ABSTRACT
Results from landscape ecology (e.g., Franklin and Foreman, Hof and Joyce) and wildlife biology (e.g., Giles), however, suggest that interactions among spatially dispersed stands may critically influence the functioning of forest ecosystems. Consider a simple example. Forest managers may produce timber products while managing the ecosystem to provide food and shelter for wildlife. If a young forest stand provides a critical food supply but is separated from critical shelter (nesting or calving habitat) by an ecological barrier (perhaps an impassable river or an active zone of clear-cutting), wildlife may be unable to access
Present-day ecosystem management represents a structural shift in the philosophy of resource management. Public forest management and policy objectives have developed from initial interest in commercial products to broad mandates for multiple-use, multiresource management (Behan), as managers balance the consumptive needs of society with a desire to maintain biological diversity and functions of ecosystems (Swallow). Ecosystem management complicates policy decisions considerably. Existing economic models of forest management are inadequate because they fail to account for the spatial relationships that determine many ecosystem functions. The traditional forest management approach is primarily based on the stand-level harvest models of Faustmann as ex-
The next section presents a forest-level model that explicitly accounts for ecological interactions among forest stands and the owner’s decisions on when and which stands to harvest. We then present a simulation model which emphasizes the simplest forest ecosystem, one consisting of two forest stands. The simulation results, for example, show that explicit recognition of ecological interactions, even between identical forest stands, may prescribe an optimal harvest pattern which not only recommends specialization over time but also specialization across space. Such spatial and temporal specialization leverages opportunities to provide ecosystem goods that simply are not apparent from models focused on the single stand. Simulation results show that individual harvest ages within the steady-state cycle may be substantially different from the Hartman rotation age for an individual stand.2 The final section discusses the policy relevance of the model and its implications in forest ecosystem management.
