Marine Protected Areas (MPAs) include many subclasses (e.g. marine sanctuaries, marine parks, wildlife refuges, fisheries closures, no-take MPAs, multiple-use MPAs, marine reserves, ecological reserves) all of which can be defined based mainly upon the level of protection and the primary conservation goal (see www.mpa.gov; Lubchenco et al. 2003). MPAs, and especially the marine reserves subclass (i.e. ‘areas of the ocean completely protected from all extractive and destructive activities’; Lubchenco et al. 2003) represent the extreme case of the precautionary approach to managing marine resources (e.g. Lauck et al. 1998). The strong and rapidly growing interest in MPAs (and particularly in marine reserves) is reflected in the dramatic increase in the number of publications devoted to them (reviewed in Jones 2002, Gell & Roberts 2003, and the articles in ‘The Science of Marine Reserves ’, a supplemental issue of Ecological Applications, Vol 13, Iss 1, freely available for download at www.esa-journals.org/esaonline/?request=getstatic&name=s1051-0761-013-01-0001). In addition, there are now a number of sites on the World Wide Web that are either totally devoted to MPAs, or include relevant information on them: UNEP’s World Conservation Monitoring Centre (www.unep-wcmc.org/ protected_areas), the Partnership for Interdisciplinary Studies of Coastal Oceans (PISCO, www. piscoweb.org), and several others. This intense interest is at least partly related to MPAs having been identified and advocated as a conservation (of habitat and biodiversity) and managerial (of fisheries) tool of central importance in the Ecosystem Approach to Fisheries (EAF) (e.g. Agardy 2000, Stergiou 2002, Halpern & Warner 2003, Lubchenko et al. 2003, Pauly & MacLean 2003, Hilborn et al. 2004). It is hoped that MPAs will be beneficial in (1) rebuilding overexploited fish stocks, (2) preserving habitat and biodiversity, (3) maintaining ecosystem structure, (4) buffering against the effects of environmental variability, (5) serving as a control group against which populations in exploited regions can be compared, among others. Clearly, the choice of location, spatial extent (horizontal and vertical), and number of MPAs is critical if they are to meet these goals. It is to this issue that we devote our attention here. Halpern & Warner (2003) state, ‘Most reserve locations and boundaries were drawn by a political process that focused on economics, logistics, or public acceptance, while largely overlooking or ignoring how the complex ecology and biology of an area might be affected by reserve protection.’ In this sense, establishing the locations and boundaries of MPAs can be seen as analogous to the imperfect process associated with establishing stock management grids—a process that has never really managed to incorporate the key realities of population dynamics of the exploited species. While there is a growing consensus on the need for MPAs, at this point in time there is no clear and well-founded basis upon which their location, spatial extent and number can be decided. In fact, rationales/frameworks that are based upon principles of theoretical and applied ecology have only recently been tapped to address these key questions (e.g. Roff & Evans 2002, Botsford et al. 2003, Roberts et al. 2003a,b, Shanks et al. 2003, Fisher & Frank 2004). Much of this work focuses on the manner in which different aspects of the life histories of marine organisms—spawning locations, dispersal, larval retention and export, juvenile nursery areas, etc.—affect MPA design. In this context, we contend that an ecoevolutionary framework already exists, grounded in marine ecology and fisheries oceanography, that is completely consistent with EAF and MPA objectives. The Member-Vagrant Hypothesis as a framework for defining the location, size and number of MPAs. The Member-Vagrant Hypothesis (MVH), the development of which can be traced through a series of publications by Mike Sinclair and Derek Iles (Iles & Sinclair 1982, Sinclair 1988, 1992, Sinclair & Iles 1988, 1989), defines 4 attributes of populations that are involved in the regulation of their size. The ’population richness’ refers to the number of discrete self-sustaining populations (henceforth simply ’populations’) exhibited by any given species. Species such as herring, cod, mackerel, the salmonids, and many others are population rich. The ‘spatial pattern’ relates to the geographic distribution of these populations. Population rich species are usually also broadly distributed (the north Atlantic region is so far the best studied in this regard). Population richness and spatial pattern are species-level characters. The ’absolute abundance’ refers to the instantaneous size of the various populations of any given species, and this size—which can range over several orders of magnitude—varies over time (thus, its ’temporal variability’). These last 2 components of the MVH are population-level characteristics. Sinclair & Iles have applied the MVH to describe the richness, pattern, abundance and variability of several economically im271