Zusammenfassung
In this dissertation I examine the existence of multiple stable treecover states of the Earth’s forest ecosystems, with a primary focus on the boreal region. Combining remotely-sensed observations, data analysis, and conceptual models, I identify areas with alternative vegetation states under the same environmental conditions, and I explore their possible dynamics under current and future conditions.
In recent years, it has been found that the distributions of remotely sensed tree cover in boreal and tropical ecosystems have three distinct modes, corresponding to treeless, open woodland, and forest states. In light of this pattern, it has been suggested that these modes reflect the presence of alternative tree-cover states. As a response to climate change, these ecosystems could undergo critical regime shifts.
For the tropics, it has been shown that a positive feedback between fire and vegetation can act as switch between the three states. For the boreal forest, it has been shown that the observed multimodality is not caused by temperature and precipitation patterns. In the first part of this thesis, by means of generalised additive models, I show that the relationship between tree cover and eight remotely-sensed environmental variables varies within the boreal region. Using a classification, I identify areas which exhibit alternative tree-cover states under similar environmental conditions. These regions show a reduced resilience and can shift between states.
In the second part of the thesis, I develop and employ a conceptual model to show that tree-cover multistability in the boreal region can emerge through competition between species with different evolutionary traits. By forcing the model with varying permafrost conditions, I show that the asymmetry in tree-species distribution between North America and Eurasia could be due to permafrost presence.
In the third part of the thesis, employing projected environmental conditions from the Representative Concentration Pathway (RCP) 2.6 and 8.5 scenarios, I identify potentially multistable areas during the last decade of the 21st century. By including a simple effect of CO2 on plant growth in the conceptual model, I simulate the dynamics of multistable zones under projected environmental conditions. I show that the two scenarios exhibit opposite trends regarding the extent of multistable areas, and that the resilience of Eurasian species might increase, while North American forests might lose stability. In each part of this thesis, I consider limits and advantages of the tools at hand, implying that only through their combined use we can advance our knowledge of tree-cover multistability and improve the representation of the boreal forest in climate models.
