DELINEATING A WETLAND RESTORATION BOUNDARY USING FIELD-BASED AND GEOSTATISTICAL METHODS

Mountain bogs improve water quality and provide critical habitat to endangered species and are one of the rarest and most threatened types of wetlands in the southern Appalachian region. As a part of a range of wetland restoration projects proposed by the National Forest Service within the Paint Creek watershed within the Cherokee National Forest, the 81-acre Devil’s Kitchen Branch Bog in Greene County, TN is one area where boundary expansion is of interest. The aim of this research was to produce a probability map for wetland restoration and possible expansion using soil hydric properties in a 40-acre plot within the bog. Soil cores 0.4 m in length were extracted from 58 locations and evaluated in situ for hydric soil indicators. Soil moisture content and grain size distribution were measured in house while chemical properties like pH, Organic Matter, estimated Nitrogen Release, Bray II Phosphorus, Exchange Capacity, % base saturation of Cation, Mehlich III Extractable P, Mn, Zn, B, Cu, Fe, Al, S, Ca Mg, K, and Na were assessed by an outside laboratory. Topographic characteristics (slope, elevation, and distance to streams) were extracted for each sample point using TN LiDAR elevation data in ArcGIS. Hydric soils were present in 20 of the 58 locations and were classified as clay, clay loam, or silty clay loam. A binary logistic regression model developed to predict the presence of hydric soils retained iron and distance to streams as the best predictors. These two variables were interpolated using ordinary kriging, converted to rasters, and combined according to the regression model equation using the raster math calculator in ArcGIS Pro. The resulting probability map for hydric soil was used to propose a boundary for future wetland restoration and expansion. A probability threshold was chosen at 0.3, based on the mean of the data and the wetland restoration boundary was delineated using this 0.3 probability contour. A 16 acre area within the 40-acre plot is expected to develop successful restoration based on the hydric soil properties. This data driven approach permits flexibility in selection of a hydric soil probability threshold to delineate the wetland restoration boundary in similar mountain bogs.