4 690 plots. During the genetic algorithm implementation, two approaches (A and B) were cosidered. Approach A aimed to maximize the number of sampled species, based on pre-defined sampling intensities of 1:1 000, 1:750, 1:500, 1:250 and 1:200. Approach B, in turn, aimed to minimize the number of permanent plots to sample all harvested species. It was not possible to sample all harvested species using pre-defined sampling intensities. To meet the requirements of FSC 8th Principle, the methodology determined an optimum sampling intensity of 1:180 and 1:165 for cutting intensities of 22 m(3) ha(-1) and 30 m(3) ha(-1), respectively. There were no significant differences between the numbers of harvested species in the both cutting intensities.