Aire minimum d'échantillonnage ou nombre de points de prisme nécessaires pour établir la structure d'un peuplement inéquienne

Abstract

A complete forest survey was conducted on an 18-ha maple stand mixed with yellow birch and American beech. Diameter at breast height of all merchantable stems was recorded. The location of all those stems was reported on a grid system to allow subsequent simulation of sampling by varying the size of rectangular sample plots. Prism sampling was also carried out in the same area. Statistical testing (χ 2 ) were used to verify if randomized sampling gave the same diameter distribution as the original stand. Both sampling methods were adequate. Rectangular sample plots, 2 m wide by 5000 m long, representing an area of 10 000 m 2 (i.e., 5.6% of the original stand) or 36-point sample (2 points per hectare) gave a good representation of the entire population at a probability level of 90%, with a confidence level of α = 0.05. For both sampling methods, sampling units must be scattered as much as possible in the forest area to be surveyed.