Differences in fixed-radius plot sample size estimates when using the inherent variability among diameter classes in stand tables rather than average valuesThis article is part of a larger document. View the larger document here.
As part of many forest inventories, sample sizes needed to meet some desired level of precision are estimated. In many timber inventories, based on estimates of the inherent variability (e.g., coefficient of variation, CV), sample sizes are estimated for average total merchantable volume or for the most valuable product class such as average sawlog volume. However, often foresters are equally interested in producing accurate and precise stand tables. Stand tables contain number of trees by diameter and/or height classes while stock tables contain amount of volume by diameter and/or height classes. The most important component of most timber prospectuses in the Southeastern United States is actually stand and/or stock tables, not average values. This study attempted to quantify differences in sample size estimates of fixed-radius plots when using variability among diameter classes in stand tables as opposed to the variability in average values in the sample size equation. Several methods were proposed to quantify the CV among diameter classes within stand tables. Fixedradius plots of varying sizes from many inventories conducted across the Southeastern United States and a few from other parts of the country were examined. In most cases, sample sizes needed for stand tables were at least twice as large when compared to estimating average trees per acre. However, when using Method Four at an 80 percent probability, sample size estimates (n = 27) were similar to sample sizes for average volume per acre at 95 percent probability (n = 29).