Introduction

“Windthrow is a complex process resulting from interactions between natural and anthropogenic factors” (Ruel 1995). Understanding the interactions between these factors and the inherent risk to stands from this disturbance has numerous benefits to natural resource managers. Windthrow is affected by several factors; climate, topography, physical and biological stand attributes, soil characteristics, and silviculture all play a role in the dynamics of wind disturbance (Ruel 1995).

Understanding windthrow risk throughout the landscape can provide insights into natural vegetation patterns and habitat types. Risk evaluation can also be used to help predict how current forests may change without harvesting and subsequent impacts to forest health. Risk evaluation can help managers: (1) evaluate where to locate plantations, (2) decide which regeneration strategies are appropriate, (3) determine where thinning and other partial harvests are acceptable, and (4) determine what species composition or rotation length is desirable for individual sites. Predicting damage, or potential for damage, provides the opportunity for impacts to be considered during prescription development, allowing for revision of management objectives or incorporation of mitigative actions into management plans (Mitchell 1998).

Currently, little is known about the extent of both catastrophic and endemic wind damage in Maine. As a result, this project’s aim was to explore the nature of wind disturbance throughout the State’s large forest ownerships. To augment our understanding of the interaction between forest management and wind damage vulnerability, this project developed a generalized wind damage model that reflects topographic exposure (distance-limited Topex) (Ruel and others 1997), soil conditions (rooting depth), and stand characteristics (density, edge, height, species composition, and treatment history). Results from similar modeling projects in British Columbia suggest that these risk factors are consistent in varied locations, and this indicates that general models may be applicable to landscapes other than the ones for which they were developed (Lanquaye-Opoku and Mitchell 2005). This model was calibrated using information from published literature and experiences of regional managers. We then evaluated it using spatially explicit wind damage records from a 40,800-ha area of managed forest in northern Maine.