Resident Reef Fish Research Articles

Habitat associations and impacts on a juvenile fish host by a temperate gnathiid isopod.

The distribution and abundance of organisms is typically shaped by multiple biotic and abiotic processes. Micropredators are parasite-like organisms that are smaller than their hosts and/or prey and feed on multiple hosts during a given life stage. Unlike typical parasites, however, they spend much or most of their time free-living, associating only temporarily with hosts. In the ocean, micropredators can impact multiple fish species, and in particular can have significant lethal and sub-lethal effects on newly settled fish. Although gnathiid isopods are abundant and primary micropredators in coral reef ecosystems, their impacts are relatively unexplored within sub-tidal temperate rocky reefs. We investigated the distribution of juvenile gnathiid isopods along sub-tidal temperate rocky reefs and tested trap methodology. We also quantified both the sub-lethal and lethal impacts of feeding-stage juvenile gnathiid isopods on juvenile, post-settlement reef fish, Heterostichus rostratus (giant kelpfish). We were most interested in determining the relationship between gnathiid infestation level and fish swimming performance, in particular swimming metrics relevant to predator avoidance maneuvers. We found that Gnathia tridens was present in rocky reefs rather than embayments along the Southern California coastline and that within rocky reefs, gnathiids occurred in the highest densities in lighted traps. Surprisingly, we observed almost no influence of fish size or gnathiid sub-lethal infestation level on ambient or burst swimming performance metrics. However, burst duration was reduced by gnathiid infestation, which is important in predator avoidance. There were significant differences in survivorship among small fish compared to large fish as a result of gnathiid infestation. Larger fish survived higher numbers of gnathiids than smaller fish, indicating that parasite-induced mortality is greater for smaller fish. Investigations of the effects of micropredators on subsequent predator-mediated mortality, including the susceptibility of fishes and their individual responses to micropredators, can further contribute to our understanding of processes affecting recruitment in resident reef fish populations. Further research, especially within temperate sub-tidal ecosystems, is needed to understand and highlight the overlooked importance of micropredation in shaping fish populations within a reefscape.

A Resident Fish Guild as a Higher Trophic Level Indicator of Oyster Reef Restoration Success

Eastern oysters (Crassostrea virginica) are critical foundation species in estuarine waters, but due to a combination of natural and anthropogenic pressures, oyster abundance has declined. Restoring oyster reefs and monitoring restoration success often focuses on oyster metrics, but relatively infrequently, responses of higher trophic level species and the production of related ecosystem services are accounted for. To address this, we compare the response of a resident reef fish guild (gobies, blennies, toadfish) to standard metrics of oyster restoration success. Using lift nets and seines, natural and restored reefs were monitored over a two-year period within Mosquito Lagoon, Florida, USA. Standard metrics are indicative of restoration success; live oyster density and reef thickness increased in restored reefs after 12 and 24 months. Combined, live oyster density and reef thickness were the best predictors of annual resident reef fish abundance compared to water quality metrics. These results suggest that the benefits of restoring oyster reef habitat are conferred to broader components of the food web, with benefits accruing to reef resident fishes that are a key trophic linkage between lower trophic level foundation species and higher trophic level predators inhabiting coastal ecosystems.

Reduced access to cleaner fish negatively impacts the physiological state of two resident reef fishes

