To examine the role of climatic extremes in structuring reef fish communities in the Arabian region, reef fish communities were visually surveyed at four sites within the southern Persian Gulf (also known as the Arabian Gulf and The Gulf), where sea-surface temperatures are extreme (range: 12-35° C annually), and these were compared with communities at four latitudinally similar sites in the biogeographically connected Gulf of Oman, where conditions are more moderate (range: 22-31° C annually). Although sites were relatively similar in the cover and composition of coral communities, substantial differences in the structure and composition of associated fish assemblages were apparent. Fish assemblages in the southern Persian Gulf held significantly lower estimates of abundance, richness and biomass, with significantly higher abundances of smaller sized individuals than Gulf of Oman assemblages. Functionally, southern Persian Gulf sites held significantly lower abundances of nearly all the common fish trophic guilds found on Gulf of Oman sites, although higher abundances of herbivorous grazers were apparent. These results suggest the potential for substantial changes in the structure of reef-associated fish communities, independent of changes in habitat within an environment of increasing fluctuations in oceanic climate.

As the science of connectivity evolves, so too must the management of coral reefs. It is now clear that the spatial scale of disturbances to coral reef ecosystems is larger and the scale of larval connectivity is smaller than previously thought. This poses a challenge to the current focus of coral reef management, which often centers on the establishment of no-take reserves (NTRs) that in practice are often too small, scattered, or have low stakeholder compliance. Fished species are generally larger and more abundant in protected reserves, where their reproductive potential is often greater, yet documented demographic benefits of these reproductive gains outside reserves are modest at best. Small reproductive populations and limited dispersal of larvae play a role, as does the diminished receptivity to settling larvae of degraded habitats that can limit recruitment by more than 50%. For "demographic connectivity" to contribute to the resilience of coral reefs, it must function beyond the box of no-take reserves. Specifically, it must improve nursery habitats on or near reefs and enhance the reproductive output of ecologically important species throughout coral reef ecosystems. Special protection of ecologically important species (e.g., some herbivores in the Caribbean) and size-regulated fisheries that capitalize on the benefits of NTRs and maintain critical ecological functions are examples of measures that coalesce marine reserve effects and improve the resilience of coral reef ecosystems. Important too is the necessity of local involvement in the management process so that social costs and benefits are properly assessed, compliance increased and success stories accrued.

The extent of larval dispersal on coral reefs has important implications for the persistence of coral reef metapopulations, their resilience and recovery from an increasing array of threats, and the success of protective measures. This article highlights a recent dramatic increase in research effort and a growing diversity of approaches to the study of larval retention within (self-recruitment) and dispersal among (connectivity) isolated coral reef populations. Historically, researchers were motivated by alternative hypotheses concerning the processes limiting populations and structuring coral reef assemblages, whereas the recent impetus has come largely from the need to incorporate dispersal information into the design of no-take marine protected area (MPA) networks. Although the majority of studies continue to rely on population genetic approaches to make inferences about dispersal, a wide range of techniques are now being employed, from small-scale larval tagging and paternity analyses, to large-scale biophysical circulation models. Multiple approaches are increasingly being applied to cross-validate and provide more realistic estimates of larval dispersal. The vast majority of empirical studies have focused on corals and fishes, where evidence for both extremely local scale patterns of self-recruitment and ecologically significant connectivity among reefs at scales of tens of kilometers (and in some cases hundreds of kilometers) is accumulating. Levels of larval retention and the spatial extent of connectivity in both corals and fishes appear to be largely independent of larval duration or reef size, but may be strongly influenced by geographic setting. It is argued that high levels of both self-recruitment and larval import can contribute to the resilience of reef populations and MPA networks, but these benefits will erode in degrading reef environments.

The global decline in coral reefs demands urgent management strategies to protect resilience. Protecting ecological connectivity, within and among reefs, and between reefs and other ecosystems is critical to resilience. However, connectivity science is not yet able to clearly identify the specific measures for effective protection of connectivity. This article aims to provide a set of principles or practical guidelines that can be applied currently to protect connectivity. These ‘rules of thumb’ are based on current knowledge and expert opinion, and on the philosophy that, given the urgency, it is better to act with incomplete knowledge than to wait for detailed understanding that may come too late. The principles, many of which are not unique to connectivity, include: (1) allow margins of error in extent and nature of protection, as insurance against unforeseen or incompletely understood threats or critical processes; (2) spread risks among areas; (3) aim for networks of protected areas which are: (a) comprehensive and spread—protect all biotypes, habitats and processes, etc., to capture as many possible connections, known and unknown; (b) adequate—maximise extent of protection for each habitat type, and for the entire region; (c) representative—maximise likelihood of protecting the full range of processes and spatial requirements; (d) replicated—multiple examples of biotypes or processes enhances risk spreading; (4) protect entire biological units where possible (e.g. whole reefs), including buffers around core areas. Otherwise, choose bigger rather than smaller areas; (5) provide for connectivity at a wide range of dispersal distances (within and between patches), emphasising distances <20–30 km; and (6) use a portfolio of approaches, including but not limited to MPAs. Three case studies illustrating the application of these principles to coral reef management in the Bohol Sea (Philippines), the Great Barrier Reef (Australia) and Kimbe Bay (Papua New Guinea) are described.

