Thomas Perry

Date of Award


Level of Access

Open-Access Thesis

Degree Name

Master of Science (MS)


Forest Resources


Jeremy Wilson

Second Committee Member

Steve Sader

Third Committee Member

Alan White


The likelihood of windthrow or windsnap occurring in a forest stand includes numerous factors; however, past research suggests that these factors can be grouped into four broad categories: regional climate, topographic exposure, soil properties and stand characteristics (Mitchell, 1995). Of the three categories, stand characteristics are most commonly and easily modified through forest management. Vulnerability to wind damage in Maine may increase in the future because of three trends influencing stand conditions. One, Maine forests contain a considerable amount of balsam fir and red spruce, tree species that are considered particularly susceptible to wind damage. Two, extensive areas regenerated after the 1970's and 1980's era spruce budworm outbreak are maturing. Three, partial removals currently account for over 74 percent of the area harvested annually in the state (McWilliams et al. 2005). Two approaches to augment our understanding of the interaction between forest management and wind damage vulnerability in Maine forests were developed The first approach combined information from the base of scientific wind disturbance literature with more localized information from Maine's forest resource managers. Forest resource professionals were surveyed through phone calls and professional meetings to gather information about wind damage over their careers. The second approach developed a general vulnerability to wind damage model that reflects topographic exposure (distance limited TOPEX (Ruel et al. 1997), restricted rooting depth, elevation, and stand characteristics (height, density, edge, treatment history, and species composition). Results of the first approach reveal serious limitations in information about wind damage statewide. However, numerous patterns and trends were identified. Damage differs by storm type and storms impact the state on a continuum of storm intensity, frequency, and scale. Numerous factors influence the damage potential of these wind events on forests. These factors include topographic exposure, soil conditions and stand characteristics. Damaging storms appear to originate from the southwest most frequently and impact softwoods more severely than hardwoods. Frequent low-intensity winds tend to eliminate softwoods from hardwood dominated stands. The general vulnerability to wind damage model is based on Mitchell's (1998) conceptual windthrow triangle and is built from eight component variables describing stand, soil, and topographic characteristics. The model is built and calibrated from composite variables which combine the component variables into distinct site and stand components. The model was tested on a 40,800 hectare forest area in northern Maine with spatially explicit wind damage records. To avoid problems with spatial autocorrelation ten random samples were drawn from the study area and evaluated individually with a Mann-Whitney non-parametric comparison of means test (alpha 0.05). Results from the ten samples were pooled, and a one sample comparison of means t-test was used to analyze the consistency of the results from the ten individual samples (alpha 0.05). The final model identifies significant and consistent differences between damaged and undamaged areas (p-value 0.000). When evaluated individually, not all model components were significantly different (e.g., density, edge, exposure and species composition). Variables describing thinning, stand height, and elevation had the greatest differences between means of the populations of damaged and undamaged stands in the study area. The general model developed proved useful on the study area and by design should be transferable to diverse regions throughout the state.