Body size is one of the most important biological characteristics of species. Among reef fishes, size is globally unevenly distributed, species being on average larger in the Atlantic and tropical eastern Pacific (TEP) than in the Indo-Pacific. This is related to species diversity and also probably linked to the history of these regions. Species size is related to numerous biological and ecological characteristics. Here we show the links between size and diet, home range, depth range, and geographical range. The large-scale distribution of species size has important implications at the local level. In particular small species tend to be more abundant than larger ones, which combined with the relationship between species diversity and species size implies higher abundances in species rich zones. Large species being in general more vulnerable to fishing than small ones, this also implies that reef fish assemblages from the Atlantic and TEP will be more vulnerable to a given rate of fishing mortality than those from the Indo-Pacific, all other factors being equal. A conceptual framework on the relationships between species richness, abundance, and biomass is proposed. It shows that the species size distribution within a regional pool, combined with regional and local species richness, may determine in great part the potential response of local assemblages to perturbations.

Body size is a central species parameter in ecology. Ecologists have been looking for general rules linking species' body size to many parameters [27, 36], in particular to geographical distribution [274, 917], temperature (see review by Angilletta [80]), metabolism (see initial theory by Brown [358]), abundance within populations and assemblages at various spatial scales (see review by White [2660]), and vulnerability [e.g. 322, 645, 1873, 1965]. In these domains, there is, however, far more literature on birds and mammals than on fish, and there is at present only a limited amount of work specifically addressing the role of fish body size on coral reefs except the initial review by Munday and Jones [1786] to which most subsequent literature refers.

Introduction

Human-induced disturbance, e.g., through unsustainable fishery activities, can affect the ecology of coastal areas (Jackson et al., 2001; Lotze et al., 2006) and reduce the associated goods and services required for human welfare (Worm et al., 2006). Major concerns are rising over observed declines in the abundance of particular species as well as reductions in functional diversity and changes in food web structure (Pauly et al., 1998; Micheli and Halpern, 2005; Villéger et al., 2010). As a result, the conservation and restoration of marine biodiversity and functions is a major concern (Balmford et al., 2005). Marine protected areas (MPAs) are recommended to promote the recovery of exploited populations and conserve or restore habitats, ecosystems, and biodiversity (Lubchenco et al., 2003). Marine protected areas have also been recommended as tools for ecosystem-based management (EBM) of marine resources (Pauly et al., 2002; Halpern et al., 2010) and for restoration of ecosystem function (Mumby et al., 2007; Mouillot et al., 2008). However, some studies have failed to detect significant differences in ecological response variables between no-take and reference areas, or have criticized the design of existing studies, fuelling the view that the importance of MPAs as a management tool may have been overstated (Hilborn et al., 2004; Murawski et al., 2005; Osenberg et al., 2006).

Some attempts have been made recently to generalize the ecological effects of MPAs (Mosquera et al., 2000; Côté et al., 2001; Halpern, 2003; Micheli et al., 2004; Guidetti and Sala, 2007; Claudet et al., 2008; García-Charton et al., 2008; Harmelin-Vivien et al., 2008; Lester and Halpern, 2008; Lester et al., 2009; Molloy et al., 2009; Claudet et al., 2010). However, ecological effects can vary both in direction and magnitude (Micheli et al., 2004; Claudet et al., 2008; Guidetti et al., 2008; Lester et al., 2009; Molloy et al., 2009; Claudet et al., 2010). This heterogeneity in response to protection may stem from differences in the design or age of MPAs, life histories of focal taxa, the socio-cultural context within which MPAs are established, or a combination of these effects.