“…The Arabian Peninsula is routinely considered as the corridor where migrating East African populations would have passed during a single or multiple dispersal events (Clark, 1989;Foley, 1994, 1998;Stringer, 2000;Derricourt, 2005;Beyin, 2006;Mellars, 2006aMellars, , 2006bRose, 2007a;Marks, 2008;Shea, 2008;. These episodes of demographic expansion are thought to correspond with wet phases, whereas Arabia would have been deserted during arid and hyper-arid events (Rose, 2004b;Petraglia, 2005;Rose, 2007b;Carto et al, 2009;Parker, 2009). This seems at odds with the fact that during Pleistocene glacial maxima, correlated with periods of Red Sea level lowstand, narrow channels less than 4e5 km wide would have favored sea crossings between Africa and southern Arabia (Winney et al, 2004;Lambeck et al, 2011;but see;Fernandes et al, 2006).…”