Patterns from the past: modeling Public Land Survey witness tree distributions with weights-of-evidence

Abstract

The Public Land Survey (PLS) witness tree data provide one of the few quantitative data sets of pre-and early-European settlement composition and structure of the forests and woodlands in the western United States. However, quantifying the areal extent of individual woody species from PLS records has proven difficult due to the coarse sampling structure of the data. Several attempts have been made to convert the discrete PLS witness tree data into continuous distributions through the use of various interpolation techniques. While these methods may adequately represent the spatial patterns of individual species over large areas, they fail to consider the numerous environmental covariates that can influence the distribution of individual tree species at finer scales. A more statistically rigorous method calls for combining species–environment relationships to estimate the areal extent of individual species from point data. In this study, we utilize weights-of-evidence (WofE), a discrete multivariate method, to estimate the probable historical distribution of six important woody plant taxa of the cross timbers of south-central Oklahoma. We successfully created posterior probability distribution maps for Quercus stellata, Q. marilandica, Q. velutina, Carya texana, C. illinoinensis, and Juniperus spp. Each posterior probability map was classified into four predictive categories, thereby enabling better estimations of the historical distribution of individual taxon from coarse-resolution PLS data. Model validation indicated that the WofE method effectively estimated the posterior probabilities of all taxa under consideration.