Abstract

Optimal forest harvesting is a problem that dates back many centuries. Modern forest-management needs models taking into account the relatively long rotation, the multiaged structure, the age-dependent timber content of trees, and the multiple services forests provide. Approaches to characterize the optimal management policy range from models that represent the forest by a unique state variable that can usually be solved analytically (i.e., models that consider forests composed by a unique even-aged stand or allow a uneven aged forest but only consider its total biomass), to much more sophisticated linear and integer programming harvest scheduling models. In this survey, we focus on dynamic optimization problems where the forest is represented with an age-class structure. These models present richer dynamics than one-variable models while preserving their analytic tractability to some extent.