Artificial reefs are often promoted as mitigating human impacts in coastal ecosystems and enhancing fisheries; however, evidence supporting their benefits is equivocal. Such structures must be compared with natural reefs in order to assess their performance, but past comparisons typically examined artificial structures that were too small, or were immature, relative to the natural reefs. We compared coral and fish communities on two large (>400,000 m3) and mature (>25 year) artificial reefs with six natural coral patches. Coral cover was higher on artificial reefs (50%) than in natural habitats (31%), but natural coral patches contained higher species richness (29 vs. 20) and coral diversity (H′ = 2.3 vs. 1.8). Multivariate analyses indicated strong differences between coral communities in natural and artificial habitats. Fish communities were sampled seasonally for 1 year. Multivariate fish communities differed significantly among habitat types in the summer and fall, but converged in the winter and spring. Univariate analysis indicated that species richness and abundance were stable throughout the year on natural coral patches but increased significantly in the summer on artificial reefs compared with the winter and spring, explaining the multivariate changes in community structure. The increased summer abundance on artificial reefs was mainly due to adult immigration. Piscivores were much more abundant in the fall than in the winter or spring on artificial reefs, but had low and stable abundance throughout the year in natural habitats. It is likely that the decreased winter and spring abundance of fish on the artificial reefs resulted from both predation and emigration. These results indicate that large artificial reefs can support diverse and abundant coral and fish communities. However, these communities differ structurally and functionally from those in natural habitats, and they should not be considered as replacements for natural coral and fish communities.

Dispersal in marine systems is a critical component of the ecology, evolution, and conservation of such systems; however, estimating dispersal is logistically difficult, especially in coral reef fish. Juvenile bicolor damselfish ( Stegastes partitus ) were sampled at 13 sites along the Mesoamerican Barrier Reef System (MBRS), the barrier reefs on the east coast of Central America extending from the Yucatan, Mexico to Honduras, to evaluate genetic structure among recently settled cohorts. Using genotype data at eight microsatellite loci genetic structure was estimated at large and small spatial scales using exact tests for allele frequency differences and hierarchical analysis of molecular variance (AMOVA). Isolation-by-distance models of divergence were assessed at both spatial scales. Results showed genetic homogeneity of recently settled S. partitus at large geographic scales with subtle, but significant, genetic structure at smaller geographic scales. Genetic temporal stability was tested for using archived juvenile S. partitus collected earlier in the same year (nine sites), and in the previous year (six sites). The temporal analyses indicated that allele frequency differences among sites were not generally conserved over time, nor were pairwise genetic distances correlated through time, indicative of temporal instability. These results indicate that S. partitus larvae undergo high levels of dispersal along the MBRS, and that the structure detected at smaller spatial scales is likely driven by stochastic effects on dispersal coupled with microgeographic effects. Temporal variation in juvenile cohort genetic signature may be a fundamental characteristic of connectivity patterns in coral reef fishes, with various species and populations differing only in the magnitude of that instability. Such a scenario provides a basis for the reconciliation of conflicting views regarding levels of genetic structuring in S. partitus and possibly other coral reef fish species.

Atmospheric carbon dioxide concentration is expected to exceed 500 parts per million and global temperatures to rise by at least 2°C by 2050 to 2100, values that significantly exceed those of at least the past 420,000 years during which most extant marine organisms evolved. Under conditions expected in the 21st century, global warming and ocean acidification will compromise carbonate accretion, with corals becoming increasingly rare on reef systems. The result will be less diverse reef communities and carbonate reef structures that fail to be maintained. Climate change also exacerbates local stresses from declining water quality and overexploitation of key species, driving reefs increasingly toward the tipping point for functional collapse. This review presents future scenarios for coral reefs that predict increasingly serious consequences for reef-associated fisheries, tourism, coastal protection, and people. As the International Year of the Reef 2008 begins, scaled-up management intervention and decisive action on global emissions are required if the loss of coral-dominated ecosystems is to be avoided.

