Optimizing Forest Landscape Composition for Multiple Ecosystem Services and Stakeholders Under Uncertainty

Determining the desirable composition of a forested landscape and its associated ecosystem services (ES) is challenging because the solutions must reconcile the preferences of various forest stakeholders and account for uncertain data. By combining an online survey of forest professionals in Slovenia (n = 130) and forest professionals, forest scientists, nature conservationists and forest owners in Germany (n = 649) with multi-objective robust optimization, we derived compromise portfolios of forest types. These portfolios minimize the trade-offs between five ES (avalanche protection, carbon storage, recreation, timber production and water protection) and account for the varying capacity of eight forest types to supply ES. The resulting optimized forest landscape compositions always comprised at least two forest types. In both countries, uneven-aged native deciduous and conifer forest in mixed stands were prominent in the optimized portfolios. In Germany, however, the optimized portfolio also contained exotic species in mixtures, whereas forest stands without active management were notable for several ES in Slovenia. Unmanaged forest stands were also selected in the forest composition optimized for nature conservationists in Germany: the nature conservationists’ portfolio diverged strongly from those of the other stakeholders. Our results illustrate that diversified forested landscapes provide multiple ES, but also secure the provision of a single ES when accounting for uncertainty. The optimal forest compositions obtained by multi-objective robust optimization may provide a useful starting point for participatory planning approaches to identify the most socially-acceptable strategies for adapting forest management to an uncertain future.