Post-Disturbance Recovery Shifts in Boreal Evergreen Landscapes: Impacts on Carbon Dynamics and Land Surface Properties.

In the evergreen boreal forest, field studies show that vegetation does not always regenerate to its previous state after disturbance but instead transitions to systems dominated by deciduous trees or non-forest vegetation. Gaining a better understanding of drivers and impacts of post-disturbance recovery is thus crucial to accurately project future vegetation dynamics and associated impacts on the carbon, water, and energy balance of the region. We here perform simulations with the dynamic vegetation model LPJ-GUESS to investigate (1) if observations of post-disturbance recovery dynamics can be reproduced in the model, (2) which environmental factors control such shifts, and (3) how these in turn influence land surface properties such as albedo and evapotranspiration. We find that post-disturbance recovery trajectories can be clustered into distinct response patterns of recovery and shifts to alternative plant types. These shifts occur even in places where multiple plant types can in theory establish in the model and thus emerge due to shifts in competitive advantage mediated by warming and soil properties. We further find that shifts from forested to non-forested ecosystems have strong impacts on land-surface properties while shifts between different forest types are less impactful. We conclude that LPJ-GUESS is capable of reproducing observed disturbance-induced changes in vegetation dynamics following disturbances. Post-disturbance recovery is a key process driving accelerated vegetation change under climate change, further stressing the importance of accurately representing disturbance impact and recovery processes in land surface and coupled modeling.