Lutjanus apodus (Schoolmaster) were collected from several mangroves and coral reefs at Turneffe Atoll, Belize, in order to investigate whether elemental concentrations from the otolith edge could be used as a means to identify the habitat (mangrove or coral reef) and site (9 mangrove sites and 6 reef sites) from which they were collected. Results of a two factor nested MANOVA (sites nested within habitat) indicated significant differences in elemental concentrations between habitats (i.e., mangrove versus reef) as well as among sites. When separate Linear Discriminant Function Analyses (LDFA) were used to assess whether the spatial variability in otolith chemistry was sufficient to differentiate individuals to their respective habitats or sites, the results indicated that fish were classified (jackknife procedure) with a moderate to poor degree of accuracy (i.e., on average, 67% and 40% of the individuals were correctly classified to the habitat and site from which they were collected, respectively). Using a partial Mantel test we did not find a significant correlation between the differences in otolith elemental concentrations between sites and the distance between sites, while controlling the effect of habitat type (mangrove or reef). This suggests that for mangrove and reef sites at Turneffe Atoll, Belize, the overlap in terms of L. apodus otolith elemental concentrations is too high for investigations of fish movement. Finally, by comparing previously published Haemulon flavolineatum otolith chemistry to that of L. apodus we assessed whether these species showed similar habitat and/or site specific patterns in their otolith chemistry. Although both species were collected from the same sites our results indicated little similarity in their elemental concentrations, thus suggesting that habitat and site elemental signatures are species specific.

Otolith chemistry can be used to assess pelagic larval fish connectivity by comparing spatially variable otolith edge chemistry (corresponding to the site of collection) to otolith core chemistry (corresponding to the site of hatching). However, because the otolith's edge and core represent different life stages, the deposition of elements may differ, thus complicating direct comparisons of edge and core chemistry to investigate connectivity. Here we present data from a field experiment in which otoliths from embryos (3 days post-fertilization) and juveniles of Stegastes partitus were collected at the same site and time, and chemically analyzed to assess whether elemental concentrations of otoliths vary ontogenetically. Separate multivariate analyses, each investigating the spatial/temporal variability in the chemistry of either embryo otoliths or the edges of juvenile otoliths, revealed significant differences, suggesting an environmental influence to the chemical signals of otoliths. A nested multivariate analysis assessing whether otolith chemistry varied with life history stage (i.e., ontogenetic variability) indicated that elemental concentrations of embryo otoliths were significantly greater than that of juvenile otolith edges. Specifically, embryo elemental concentrations of Mn, Zn, Sn, Ba, Ce, and Pb were between 2 and 163 times greater than those of the corresponding juvenile otoliths, and thus the environment was not the primary determinant of embryo otolith chemistry. Consequently, caution is warranted when interpreting environmental patterns of otolith cores, particularly when using them as a proxy for natal signatures.

A nested grid ocean circulation modeling system is used to examine the circulation, dispersion, and hydrodynamic connectivity of surface waters on the Belizean shelf. The nested grid system consists of a coarse‐resolution (∼19 km) outer model of the western Caribbean Sea, an intermediate‐resolution (∼6 km) middle model of the southern Meso‐American Barrier Reef System (MBRS), and a fine‐resolution (∼2 km) inner model of the Belizean shelf. The nested system is forced by climatological monthly mean surface forcing and integrated over 5 years. The near‐surface circulation on the Belize shelf produced by the inner model is characterized by a strong and persistent northwestward flow as a direct influence of the Caribbean Current on the northwestern shelf and a weak and spatially variable flow on the inner and southern shelf. The monthly mean model currents are used to calculate retention and dispersion of conservative, near‐surface particles carried by the ocean currents. The near‐surface dispersion is relatively higher in areas seaward (east) of Lighthouse and Glovers Reef atolls and lower on the inner shelf, particularly within the Inner Channel and in the vicinity of South Water Cay. To examine hydrodynamic connectivity of reefs in the surface waters of the Belize shelf, we calculate upstream and downstream retention areas for coral reefs at Turneffe Islands and Glovers Reef atolls. The potential sources of passive, near‐surface particle supply reaching these two reef atolls within 30 days include both the shallow waters surrounding the two sites, the deep waters between them, and the coastal waters of the Bay Islands (Honduras). The 30‐day downstream retention areas of the Turneffe and Glovers Reef atolls cover the central and southern Belize shelf, respectively.

