Publications

We created and tested several predictive occurrence models to determine the ecoregional distribution of native plant taxa and applied the preferred model to plants of interest to the Bureau of Land Management, particularly those that may be used in stabilization and rehabilitation activities on federal lands. The models were designed to predict the ecoregions in which plants occurred based on county records for those plants, with particular attention to cases where counties were not wholly encompassed within a single ecoregion. The models were used to place plants into three confidence categories for each ecoregion: presumed or probably present, possibly present, and presumed absent. Across the area we studied, our preferred model placed an additional 6.9 percent of all possible taxon-ecoregion combinations in the highest confidence category compared to limiting this category to plants occurring in counties that were wholly encompassed within a single ecoregion. This was the case even though we chose the preferred model primarily because it was the most conservative (it minimized cases where a taxon determined to be presumed/probable in an ecoregion was actually absent).