Masters Theses

Date of Award

5-1994

Degree Type

Thesis

Degree Name

Master of Science

Major

Forestry

Major Professor

John Rennie

Abstract

Landscape ecosystem classification is gaining acceptance by the USDA Forest Service as essential to ecosystem management of public lands. Such a management approach has previously been rejected in the southern Appalachians in favor of traditional methods such as site-index. A long history of natural and man-made disturbance combined with highly complex interactions between climate, geology, topography and soils has made ecological classification in this region inherently difficult. Consequently, very few classification efforts have integrated vegetation, landform and soils in identifying ecologically equivalent sites.

An ecological classification for the 43,800 hectares comprising the Foothills section (300 m to 610 m elevation) of the southern unit of the Cherokee National Forest is described. Vegetative cover, landform and soils data were obtained from sixty 0.04 ha plots located in stands representing late successional stages. Vegetation data were grouped by dominant cover type utilizing (1) agglomerative, hierarchical clustering and (2) detrended correspondence analysis. Detrended canonical correspondence analysis (DCCA) was used in conjunction with stepwise discriminant analysis (SWDA) to identify patterns in species composition explained by environmental variables.

Four site units were identified: Eastern Hemlock - Sugar Maple - American Beech - White Ash, Eastern Hemlock - Rosebay Rhododendron, Chestnut Oak, and Chestnut Oak - Scarlet Oak - Red Maple. A recently developed "Landform Index" that quantifies slope type and degree of protection by adjacent land masses was identified through DCCA and SWDA as the most important predictor of species variation between site units. The strength of the correlations between elevation and several soil thickness variables with DCCA axis 1 indicated vegetation varies along a moisture gradient. No ecological meaning was attributed to the second axis.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS