Tree cover transitions in tropical landscapes: hypotheses and cross-continental synthesis

The term ‘forest’ is first of all an institutional marker; there are forests without trees and treesoutside forests (van Noordwijk et al 2009, PB 15). Moreover, debates about forest transitionhave often tried to replace time as the primary X-axis with macroeconomic variables that indicatethe changing roles of forest areas once economies develop. However, in the context of tropicalcountries, the existing hypotheses lack agency- and context-specific explanations. While thelogarithm of human population density accounts for 70-80% of variation in the national forestcover fraction, forest transition points can occur at almost any population density and forestcover fraction (Figure 3.1). They seem to be more likely, however, in countries that already hadabove-average forest cover in relation to their population density. The identified limitations offorest transition theory include the loose definition of forests (combining primary, secondary andplanted forest types), a lack of detail about the forest cover dynamics involved, including its spatialand (multi-)temporal scales, and few explanations of context-specific transitions (Perz 2008).