Many “client” coral reef fishes have their ectoparasites removed by the “cleaner” wrasse Labroides dimidiatus in mutualistic interactions. Clients regularly receiving cleaning services reportedly benefit from increased growth and cognitive performance, but the underlying physiological changes that covary with such benefits are unknown. Here, we tested whether reduced access to cleaning services affects physiological state in two species of damselfish, Amblyglyphidodon curacao and Acanthochromis polyacanthus. We performed an in situ removal experiment on the Great Barrier Reef, Australia, whereby 47% of cleaners on a natural reef were removed. Since cleaners occupy defined territories (called “cleaning stations”), this removal created areas where small, resident clients, including A. polyacanthus and A. amblyglyphidodon, had no access to cleaning services. One month following cleaner removal, we measured body condition and collected blood samples from both damselfish species from territories with and without access to cleaners. Blood was used for estimating haematocrit levels, hormonal analyses, and immune cell counts. We tested for correlations among all these parameters to explore potential trade-offs in terms of growth, aerobic capacity, immune activation, and/or reproduction as a result of the loss of cleaning benefits. In both species, we found that fish without access to cleaners had lower haematocrit, testosterone levels, and lymphocyte counts than fish with access. There were no significant changes in fish body condition, leukocytes, granulocytes, or plasma cortisol levels between fish with access to cleaners or not. However, testosterone levels correlated negatively with the proportion of granulocytes in the blood of fish with access to cleaners. Our results suggest that even a relatively short-term reduction in access to cleaning services can have negative physiological outcomes for clients. Thus, the presence of cleaners on coral reefs appears to have important benefits for coral reef fish community health.

Documenting baseline value chains of Palau's nearshore and offshore fisheries prior to implementing a large-scale marine protected area

Large-scale marine protected areas (LSMPAs) are increasingly enacted to manage marine resources and meet international targets, yet their socio-economic impacts are not well documented. In 2015, Palau passed the Palau National Marine Sanctuary (PNMS) Act, which will close 80% of its exclusive economic zone to fishing by 2020. An assessment of Palau's fisheries value chains can provide a baseline to assess the policy's impacts on the domestic market. A literature review and key informant interviews were conducted between August 2016 and June 2018 to: (i) estimate the current market size and structure; (ii) document the volumes, forms, and prices of fish throughout the market chain; and (iii) estimate fish consumption by different end consumers. The analysis finds the vast majority of reef fish are locally consumed by residents and the vast majority of tuna are exported internationally. Tourist consumption of reef fish (estimated for the time frame of the study during which tourist numbers varied considerably), anecdotally highlighted as a threat to reef fisheries sustainability, is relatively small. About half of resident reef fish consumption is non-commercial, highlighting the potential of non-market-based interventions. Commercial fish supply was reportedly inconsistent, a current challenge, but potential opportunity for a new, offshore domestic fishery. Finally, prices have remained stable, a characteristic of the market that may be attributed to the importance of social relationships in Palauan culture. Our results suggest key actions for PNMS implementation, can be used to evaluate future impacts of the PNMS, and provide planning insights for LSMPAs regionwide.

South‐western Atlantic reef fishes: Zoogeographical patterns and ecological drivers reveal a secondary biodiversity centre in the Atlantic Ocean

AbstractAimTo present an updated database of fish species recorded on south‐western Atlantic reef environments and to explore the ecological drivers of the structure, the latitudinal gradient of biodiversity and the centre of endemism in this peripheral province.LocationSouth‐western Atlantic (SWA): Brazilian and Argentinian Provinces.MethodsA database composed of 733 fish species along 23 locations in the SWA (00°55′ N to 43°00′ S) was compiled based on primary data, literature and museum records. Cluster and beta diversity analyses were carried out to evaluate faunal overlaps among locations and subprovinces. “Target‐area‐distance effect” and “stepping stones dispersal” hypotheses for assemblage composition were tested through Mantel tests. Relationships between the distribution patterns and ecological traits of reef fish species were investigated through generalized linear mixed‐effect models.ResultsOut of the 733 fish species, 405 are SWA resident reef fishes, of which 111 (27%) are endemics and 78 are threatened with extinction. Cluster analysis detected six subprovinces in the SWA structured following the target‐area‐distance model, and with no evidence for a latitudinal gradient in diversity. The greatest overall richness and endemic species richness were found in the east–south‐eastern region. Depth range, habitat use and body size were the main drivers of SWA reef fish assemblage structure.Main conclusionsThe Brazilian and Argentinian coasts constitute different provinces structured by oceanographic barriers and environmental filters. Similarities among oceanic islands indicate connectivity driven by stochastic and ecological factors. Species richness and endemism indicate that peripheral provinces may also bear centres of origin and biodiversity, patterns driven by parapatric/ecological speciation and the overlap between tropical and subtropical reef fish species. Ecological drivers of reef fish distribution, such as habitat specialization and body size, support hypotheses of speciation in the periphery. New approaches for spatial planning, marine protected areas and off‐reserve marine management are essential for the conservation and sustainability of SWA reef fishes.