We investigated whether the otolith chemistry of Haemulon flavolineatum (French grunt), a nocturnally active fish, could be used as a means to differentiate individuals occupying mangrove and coral reef habitats. In 2003, adults were collected from 9 mangrove and 10 coral reef sites throughout Turneffe Atoll, Belize. Concentrations of trace elements were measured at the edge of sagittal otoliths by laser ablated inductively coupled plasma mass spectrometry. Results of a two-factor nested MANCOVA (sites nested within habitat and covariate of fish size), used to investigate whether significant differences in otolith elemental concentrations existed between habitats (i.e., mangrove versus reef) and among sites, indicated significant differences between habitats, in terms of lithium, magnesium, zinc, and rubidium (fish from mangroves had greater concentrations than those from coral reefs), as well as among sites (for several elements). Because elemental variability existed between habitats and among sites, we asked whether this variability was sufficient to differentiate habitats and sites using separate linear discriminant function analyses (LDFA). LDFA indicated that fish were classified to the habitat (mangrove or reef) from which they were collected with a moderate degree of accuracy (correct classification of 74% and 79% for mangrove and coral reef fish, respectively), but were poorly classified to the site from which they were collected (average correct classification of 46% with a range of 0–89%). Otolith microchemical investigations of H. flavolineatum at Turneffe Atoll can be used to identify movement between habitats, yet due to the lack of unique site-specific chemical signatures likely caused by the nocturnal movement of individuals, it will not be possible to identify specific sites from which reef fish originated.

Investigating the resilience of fish assemblages requires large-scale experiments, but large-scale manipulations are rarely possible, and "natural experiments" frequently must take their place. Here we report a unique opportunity to explore the resilience of reef fish assemblages. The underground nuclear testing program conducted by the French Armed Forces at Mururoa Atoll, while not intended as an ecological experiment, caused multiple instantaneous removals of fish over areas of approximately 12.5 km(2), without otherwise affecting their environment or the opportunity for recolonization. We show that fish assemblages responded rapidly to these intense; large-scale perturbations, restoring assemblage structure within 1-5 yr. Reef fishery conservation efforts must emphasize protection of habitat, because reef fish assemblages are resilient even to intensive, localized harvesting so long as the structural and biotic integrity of their habitat is maintained and neighboring sites are able to supply recruits.

As well as serving valuable biodiversity conservation roles, functioning no-take fishery reserves protect a portion of the fishery stock as insurance against future over-fishing. So long as there is adequate compliance by the fishing community, it is likely that they will also sustain and even enhance fishery yields in the surrounding area. However, there are significant gaps in scientific knowledge that must be filled if no-take reserves are to be used effectively as fishery management tools. Unfortunately, these gaps are being glossed over by some uncritical advocacy. Here, we review the science, identify the most crucial gaps, and suggest ways to fill them, so that a promising management tool can help meet the growing challenges faced by coastal marine fisheries.

Despite their importance at smaller scales, microhabitat characteristics appear to have only a minor influence on recruitment of reef fishes at scales of >50 m. Abundance of young-of-year (YOY) recruits of 104 species was measured in late summer of 3 yrs at three sites on each of seven reefs of the Capricorn-Bunker Group, southern Great Barrier Reef. Abundance of all YOY recruits and of each of the 15 most common species varied substantially, with numerous interactions between year and either reef or site (within reefs). Sites varied in substratum composition, but these differences did not explain much of the pattern in reef fish recruitment. When microhabitat principal components were included as covariates, 14 of the 16 analyses still showed significant spatio-temporal interactions, and all 16 exhibited significant recruitment variation. Redundancy analysis of the fish assemblage revealed 51.6% of the variation in recruitment could be explained: 18.6% by spatio-temporal scales, 6.7% by microhabitat, and 26.3% shared by spatio-temporal and microhabitat variables. Responses to microhabitat, at or soon after settlement, help determine distribution of recruits at small scales, but factors other than microhabitat appear to play major roles in determining the spatial and temporal variation in recruitment of coral reef fishes at larger spatial scales.

Coral reefs contain the most speciose communities of fishes on this planet, so it is appropriate to use these to explore how fish species are organized into communities. While descriptive data suggest that the diverse communities of fish on coral reefs are equilibrial assemblages of species, all finely adapted to specific and unique ecological roles, these are highly dynamic, non-equilibrial assemblages with structure driven more by patterns of recruitment and loss of individual fishes, than by patterns of resource allocation among differently adapted phenotypes. As a consequence, local assemblages differ in structure, and structure wanders through time. Individual fish are confronted by different mixes of species in different times and places. The recruitment process that drives these dynamics is complex, being governed by several mechanisms, and local populations receive some portion of their recruitment from distant sources. Information on this connectivity among local populations is critically important for management which is based increasingly on use of marine protected areas (no-take zones) both to conserve, and to provide sustainable fisheries. At present, however, we do not know the spatial scale or the extent of this connectivity, and this critical knowledge gap impedes both management, and fundamental understanding.

