Date of Award


Degree Type


Degree Name

Master of Science


Ecology and Evolutionary Biology

Major Professor

H. R. DeSelm

Committee Members

Clifford C. Amundsen, Dewey L. Bunting


The white pine-hardwood type was described by Miller in 1938; however, subsequent researchers of the vegetation of the Great Smoky Mountains National Park have not described this type. A field study of white pine-hardwood vegetation was conducted in the Park from June to October, 1977. The objectives were to relocate and plot sample the white pine-hardwood stands to 1) group samples into vegetation types based upon the importance of white pine and its associated taxa, 2) use quantitative vegetation analysis procedures to describe the white pine-hardwood vegetation types, 3) assess the relationship of the vegetation types to environmental characteristics, 4) examine the successional status of the types, and 5) provide a basis for further ecological studies of these types in the Park.

Data were analyzed from 144 sample plot locations in the western portion of the Park in Tennessee at low to middle elevations (312 to 716 meters). Circular 0.0406 hectare (1/10 acre) plots were located in areas which had been previously mapped by Miller in 1941. Canopy (over 10 cm), sapling (2.5 to 10 cm), subsapling (2.5 cm diameter and one meter high), and herbaceous data were tallied in each plot. Site properties were collected in each plot. Laboratory determinations of soil pH and texture of both the A and B horizons were made.

Canopy data were used to group plots into vegetation types using an agglomerative clustering technique (Orloci, 1967). The seven communities identified were: white pine-Virginia pine, white pine-red maple, white pine-hemlock, white pine-chestnut oak, white pine-white oak, white pine-northern red oak, and white pine types. Relative densities of tree taxa in the canopy, sapling, subsapling, and seedling strata were compared to determine the reproductive success of each type.

Disturbance evidence and historical accounts were analyzed to assess the successional status of the types. Most types have been disturbed through cultivation, logging, and fire although portions of the white pine-chestnut oak, white pine-white oak, white pine-northern red oak, and white pine types occurred on sites of limited human disturbance. The absence of chestnut stumps and the low proportion of sprouts indicated that it had a minimal former presence in the white pine-hardwood types.

Simple linear correlations among and between site, soil, and vegetation characteristics were computed. Significant correlations among soil characteristics indicated that slope angles increased as microtopographic position increased such that steep slope angles occurred predominately downslope. Site and soil correlations indicated that stone volume was negatively correlated with elevation: lower elevations had an increase in stone material. The increased acidity of litter and the increased leaching due to additional precipitation at higher elevations contributed to a decrease in soil pH.

Discriminant analysis of the community types using vegetation data indicated that 95 percent of the types were distinct as classified by the cluster procedure. Discriminant analysis using selected environmental variables indicated that some types were not as distinct environmentally as they were vegetationally. Discriminating factors related to soil moisture conditions such as stone percentages, horizon thickness, and total available water were important on the first discriminant axis. The second discriminant function appeared to be related to both soil moisture phenomena and slope position, which contributed to the concept that the white pine-hardwood types were segregated by available soil moisture. The classification success was low with only 41 percent of the plots correctly classified. The inability of the measured environmental variables to exactly distinguish the types may be attributed to the successional relationships among the types.

Canonical analysis was used to display the arrangement of the seven vegetation types along the first two canonical axes. The centroids of each type were arrayed along the first axis in an order which closely corresponded to the first and second axes in the discriminant analysis of the environmental variables. From the canonical analysis, it is inferred that soil moisture was important in segregating the white pine-hardwood types.

The diameter distribution of white pine may be of considerable value in inferring the age distribution and stand history of a forest. A direct sampling of white pine increment cores was conducted. Regression analysis was used to determine the best fit of the collective white pine-hardwood type as well as the individual types. The white pine-Virginia pine, white pine-red maple, white pine-hemlock, and portions of the white pine type were represented by relatively even-aged stands resulting from large scale disturbances. The white pine-oak types and portions of the white pine type more closely represented all-aged forests.

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