Risk-based Selection of Forest Regeneration Methods
Abstract
A stochastic optimization model is developed to make a selection between the planting method and the seed-tree method, taking into account the uncertainty of, and the legal requirement on, the stocking level of the established seedlings in a given year after regeneration action. Uncertainty is quantified as the variation of the mortality rate of planted seedlings for the planting method, and as the prediction error for the seed-tree method. The objective of the forest landowner is assumed to maximize the expected net present value (NPV). Numerical simulations show that the owner should prefer the seed-tree method to the planting method for Scots pine stand. However, if the risk-free selection model is used, it overestimates the expected NPV by about 2%. Sensitivity analysis shows that a less restrictive forest act could improve the expected net present value both for the planting method and the seed-tree method. Sensitivity analysis also shows that decreasing the level of variation of the mortality rate (or prediction error) increases the expected NPV.