Large areas of savannas are found in Africa in climatic zones favourable to humid tropical forests: they are relicts of past dry periods and forest domains are naturally expanding. Men have influenced the transgression, especially by starting savanna fires. FORSAT is a stochastic cellular-automaton model dedicated to the forest-savanna mosaic on a landscape scale, taking into account savanna fires, vegetation succession cycle (vegetation types are discriminated by their ecological function: herbs, pioneer seedlings, pioneer adults, forest) and pionneer seed dispersal and recruitment. The model is validated by comparison between its emergent behavior and biogeographical field studies. Two parameters (an environmental factor and a man-related fire frequency) control an van-der-Waals-like phase transition between forest and savanna. This brings to three types of evolution: progression or regression of forest edge, formation and coalescence of clumps in savanna and global afforestati...
Charly Favier, Marc A. Dubois