COS 104-5
Tree harvest, fire, and drought can drive state transitions in savannas
Deciphering the mechanisms by which ecosystems shift between alternate stable states is a central goal of modern ecology, and a necessary task to inform management decisions. Several lines of evidence suggest savanna and forest are alternate stable states at the mesic end of the savanna distribution, and at the arid end savanna and grassland may also be alternate stable states. While mean annual precipitation explains broad patterns in biome distribution, fire has been invoked to explain why savanna often persists where the climate is amenable to the forest state. Given that some 240-450 million tons of wood is harvested annually from African forests and savannas, we ask whether tree harvest also plays a role. To address this question we incorporate tree harvest into a model of grasses, savanna saplings, savanna trees, and forest trees proposed by Staver and Levin (2012). We use assumptions about the differential demographic responses of savanna trees (they resprout) and forest trees (they do not resprout) to show how tree harvest influences not only tree cover but also tree demography.
Results/Conclusions
In general, the model suggests savannas are resilient to tree harvest due to adaptations savanna trees have evolved to cope with frequent fire (resprouting). However, at the arid and mesic ends of the grassland-savanna-forest continuum, tree harvest can have a profound and potentially irreversible impact. At the savanna-forest ecotone the model produces a runaway feedback between tree harvest and fire frequency that can drive a biome transition from forest to savanna that will likely be persistent without active fire suppression. At the grassland-savanna ecotone tree harvest can make a savanna more vulnerable to periodic droughts by disrupting the storage potential held within long-lived adult savanna trees, which results in a transition to a grassland state. These results are not just of theoretical interest. Demand for wood fuels is projected to increase in sub-Saharan Africa, as is the use of charcoal, which requires more wood per unit energy. Thus, as demand for wood fuels, especially charcoal, increases in the coming decades we can anticipate the boundaries of the savanna biome relative to grassland and forest to shift.