Modeling Timber Harvests In The Northeastern United States

Open Access
Kovach, Kyle Ryan
Graduate Program:
Forest Resources
Master of Science
Document Type:
Master Thesis
Date of Defense:
July 06, 2016
Committee Members:
  • Michael Jacobson, Thesis Advisor/Co-Advisor
  • Marc Mcdill, Committee Member
  • Laura Leites, Committee Member
  • Timber Harvest
  • Probability
  • Forest Resources
  • Statistics
  • Environmental Economics
  • Natural Resources
Timber harvesting activity takes place for a variety of reasons. Environmental, social, and economic factors all influence where and when a harvest occurs. Using the FIA database and statistical modeling, our research has shown that there are factors statistically significant to a harvest at both the plot and tree level in the northeastern United States. These factors also vary with the proportion of basal area of timber removed in a harvest. Using logistic regression to examine these significant factors, we have determined that the value of standing timber has an effect on harvest probability at the stand level. At a tree harvest level, there is no significant relationship between harvesting decision and individual tree value. This may be due to the factors considered when a site is being considered for a harvest differing from those when individual trees are being selected. A significant difference also exists for harvests on private versus public land, with private owners tending to take larger, more valuable trees, but less overall stand volume. Diameter tends to be a strong predictor of harvest probability at both the plot and tree level, and suggests that there is an increasing harvest probability up to a maximum plot average or individual tree size, after which, the probability of a harvest decreases. Model variable statistical significance for volume/size metrics such as diameter and cubic feet is quite high. The inability to utilize strong explanatory variables to describe a large portion of the variation inherent in timber harvests is a problem that plagues similar studies, even at smaller scales, suggesting that more research needs to be done on what environmental, social, or economic variables influence timber harvests, and whether these factors are national or regional. In attempting to explain what significantly influences timber harvests with our models, we can statistically say that volume, size, and to a degree, value, are all factors influential to timber harvesting in the northeastern United States.