Abstract
Historical radial increment data based on tree ring analyses from the close-to-nature experimental forest management unit Smolnícka Osada in Central Slovakia were used for retrospective modelling of changes in forest dynamics to estimate the sensitivity of management planning goals under climate change. Four example years representing historical periods with typically different species-specific patterns of radial increment in mixed beech-fir-spruce forest (1910, 1950, 1980, and 2014) served as virtual starting points for the modelling. An advanced density-dependent matrix transition model was utilised for modelling stand dynamics. An integrated tool for nonlinear financial optimisation searched for an optimal management equilibrium. In addition to transition probabilities adjusted from increment data, some assumptions for changes in ingrowth and mortality related to the increment, as well as a case study concerning the reduced ingrowth changed by game browsing intensity, were tested for modelling more realistic historical ecological conditions. The sensitivity study revealed changes in the optimal management equilibrium represented by optimal basal area, tree species composition, diameter distribution and target harvest diameter over time due to the adapted ecological modelling. The main lesson of the past for the future is to avoid placing too much trust in the simple extrapolation of current trends, such as the observed continual decline in spruce related to climate change, but to be aware of temporal and possibly reversible processes, such as the observed extensive fir recovery after the reduction of air pollution. Tree species diversity appears to be the best option for the uncertain future.