Recovery potential of the world's coral reef fishes.

Continuing degradation of coral reef ecosystems has generated substantial interest in how management can support reef resilience. Fishing is the primary source of diminished reef function globally, leading to widespread calls for additional marine reserves to recoverfish biomass and restore key ecosystem functions. Yet there arenoestablished baselines for determining when these conservation objectives have been met or whether alternative management strategies provide similar ecosystem benefits. Here we establish empirical conservation benchmarks and fish biomass recovery timelines against which coral reefs can be assessed and managed by studying the recovery potential of more than 800 coral reefs along an exploitation gradient. We show that resident reef fish biomass in the absence of fishing (B0) averages ∼1,000kgha(-1), and that the vast majority (83%) of fished reefs are missing more than half their expected biomass, with severe consequences for key ecosystem functions such as predation. Given protection from fishing, reef fish biomass has the potential to recover within 35years on average and less than 60years when heavily depleted. Notably, alternative fisheries restrictions are largely (64%) successful at maintaining biomass above 50% of B0, sustaining key functions such as herbivory. Our results demonstrate that crucial ecosystem functions can be maintained through a range of fisheries restrictions, allowing coral reef managers to develop recovery plans that meet conservation and livelihood objectives in areas where marine reserves are not socially or politically feasible solutions.

Gender-specific benefits of eating eggs at resident reef fish spawning aggregation sites

Food resource availability has a fundamental role in shaping consumer populations through changes in energy intake. Tropical reef fish spawning aggregations provide a resource pulse for reef-based planktonic egg predators, and an opportunity to quantify the energetic repercussions of natural changes in food availability in tropical reef fishes. We examined the effects of the consumption of eggs from the broadcast spawning surgeonfish Ctenochaetus striatus on the allocation of energy to body condition, growth and reproduction in the planktivorous egg-predator damselfish Abudefduf vaigiensis. Fish that fed on eggs at resident fish spawning aggregation sites (FSASs) had significantly greater lipid storage in liver vacuoles compared to conspecifics from non-FSASs. Growth of male A. vaigiensis was faster at FSASs than non-FSASs. However, we found no differences in the growth of females among sites. Female A. vaigiensis from FSASs in vested more into reproduction, having larger gonadosomatic indices (GSI) than females from non-FSASs, while there was no difference in the GSI of males among FSASs and non-FSASs. At the same locations, no differences were found in the life-history traits of the reference species, Pomacentrus moluccensis, which does not consume eggs. This study demonstrates the role of natural variations in food availability on energetic processes in reef fish. Furthermore, the sex-specific energy allocation strategy highlights the complexity of the interaction between natural variations in food availability and life-history strategies. This study demonstrates that the conservation of FSASs may also benefit trophically linked reef fishes.

Small home range in southern Australia's largest resident reef fish, the western blue groper (Achoerodus gouldii): implications for adequacy of no-take marine protected areas

No-take marine protected areas (MPAs) represent an effective biodiversity conservation tool for a range of species including resident reef fishes that are intrinsically vulnerable to overfishing and other localised impacts. The western blue groper (Achoerodus gouldii) is the largest permanent-resident reef teleost in southern Australian waters and has the second-oldest recorded age of any labrid at 70 years. Acoustic telemetry was used to investigate whether adult A. gouldii can be effectively protected within adequately sized no-take MPAs. Ten passive acoustic receivers tracked the movements of 15 A. gouldii individuals (69–112-cm total length; 7–31-kg weight) at a site off north-western Kangaroo Island, South Australia. Most of the fish displayed high site fidelity (91–100% residence time for 10 of the 11 fish with useful data) for a narrow strip of fringing coastal reef (~1-km length by ~40-m width) throughout a 12-month period. Mean home-range along-shore length and area were estimated at 1076 ± 83 m (s.e.), and 45 188 ± 3497 m2, respectively (n = 11 fish with useful data). Comparison with other resident temperate-reef teleosts indicated no relationship between fish length and home-range area, and that A. gouldii has a relatively small home range. The high site fidelity and small home range of adult A. gouldii individuals makes localised populations amenable to a high level of protection within no-take MPAs.

