Optimizing thinning and rotation in a stand of Pinus sylvestris on a drained peatland site

A non-linear programming algorithm was combined with a stand simulator consisting of individual-tree, distance-dependent diameter growth, mortality and thinning models, and a distance-independent height growth model. The Hooke and Jeeves' direct search method was used for determining the optimum thinning schedule and rotation length. The optimum solutions for both deterministic and stochastic optimization, with stochastic growth predictions, were calculated. In the optimization, stand density was regarded as a function of the distance to ditch. On the rather infertile site, it was most profitable to keep the stand about twice as dense near ditches and the extraction road, which was placed midway between two adjacent ditches. Stochasticity did not affect the optimum regimes markedly nor did it increase the expected returns of the optimum regimes.