Assessing the representativeness of the oldest permanent inventory plots in Northern Arizona ponderosa pine forests

A network of permanent plots established between 1909 and 1913 (the Woolsey plots) contains the oldest measured data in northern Arizona ponderosa pine forests. These forest inventory data offer a unique opportunity to reconstruct pre-settlement reference conditions, as well as detect and quantify changes in southwestern forest structure and composition. However, the selection of plot locations in the early 1900s followed a subjective nonrandom approach. To assess the applicability, or inference space, of results obtained from these historical plots, we compared their environmental characteristics (terrestrial ecosystem unit [TEU, based on a U.S. Forest Service (USFS) ecological classification system], site index, elevation, insolation index, and soil parent material) as well as contemporary forest structure (trees per hectare, basal area, and quadratic mean diameter) with two large inventory samples: USFS Forest Inventory and Analysis (FSFIA) and Arizona State Land Department Continuous Forest Inventory (AZCFI). Analytical methods included multivariate permutation tests, ratios of variance, and Kolmogorov-Smirnov two-sample tests. Results indicated that the Woolsey plots (1) were neither historically nor contemporarily representative of the entire study area because of environmental and current forest structural differences with respect to the FSFIA and AZCFI and (2) may be considered historically representative of their corresponding TEUs. Our study supports the use of TEUs for defining the applicability of information obtained from the Woolsey plots.