Spatial harvest scheduling in the Southeastern United States
MetadataShow full item record
The use of spatial harvest scheduling processes has increased over the past 15 years due to regulatory and voluntary programs that affect the spatial and temporal arrangement of management activities across a landscape. A number of papers have been presented in the literature that describe and compare the performance of spatial harvest scheduling algorithms on small sets of management problems. Applications of a single planning process to a broad range of ownership sizes and spatial configuration of ownership is lacking. In this research, we assess whether there is a set of ownership patterns, ownership sizes, or initial age class distributions that will be more highly affected by potential harvest scheduling constraints than others. This research represents one of the most extensive assessments of spatial harvest scheduling constraints ever performed for southeastern U.S. forest conditions and indicates that small landowners, and landowners with young age class distributions, will be most affected by a commonly used (but voluntary at this point) set of clearcut adjacency constraints (240 acre maximum clearcut, 2-year green-up). A meta heuristic, which includes threshold accepting, 1-opt tabu search, and 2-opt tabu search performed as well, or better, than threshold accepting and tabu search by themselves. The combination of search characteristics (speed, diversification, and intensification) show that forest plans developed with heuristics will benefit from multiple search strategies. Finally, we assessed whether a recent development (raindrop heuristic) would be of value in forest planning problems that include area restriction adjacency constraints. While the modified raindrop heuristic is computationally intensive, it requires only two parameters and can produce as good, or better solutions than threshold accepting or tabu search.