Marine and fisheries scientists are increasingly using metapopulation concepts to better understand and model their focal systems. Consequently, they are considering what defines a metapopulation. One perspective on this question emphasizes the importance of extinction probability in local populations. This view probably stems from the focus on extinction in Levins' original metapopulation model, but places unnecessary emphasis on extinction-recolonization dynamics. Metapopulation models with more complex structure than Levins' patch-occupancy model and its variants allow a broader range of population phenomena to be examined, such as changes in population size. age structure and genetic structure. Analyses along these lines are critical in fisheries science, where presence-absence resolution is far too coarse to understand stock dynamics in a meaningful way. These more detailed investigations can, but need not, aim to assess extinction risk or deal with extinction-prone local populations. Therefore, we emphasize the coupling of spatial scales as the defining feature of metapopulations. It is the degree of demographic connectivity that characterizes metapopulations. with the dynamics of local populations strongly dependent upon local demographic processes. but also influenced by a nontrivial element of external replenishment. Therefore, estimating rates of interpopulation exchange must be a research priority. We contrast metapopulations with other spatially structured populations that differ in the degree of local closure of their component populations. We conclude with consideration of the implications of metapopulation structure for spatially explicit management, particularly the design of marine protected area networks.

Patchily distributed demersal marine organisms that possess a pelagic larval stage have the potential to form complex linkages among sedentary adult populations through larval dispersal. Unfortunately, determining the extent of this inter-population connectivity in specific cases is quite difficult: larval dispersal is a complex result of several biotic and non-biotic processes, and marine larvae are minute creatures that are difficult to track or manipulate. Nevertheless, effective management of such open populations requires that we know how they are organized and interact. Coral reef fishes and many decapod crustaceans share patchy demersal distributions and pelagic larvae. We review progress made in determining connectivity among crab, lobster and reef fish populations, considering the questions that need to be answered and the techniques that have been applied. We highlight common research strengths and gaps among researchers working on reef fishes and benthic decapods, as well as those areas given greater attention by one group, the lead of which the other group will need to follow. Finally, we suggest that greater progress towards understanding dispersal and connectivity in the future will best be achieved by integrating the approaches we review into large-scale, multidisciplinary, collaborative research programs developed and carried out in association with natural resource managers.

The structure and dynamics of reef fish assemblages on small patch reefs in the Caribbean (St. Croix, US Virgin Islands) were compared with those studied previously (Sale et al. 1994) on the Great Barrier Reef (One Tree Reef), Australia, in order to test whether assemblage struc- ture and function are consistent despite overall taxonomic differences. St. Croix reefs were monitored over 5 yr (1991 to 1996), while those at One Tree Reef had been monitored for 10 yr (1977 to 1986). Analyses of species richness and abundances for individual patch reef assemblages revealed con- siderable change in assemblage structure between censuses. Specifically, number of individuals and species varied among summer and winter censuses, with summer censuses generally showing greater numbers. Among years but within a season, there was significant variation in the number of indi- viduals (summer and winter) and species richness (summer only) at St. Croix, while only species richness varied yearly (summer and winter) at One Tree Reef. In addition, when the proportional abundances of species observed on each patch reef were compared among censuses on the same patch reefs but at different times (measured as Czekanowski similarity values), results indicated that faunal similarity declined with increasing length of interval between censuses. The mean, between- census Czekanowski similarity values for individual assemblages were 0.56 and 0.59 for St. Croix and One Tree Reef, respectively. These results suggest that there is substantial temporal change in assem- blage structure at both locations. Species in each location were classified into 12 trophic guilds in order to determine whether the assemblages at St. Croix and One Tree Reef showed similar trophic structure despite different species present. Results indicated that guild structure varied among censuses in both locations and that average guild structure differed substantially between these locations. At St. Croix, diurnal and nocturnal planktivores were proportionately more important, while site-attached herbi- vores, wide-ranging grazing herbivores, wide-ranging browsing herbivores, sedentary carnivores, and demersal planktivores were proportionately more important on the Australian patch reefs. Results confirm that comparable assemblage dynamics occur for patch reef assemblages at St. Croix and One Tree Reef. However, there are also clear differences in structure and dynamics in these 2 locations, despite the very similar physical and hydrographic habitats occupied.