Date of Award

Summer 8-14-2015

Level of Access Assigned by Author

Open-Access Thesis

Degree Name

Master of Science (MS)


Forest Resources


Robert Seymour

Second Committee Member

Shawn Fraver

Third Committee Member

Aaron Weiskittel


Land managers today are increasingly called upon to retain and restore late-successional features on harvested landscapes in order to reverse the current global decline of large, old trees and their associated elevated levels of biodiversity and ecosystem function. Such retention practices are commonly thought to result in increased rates of mortality as a result of exposure to wind, thus compromising management objectives. This study investigated the survival and growth dynamics of the reserve trees retained in harvested gaps (n=787) established 20 years prior in the Acadian Forest Ecosystem Research Project (AFERP) in east-central Maine.

A high (relative to similar treatments implemented throughout the world) overall survival rate (91.6 ± 0.02%) across 18 species is attributed to variables associated with vigor, post-treatment growth rate and crown ratio. For species with a sample size greater than 50, Thuja occidentalis (81.0% survival rate) performed the poorest, while Tsuga canadensis (97.9% survival) performed the best of all reserve trees with a sample size greater than 50.

Distance to original gap edge was significantly positively associated with mortality and not windthrow, while distance to expanded gap edge was significantly positively associated with windthrow. These phenomena could be explained by (1) trees dying as a result of shock soon after gap creation (i.e. significant positive correlation between mortality and distance to original gap edge) and wind stress increasing at the micro-site level closer to the center of gaps as the gaps expanded (i.e. significant positive correlation between windthrow and distance to expanded gap edge).

Post-treatment basal area increment (BAI) was assessed using increment cores extracted from the five predominant reserve tree species in AFERP: Acer rubrum (n=108), Picea rubens (n=127), Pinus strobus (n=99), Thuja occidentalis (n=65), and Tsuga canadensis (n=129). Three different response variables were tested: proportional increase, annual post-treatment BAI, and absolute increase. Overall treatment effects on BAI were minimal, while species varied in their responses based on size, age, shade tolerance and spatial location within gap. In general, growth responses diminished with greater tree size, age, and shade-intolerance and increased with distance from the nearest gap edge. While all species demonstrated positive growth responses relative to their pre-treatment BAI and to paired analogues selected from an untreated control, hemlock was consistently the most responsive species to treatment.

The duration of sustained accelerated growth was negatively influenced by pre-treatment BAI, tree slenderness (height/diameter), and positively influenced by crown ratio and volume. In pairwise comparisons of species, hemlock had significantly higher incidence of sustained accelerated growth than white pine (p=0.00), red spruce (p=0.02), and paper birch (p=0.03). Red maple and white cedar had a marginally higher incidence of sustained accelerated growth than white pine (p=0.06 and p=0.07, respectively).

Benefits and trade-offs of the different response variables exist, with the proportional increase demonstrating sensitivity toward small and slow-growing trees and the annual post-treatment BAI and absolute increase demonstrating sensitivity toward large and fast-growing trees, namely white pine. Of the three response variables tested, the absolute increase demonstrated the best balance of sensitivity to tree size and spatial variables, while performing moderately well in terms of relative marginal and conditional R2 values when compared to the proportional increase and annual post-treatment BAI. The growth responses quantified in this study can be used to inform growth projection models of uneven-aged treatments and stand dynamics of ecological forestry silvicultural prescriptions and illustrate the need for a metric capable of quantifying growth responses of trees across a range of size-classes.

Based on these findings, silvicultural prescriptions should gradually increase exposure of retained trees and target the retention of trees with high values in variables associated with vigor, in order to improve survival rates of reserve trees. With respect to growth responses, selecting reserve trees with high values in attributes associated with vigor, i.e. crown ratio, pre-treatment growth, of intermediate age-, size-, and crown-class, and high shade tolerance will promote strong growth responses of reserve trees. Abiding by these guidelines strictly, however, will compromise the objective of retaining a diverse cohort of reserve trees of varying age, size, and species. Selecting the individual of each species and size-class with the highest value for attributes associated with vigor is therefore the best option for both desired survival and growth outcomes.