Date of Award
Master of Science
Garland R. Wells
J. W. Barrett, Edward R. Buckner
The study was conducted using hardwood diameter and mortality data from two five-year remeasurements of permanent continuous forest inventory (C.F.I.) systems. These C.F.I, plots were located on three Forestry Field Stations managed by the Forestry Department, The University of Tennessee. These three Field Stations were: Ames Plantation (located in Fayette and Hardman Counties), Highland Rim (located in Franklin County), and the Cumberland Station (located in Morgan and Scott Counties). These three Field Stations furnished four experimental forest tracts for study which were: Ames Plantation, Highland Rim, and the Morgan and Scott Counties, which were parcels of the Cumberland Field Station. C.F.I, data were limited to diameter ranges of 5.0 to 10.5 inches. Five-year diameter distributions were established with one-half-inch diameter classes. Markov chain theory was then used to project and evaluate these distributions. The objectives of this study were to compare present and projected hardwood diameter distributions and their resulting differences over the four experimental forest tracts. Comparisons were made among tracts, individual selected species on different tracts, and selected species within each tract. Markov chains were selected in order to quantify total stand variation in some of the major hardwood species of Tennessee. Markov chain analysis proved to be a more useful mathematical tool of management science than was anticipated. When data were ade-quate, four basic categories of results were obtained: initial and projected diameter distributions, projected time periods to reach larger diameter classes, expected time periods that trees can be expected to remain in each diameter class, and projected mortality percentages. The analysis revealed extremely long time period projections for both entire tracts and individual species. The ability of stands to differentiate within diameter classes over entire forest tracts was, to some extent, related to tract productivity. However, where all hard-woods were concerned, Ames Plantation did not perform according to expectations. Projected mortality differences between the Morgan and Scott County forest tracts were unexpected, given their close proximity and similar species composition. The majority of individual species performed according to traditional growth expectations. Ames Plantation possessed both the slowest and fastest growing species. Respectively, these were sweetgum (Liquidambar styraciflua) and southern red oak (Quercus falcata). Incon-sistencies did exist in the sweetgum at Ames. Both growth rates and mortality levels were much less than anticipated for this species. In addition, projected growth rates revealed similarity between yellow-poplar (Liriodendron tulipifera), red maple (Acer rubrum), and chestnut oak (Quercus prinus) on the Morgan and/or Scott County forest tracts.
Boyd, Wayne Madison, "Forest stand projections under conditions of uncertainty--a Markov chains approach. " Master's Thesis, University of Tennessee, 1970.