A Sampling Procedure for Estimating Selected Physical Properties of Wood in a Forest

A sampling plan was developed for economically determining strength and related properties of wood at any desired level of precision. These properties were measured, according to the plan, for 40 trees of Tsuga heterophylla (Raf.) Sarg. (western hemlock) from Clatsop County and Mt. Hood National Forest in Oregon. Bolts were taken from each tree, sticks were prepared from the bolts, and specimens for each of the tests from each stick. Specific gravity, static bending, shear parallel to the grain, and compression parallel to the grain were tested. An equation was presented to calculate the optimum number of trees for the 95 percent half-confidence interval of the species mean. In order to estimate averages within 5 percent of the true mean maximum crushing strength parallel to grain, as many as 49 trees should be sampled with one specimen tested from each tree, while 15 trees are sufficient to determine specific gravity. Testing of two randomly selected specimens from each tree results in fewer trees being required for a given precision. An estimated cost analysis indicated that the selection of two specimens from each tree minimized the costs for estimating maximum crushing strength and radial shear while minimum costs for measuring the other properties were obtained by testing one specimen from each tree.

You must be logged in to download any documents. Please login (login accounts are free) or learn how to Become a Member