Ecological classification systems have been developed for many ecosystems around the globe. Classification systems provide a set of probabilities concerning biotic components based on some observable conditions on the landscape. The National Hierarchical Framework of Ecological Units (NHFEU) has been adopted by the United States Department of Agriculture Forest Service (USFS) to characterize their publicly managed forests in the U.S.
A static model was created in a GIS to map Ecological Land Types (ELTs) across the Shawnee National Forest in Southern Illinois, USA. Six ELTs were characterized in empirical field data across seven subsections (NHFEU ecological unit), then modeled and mapped based on digital topographic and soils datasets. The final model yielded more 150 thousand polygons covering the 660 thousand hectare study area.
Witness tree data collected in the early 19th century by the U.S. Department of Interior General Land Office Survey (GLOS) were digitized at nearly 6 thousand points in the study area. These witness trees provide evidence of pre-European settlement forest conditions in the Shawnee National Forest, and are matched in this analysis with modeled ELTs.
Results of this research will ultimately be used as a management tool for USFS personnel in the Shawnee. There are two specific objectives of this project. First, a detailed examination of the GIS model construction, parameters utilized, and sensitivity analyses are provided to substantiate the link between drawn ecological conditions and the modeled digital landscape. Second, a fuzzy set accuracy assessment was constructed using forest species and diameter of witness trees at GLOS points, as a validation tool. Results from the analysis demonstrate that forest composition in Southern Illinois has undergone substantial changes in the past two centuries.