Optimization models on continuous cover forestry are complicated and typically incompatible with rotation models. This dichotomy is theoretically unsatisfactory and makes the choice between clearcuts and continuous cover forestry vague. We present a theoretically sound and empirically detailed generalized setup with an optimal clear-cut regime (or even-aged management) and optimal continuous cover regime (or uneven-aged management) as special cases. It includes a size-structured growth model, variable and fixed harvesting costs, and allows for the completely flexible optimization of harvest timing in both regimes. Flexible harvest timing becomes essential when optimizing the transition from clear-cut regimes toward continuous cover forestry. The model is applied to Norway spruce (Picea abies (L.) Karst.) and solved as a dynamic mixed-integer problem. Low or moderate site productivity, an interest rate above 2%, and a high artificial regeneration cost support the optimality of continuous cover forestry. In its most general form, the optimal clear-cut regime does not exist when the continuous cover regime is globally optimal, and when it exists, the rotation period lengthens with interest rate. The optimal choice between forest management regimes may depend on the initial stand state and whether the naturally regenerated seedlings are utilized in solutions with clearcuts. Maximizing sustainable yield favors clearcuts.

Continuous-cover forestry (CCF) is expected to reduce the negative environmental impacts of peatland forestry in comparison with rotation forestry (RF), but the unknown profitability of CCF on peatlands limits its application in practice. The profitability of CCF was analyzed by simulating management scenarios with a process-based ecosystem model, EFIMOD, which was complemented to describe the interplay between tree growth and water table depth, which is typical of peatland forests. A variety of harvest intervals and post-harvest basal areas for a mature Norway spruce (Picea abies (L.) Karst.) dominated stand was simulated on a nutrient-rich peatland site. Conventional RF was simulated for comparison. CCF provided a higher profit than RF. The best financial performance was obtained with a 15-year harvest interval regardless of interest rate, although the overall profitability of CCF depended on the interest rate used. Ditch network maintenance was needed to maintain the stand growth only when the post-harvest basal area was smaller than 10 m²·ha⁻¹. There were many CCF scenarios in which the difference in the net present value of harvest revenues was within 10% compared with the best CCF scenario. Hence, there are many relatively profitable CCF harvesting alternatives for forest management in boreal spruce-dominated peatland forests.

We study the economics of carbon storage using a model that includes forest size structure and determines the choice between rotation forestry and continuous cover forestry. Optimal harvests may rely solely on thinning, implying infinite rotation and continuous cover forestry, or both thinning and clearcuts, implying finite rotation periods. Given several carbon prices and interest rates, we optimize the timing and intensity of thinnings along with the choice of management regime. In addition to the carbon storage in living trees, we include the carbon dynamics of dead trees and timber products. Forest growth is specified by an empirically validated transition matrix model for Norway spruce (Picea abies (L.) Karst.). The optimization problem is solved in its general dynamic form by applying bilevel optimization with gradient-based interior point methods and a genetic algorithm. Carbon pricing postpones thinnings, increases stand density by directing harvests to larger trees, and typically yields a regime shift from rotation forestry to continuous cover forestry. In continuous cover solutions, the steady-state harvesting interval and the diameter distribution of standing and harvested trees are sensitive to carbon price, implying that carbon pricing increases the sawlog ratio of timber yields. Additionally, we obtain relatively inexpensive stand-level marginal costs of carbon storage.