“…There is also some indication that African savannas are characterized by grassland/woodland phases occurring with a periodicity of 250-600 years (Gillson, 2004;Gil-Romera et al, 2010). Such woodland phases were not always led by the spread of encroaching species (Gil-Romera et al, 2010;Scott, Cooremans, de Wet, & Vogel, 1991), which suggests that a mere increase in woody cover, as reported by several studies since the 20th century (Hoffman, Rohde, & Gillson, 2019;van Rooyen, le Roux, van der Merwe, van Rooyen, & Geldenhuys, 2018;Wiegand, Ward, & Saltz, 2005), does not necessarily mean shrub encroachment. Accordingly, identification of encroached states should be related to the increasing cover of encroacher species and the suppression of perennial grasses (Gil-Romera et al, 2010).…”