A QUANTITATIVE ANALYSIS OF OAK SEEDLING SUCCESS IN REGENERATING OAK FOREST STANDS

Open Access
Author:
Graboski, Lake Edward
Graduate Program:
Ecology
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
March 19, 2019
Committee Members:
  • Dr. Kim C. Steiner, Thesis Advisor
  • Dr. Marc McDill, Committee Member
  • Dr. Laura Leites, Committee Member
  • Dr. James Finley (Retired), Committee Member
Keywords:
  • oak
  • seedling
  • regeneration
  • pennsylvania
Abstract:
Oaks (Quercus spp.) are becoming less dominant in general throughout the Central Hardwood region of the eastern United States, causing a shift in forest composition that impacts the ability of these forests to provide timber and ecosystem services. Insufficient oak regeneration is often cited as the main cause of these changes. In order to learn more about the recruitment of successful oak seedling regeneration, this study examined 23 mature oak stands that had reached exclusion stage of stand development (15 to 20 years after harvest). The objective was to develop predictive models of oak seedling success dependent upon initial conditions and on conditions within the first decade after logging. A model based on the aggregate height of all oak seedlings > 0.5ft (15cm) offered the best prediction of oak seedling success from pre-harvest data. Post-harvest models were more predictive, as expected, and were enhanced by contrasting the height of dominant oak seedlings to that of competing species. Of all competing tree species, red maple (Acer rubrum) and black birch (Betula lenta) were the two most frequently encountered, and important differences in their effects on oak seedling success were quantified. The increase in power associated with post-harvest height measurements was greatest between 1 and 4 years following the harvest. Stands where white oak (Quercus alba L.) composed a greater percentage of the oak regeneration had higher rates of successful oak seedlings after accounting for the sizes of oak and non-oak regeneration. Site quality and other candidate variables did not improve predictive power of the best models, but these variables were correlated with significant predictors included in models. These findings provide new information for evaluating oak regeneration both before and after harvest of mature oak forests.