Three papers on the natural disturbance model of forest management

This thesis is a collection of three papers examining the natural disturbance model of forest management, specifically how timber harvest rates might be determined under that model. The studies presented here focus on areas of the boreal mixedwood forest in Alberta, Canada.; The first paper projects the consequences on timber supply of a simple interpretation of the natural disturbance model: the rate of disturbance (harvest and wildfire) under management is set equal to the rate of natural disturbance, and the proportion of different forest types subject to managed disturbance in equal to that under natural disturbance. The paper compares the timber supply implications of three very different estimates of the natural disturbance rate to the timber supply resulting from the current sustained yield policy. This of the natural disturbance model results in major reductions in annual allowable cuts.; The second paper characterizes the natural disturbance regime for an 8.6 million ha region of the boreal mixedwood forest as a serially independent random draw from a lognormal distribution. This characterization has some important results which were identified through Monte Carlo simulations. Estimates of the mean annual burn rate for the study area are highly variable. Single parameter characterizations of the disturbance regime ( e.g. mean disturbance rate, fire cycle) are unreliable. There in no equilibrium age class structure for a forest subject to the lognormal disturbance regime. No target age class distribution can be justified an the basis that it is the “correct” natural distribution.; The third paper presents a modeling system that recognizes the variability in the disturbance regime, and uses the variability to help guide harvest rate determination. Monte Carlo simulation is used to project the probability of habitat areas for fin vertebrate species for each year of the planning horizon. These probably we used to set constraints for an optimization based forest planning model, which is used to develop trade-off curves between objective function values and constraint levels.