Coordinating Management Decisions of Neighboring Stands with Dynamic Programming

Abstract

Stands are typically thought of as the basic analysis area (AA) or building block of a forest. Historically, forest management planning has focused on how stand management activities can be coordinated across the forest to provide a sustainable flow of products over time. With increasing concerns about sustaining environmental conditions, spatial facets of the management situation also become a concern. Often, how a stand is managed impacts not only the condition of the stand, but also the condition of a “neighborhood” surrounding the stand. These impacts can have a lasting impact over time, thus adding a temporal dimension to spatial concerns. Spatial concerns add substantial complexity for analyzing management options because forests include many stands, each with its own potential unique neighborhood. Generally, it is too simplified to separate the forest into independent neighborhoods, ignoring conditions along neighborhood boundaries. In recent years adjacency constraints have received considerable attention as one policy tool for addressing spatial management objectives. Adjacency constraints set maximum limits on the size of harvest blocks. Specific management limitations for adjacency constraints vary, without any one accepted best set of rules. Typically, the rules involve only regeneration harvests associated with even-aged management. Rules may limit harvesting of adjacent stands within a given time period or set a maximum limit on the size (total area) of any contiguous harvest block. Rules involve an “exclusion period” during which harvesting is prohibited for adjacent stands around a harvest block. Definitions of exclusion periods also vary by user. Some assume a fixed time length for the exclusion period (e.g. 10 years) or tie its definition to the condition of the harvest block (e.g. not until the trees regenerated in the harvest block reach 20% of their height at rotation). Methods for incorporating adjacency constraints into forest planning models have received considerable