Economic potentials and limitations of tree-species diversification as an adaptation strategy to climate change and extreme weather events

Abstract

Increasing biophysical disturbances under climate change have severe impacts on forest ecosystems. Recent large-scale calamities in Central Europe have underlined how extreme weather events challenge our understanding of mechanisms behind forest resistance to disturbances. Large, spatially correlated disturbances also lead to high economic losses for forest enterprises, resulting in a lack of resources for financing reforestation and forest management in general. Tree-species diversification is commonly expected to be a key adaptation strategy for forestry under exacerbating climate-driven risks. This expectation mainly arises from ecological rationales and mechanisms. Bioeconomic models interpreting tree species as assets in a portfolio have been proven helpful to explore economic advantages of tree-species diversification. Results of stand-level simulations suggest that product diversification and a higher biophysical stability in mixed-species stands may also mitigate economic risks. However, the economic advantages of tree-species diversification are still under debate and have not yet been analyzed under unprecedented extreme weather events and at larger spatial scales. This thesis deals with the overarching hypothesis that tree-species diversity buffers the economic consequences of climate change and extreme weather events for forest enterprises. Three sub-hypotheses structure the analyses: H1: The economic buffering capacity of tree-species diversity towards climate change is affected by the mix of silvicultural management options applied at the stand level. H2: The economic buffering capacity of tree-species diversity towards climate change is affected by extreme weather events. H3: The economic buffering capacity of tree-species diversity towards climate change is affected by the considered spatial scale, from stand to regional scale. The four scientific articles in the thesis address these hypotheses with a combination of econometric analyses and normative simulation-optimization approaches. Paper 1 integrated alternative silvicultural management options in simulation-optimization models that apply Modern Portfolio Theory to economic tree-species selection. It assessed the economic buffering capacity of combinations of proactive and reactive management options in Norway spruce (Picea abies (L.) H. Karst) forests towards stand-level disturbances. Paper 2 applied advanced econometric time series analyses to operational data sets on harvest volume, salvage harvests, and wood revenues. Impulse Response Functions were applied to disentangle mechanisms behind decreasing wood revenues after disturbances. Integrated into an R package for wood valuation (Paper 3), the econometric results allowed for distinguishing between impacts of extreme weather events as compared to stand-level disturbances on wood revenues. Paper 4 scaled up the stand-level portfolio model to a large forest enterprise with several, spatially explicit planning units. Within this regional-scale model, scenarios of unprecedented extreme weather events resulted in spatial correlations of annuities between the planning units. The optimization accounting for these correlations compared bottom-up and top-down decision-making regarding the selected spatial scale for tree-species diversification and the resulting economic buffering capacity. The results showed that tree-species diversification outperformed other silvicultural management options at the stand level in terms of mitigating economic risks. The effect on expected economic return and the economic buffering capacity in terms of the risk-averse objective function, Value at Risk, depended on the productivity and revenues of the tree species introduced to spruce forests. The econometric analyses revealed that decreasing spruce revenues after disturbances were predominantly driven by an oversupply to wood markets rather than losses in wood quality. Simulations with extreme-event scenarios at the regional scale indicated that their adverse economic consequences exceeded the buffering capacity of local as well as across-region tree-species diversification. Rather than diversification, we found homogenization of the tree-species composition under extreme weather events. The model focused on stand types with low establishment costs instead of investing in biophysically stable stand types. The enterprise-level model showed that site heterogeneity was an additional driver of tree-species diversification beyond biophysical stability of mixed stands and economic portfolio effects. Spatial correlations related to extreme weather events and endogenous market effects affected optimal diversification strategies. Bottom-up decision-making resulted in a comparatively high local stand-type diversity. Top-down decision-making that accounted for the spatial correlations opted for diversification across the region with higher economic adaptation gains. This suggests that stand-level models systematically underestimate the buffering capacity of tree-species diversification at larger spatial scales. The methodological innovations in this thesis contributed to a more sophisticated understanding of economic consequences of extreme weather events for forest enterprises and economic potentials and limitations of tree-species diversification as an adaptation strategy. The stand-level study showed the potential of bioeconomic portfolio models to account not only for economic advantages of tree-species diversification, but also for synergies when combining different silvicultural management options. Advanced time series analysis were adopted to make operational data of forest enterprises available for scientific analysis. Particularly Impulse Response Functions were a promising method to study market mechanisms related to disturbance events and to inform future simulation modeling. The R package woodValuationDE compiles existing models and the econometric results and makes them easily available for future applications. Up-scaling the portfolio simulation-optimization model to the enterprise level allowed for integrating novel mechanisms related to site heterogeneity, endogenous market effects, and spatial correlations. These mechanisms combined with top-down decision-making resulted in systematically different solutions as compared to previous stand-level approaches. Scenarios of unprecedented extreme weather events going beyond typical sensitivity analyses allowed for overcoming limitations of studies relying only on statistical models to assess possible future economic consequences of climate change on forestry. In conclusion, tree-species diversification has the potential to buffer parts of the adverse economic consequences of increasing disturbances. Its economic advantages may, however, be limited or even superimposed by increased investment risks under unprecedented extreme weather events. For forest management, the thesis thus recommends not to rely on tree-species diversification as the one and only adaptation strategy. For forest policy, the findings suggest that the economic benefits of establishing biophysically stable and diverse forests are challenged by exacerbating disturbances. The sensitivity analyses indicate that subsidies for planting costs are promising to mitigate the investment risks of forest enterprises which establish diverse, stable forests that provide multiple ecosystem services to society. This thesis proposes reconsidering tree-species diversification as a silver bullet for buffering economic consequences of climate change and extreme weather events for forest enterprises.