Tree regeneration in models of forest dynamics – Suitability to assess climate change impacts on European forests

Climate change impacts on Europe's forests are becoming visible much sooner than previously anticipated. The increase in natural disturbances leads to tree mortality and raises concerns about the forest's adaptive potential to sustain vital ecosystem services. In this context, the regeneration phase is crucial and comprises the largest potential to adapt to new environmental conditions with long lasting implications. Yet, forest regeneration is particularly susceptible to climatic changes due to the many directly climate-dependent processes, such as seed production and germination but also seedling and sapling development. Models of forest dynamics (MFDs) are essential to describe, understand and predict the effects of changing environmental and management factors on forest dynamics and subsequently on associated ecosystem services. We review a large variety of MFDs with regard to their representation and climate sensitivity of regeneration processes. Starting with a description of the underlying biological processes, we evaluate the various approaches taking into account specific model purposes, and provide recommendations for future developments. We distinguish between models based on ecological principles and models based on empirical relationships. We found an ample mix of regeneration modelling approaches tailored to different model purposes. We conclude that current approaches should be refined to adequately capture altered regeneration trends. Specifically, refinement is needed for MFDs that rely on ecological principals, as they suffer from knowledge gaps and underrepresented processes, thereby limiting their ability to accurately simulate forest regeneration under climate change. Global vegetation models are strongly constrained by their weak representation of vegetation structure and composition, and need to include more detail regarding structural complexity and functional diversity. Models focused on timber yield often rely on strong assumptions regardin