Testing errors in estimating stand diameter distribution of high forests using angle-count sampling

Abstract

The estimation of stem number per hectare is one of the more interesting applications of the so-called «anglecount sampling». Many authors have described theoretical and practical aspects, but an analysis of the error related to diameter distribution reconstruction is lacking. The objective of this work was to test this error in high forests by using a spatial simulation of tree distribution. The real data of stem diameter distribution in 270 forest compartments were used to simulate 30 angle-count sampling points for each compartment with 100 replications, by using different basal-area factors and sampling schemes.
The results showed an error of about 5% for a single forest compartment, but it decreased to less than 3% considering all the compartments. The difference between predicted and real number of trees of different species
was relevant only at very low density (< 10 trees/ha). The total error was negatively correlated with the number of trees of each diameter class. The simulation showed that using 30 angle-count sampling points for each compartment is adequate for estimating stem diameter distribution.