Date of Award


Degree Type


Degree Name

Doctor of Philosophy



Major Professor

Jerry Olsen

Committee Members

Edward E. C. Clebsch, M. E. Springer, L. F. Seatz


Six "beech gap" forests in the Great Smoky Mountains between Tennessee and North Carolina were sampled for pattern of stem distribution and for stem size and survivorship of Fagus grandifolia. Several methods for approximating the adequacy of number of plots for stem count were compared in terms of relative efficiency. Stem count and basal area of all species were similar between north-facing and south-facing stands, and between more east-facing and west-facing sides of the same stand. Mean density of live beech ranged from 3705 to 7835 trees per hectare, and basal area ranged from 16.1 to 33.0 m2 per hectare. For living and dead stems of all species mean density ranged from 6200 to 8515 stems per hectare, but the basal area ranged from 39.6 to only 40.0 m2 per hectare. Ten randomly placed quadrats in each side of the stand provided measures of variance for stand data on sizes of live and dead stems and inferred survivorship. The distribution pattern of stems was also studied in mapped transects 5 by 100 meters in each stand.

The distribution of size classes of dead beach stems reveals a primary unstable size group (0.1 to 6.0 cm) and primary stable, secondary stable, and unstable size groups. The general features of the distribution curve are similar to those with constant mortality rates. The implications of the high mortality of small stems (0-6 cm diameters) and the high density and low density patches for release and competition are discussed. The mosaic pattern of beech distribution is thought to be the result of cyclic regeneration.

The proportion of dead beech is not different significantly between north- and south-facing slopes, but it is significantly different among stands and between aspects within the stand. No directional trend can be demonstrated for aspect differences. The proportion of non-beech species was significantly different between north- and south-facing slopes (21% vs. 11%), among sites, and between aspects within sites. These differences when added to characteristics of the survivorship curves and the distribution of spruce in the stands suggest that the south-facing stands are under environmental stress.

The distribution of stems was studied by analysis of variance and my Morisita's (1959) index of dispersion analysis. From the analysis of variance, several sizes of primary patches and mosaics were detected in all stands. The regular distribution pattern of individual stems and clumps of stems was revealed, as was some random distribution and several scales of aggregated pattern. North-facing sites had more randomly distributed stems among all trees and among all beech. No random distribution was found in the subareas nearest the ridge crests in the south-facing sites. Living beech was more randomly distributed than were all trees or all beech. Morisita's index reveals similar patterns, but it showed more regularity of pattern. Greig-Smith's (1964) method was more convenient for detecting and interpreting the mosaic pattern and for determining average size of clump and single clump area.

The complicated structure and dynamics of these stands are very strongly controlled by the root suckering of the sole dominant, beech. The dynamics of the stands are revealed by inferred survivorship and mosaic pattern in different size classes to be strongly controlled by changes in micro-environmental pattern and competition during growth of clumps. Maintenance of these deciduous island communities in the boreal conifer forests is possible largely because of rapid and cyclic regeneration in a complicated mosaic.

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

Included in

Botany Commons