Predation regimes on many oceanic islands are known to be weak. Evidence gathered in the 1970s and 1980s shows that Antillean anoles live at higher densities on fewer resources, grow more slowly, reproduce later, and live longer than mainland counterparts. Living at high densities should select for the ability to subsist on minimum resource levels, the definition of a superior resource competitor. These results imply that island anoles, and by inference, birds, experience low predation but intense competition, whereas mainland species experience the opposite. Further support is found in the phenomenon of community saturation in Antillean birds, an unexpected finding underlain by hyperdispersed body size ratios within avian foraging guilds, low alpha diversity and low species packing within guilds, all being evidence that these guilds are structured by competition. Islands should thus not be regarded as havens for weak competitors, but rather as refuges from predation. It follows that intense predation regimes prevent island species from colonizing mainlands, and that competition and/or low resource levels prevent mainland species from colonizing islands. These predictions are experimentally testable with lizards and if confirmed, could set island biogeography on a new course.