Pattern, process, and prediction in marine invasion ecology

Invasions frequently continue long after dispersal corridors have been well established. Six interrelated processes (changes in donor regions, new donor regions, changes in recipient region, invasion windows, stochastic inoculation events, and dispersal vector changes) are examined to explain this phenomenon. The combination of these processes makes it difficult to forge a list of species from potential donor regions that will never become successful invaders and, by extension, to thus define the characteristics of species that have failed to invade. Predictions relative which species will invade and when they will invade can be improved by more detailed attention to these six categories of interrelated processes that mediate invasion success. Copyright (C) 1996 Elsevier Science Limited