An alternative method for estimating the status of resident reef fish stocks, based on differential fishing effort across a marine reserve boundary

The stock status of roman Chrysoblephus laticeps was estimated in the Goukamma, a temperate South African marine protected area (MPA). Standardised catch per unit effort (CPUE) from a controlled angling survey on both sides of the MPA border was employed to extrapolate the CPUE at zero fishing mortality. Converted into biomass, the estimate (61% of unexploited biomass) lay midway between those of two biomass-per-recruit (B/R) models for the same population based on angling and diving surveys (i.e. 52% and 69% of unexploited biomass respectively). The extrapolated CPUE at zero fishing mortality (4.4 fish angler-hour−1) in this study compared well with the mean CPUE of 4.6 fish angler-hour−1 determined during a concomitant survey in the core area of the nearby Tsitsikamma MPA – the oldest, and one of the largest, MPAs in Africa. Extrapolations of CPUE have the potential to deliver reliable and consistent estimates of stock status and could offer a practical alternative to conventional B/R models.

Does access to the bluestreak cleaner wrasse Labroides dimidiatus affect indicators of stress and health in resident reef fishes in the Red Sea?

Interactions between the bluestreak cleaner wrasse Labroides dimidiatus and its client reef fish are a textbook example of interspecific mutualism. The fact that clients actively visit cleaners and invite inspection, together with evidence that cleaners eat many client ectoparasites per day, indeed strongly suggests a mutualistic relationship. What remains unknown is how parasite removal affects the physiology of clients and thereby their body condition, health, and immune function. Here we addressed these issues in a field study in Ras Mohammed National Park, Egypt. In our study area, small reef patches are inter-spaced with areas of sandy substrate, thereby preventing many species (i.e., residents, including cleaner wrasses) from travelling between the reef patches. This habitat structure leads to a mosaic of resident clients with and without access to bluestreak cleaner wrasses, further referred to as “cleaner access”, on which we focused our study. We found that residents with cleaner access had higher body condition than residents without cleaner access. However, indicators of stress like variation in cortisol levels corrected for handling time and various immune parameters were apparently unaffected by cleaner access. In fact antibody responses were significantly higher in fishes without cleaner access. This suggests that cleaner access decreases the need for active immunity and that this releases resources that might be allocated to other functions such as somatic growth and reproduction.

A change of the seaward boundary of Goukamma Marine Protected Area could increase conservation and fishery benefits

Goukamma Marine Protected Area (MPA) on the South African temperate South Coast has been shown to be effective in maintaining a spawning stock of roman, Chrysoblephus laticeps (Sparidae). The larval ecology and the oceanographic conditions in the area suggest a good potential for the enhancement of roman stocks outside the reserve through larval dispersal. A high rate of illegal fishing just inside the seaward boundary of the MPA could severely compromise its function. We suggest that a change of the seaward boundary of the reserve to coincide with a latitudinal line could increase its function as a harvest refuge for resident reef fishes such as roman, facilitate voluntary compliance and monitoring and prosecution of illegal fishing without a significant negative impact on the commercial linefishing fleet in the area. Simple adjustments such as the one proposed here could be attempted at a number of South African MPAs as they would be beneficial to achieve fishery and conservation goals alike.

Trophic links and condition of a temperate reef fish: comparisons among offshore oil platform and natural reef habitats

We explored habitat-related variability in the composition and density of small inverte- brate prey of microcarnivorous reef fishes among 2 artificial (offshore oil platform) and 2 natural reef habitats in the Santa Barbara Channel, USA, and the consequences of this variability to the condition of a resident reef fish, the painted greenling Oxylebius pictus. Amphipod crustaceans comprised the highest proportion of potential prey at all sites, but the amphipod assemblage from the platforms con- tained a high percentage (>50% by number) of 2 exotic species, Caprella mutica and Erichthonius brasiliensis, that were either absent or rare on the natural reefs. At the platforms, 80 to 90% of the caprellid amphipods consisted of C. mutica. Densities of amphipods varied among sites and were 10 to 100 × higher at one of the platforms compared with the other sites. Amphipods comprised 89 to 98% (by number) of painted greenling diet. Fish condition factor (K) was correlated with absolute and per capita amphipod prey biomass. Our results fit a scenario described for some lotic systems where exotic amphipod species proliferate in the invaded habitat. Fish harvest the exotic species with positive effects on fish condition. The potential negative effects of the exotic species on the native amphipod assemblage thus contrast with a positive effect on the condition of a higher level consumer. Our findings suggest that trophic pathways on other types of artificial structures colonized by exotic species may also differ from those of the natural reef habitat.

Configuration of small patch reefs and population abundance of a resident reef fish in a complex coral reef landscape

AbstractHabitat use by the resident coral reef anemonefish, Amphiprion frenatus, was examined in the complex coral reef landscape of Shiraho Reef, Ishigaki Island, Okinawa, Japan, using an enlarged color aerial photograph processed using image analysis software as an accurate field map. The anemonefish inhabit assemblages of the host sea anemone, Entacmaea quadricolor (clonal type), which inhabit various patch reefs in the back reef moat. We located all patch reefs inhabited by the host in the map based on snorkel observations: 297 anemonefish were found in 93 host assemblages in the study site of 2.9 ha. These patch reefs could be recognized by the reef colors, including the shadows (blackish color) in the photograph. Using image analysis software, the colors of the patch reefs were chosen and pixels with the same color values were regarded as potential habitat patches for the fish (PHPs). PHPs were 3D patch reefs (>0.5 m in height). Total areas (TA) and total perimeters (TP) of PHPs were measured in nine quadrats in the digitized aerial photograph. Host abundance and anemonefish abundance in a quadrat showed stronger correlations with the product of TA and TP of PHPs than TA alone. A site with numerous large 3D patch reefs (≥0.75 m2 in situ) can be a better habitat for the fish than other sites consisting of several huge 3D patch reefs of the same total area. The methodology applied here may be useful for assessing the quality of habitats for small resident animals in shallow subtidal reefs.

Long-term trends in catch and effort in the KwaZulu-Natal nearshore linefisheries

The boat-based linefishery is the most important marine fishery along the coast of KwaZulu-Natal, producing 40% of the total annual mass of fish landed there. Since 1900, the fishery has supported commercial and recreational sectors that compete for more than 120 species, using similar methods in the same fishing areas. Catch-and-effort data have been collected sporadically from the commercial linefishery since 1910 and the recreational linefishery since 1970. The number of commercial vessels increased from 10 in 1910 to 140 in 1995, and effective commercial effort continues to increase, despite a limit placed on the number of vessels in 1985. Recreational participation escalated from 20 skiboats in 1948 to 2 000 in 1995 and entry to this sector remains unlimited. Introduction of the beach-launched, trailable skiboat in 1945 resulted in a rapid expansion of effort to coastal areas that previously functioned as harvest refugia for resident reef fish. Total catch has declined, despite increased effort, as has catch per unit effort (cpue). There have also been substantial changes in catch composition and catches have been sustained by sequential target switching from large endemic reef fish such as seventyfour Polysteganus undulosus, red steenbras Petrus rupestris and rockcods Epinephelus spp., to smaller sparids, such as slinger Chrysoblephus puniceus, santer Cheimerius nufar and blueskin Polysteganus coeruleopunctatus, and shoaling migrants, such as king mackerel Scomberomorus commerson, geelbek Atractoscion aequidens and kobs Argyrosomus spp. The linefish resources off KwaZulu-Natal have long been incapable of sustaining the fishing effort in the region and most resident species are now overexploited. Existing linefish management measures have not been successful in limiting fishing mortality to sustainable levels and existing limits on commercial permits and recreational launches will have to be substantially reduced if this is to be achieved. Bag limits may also have to be substantially reduced, particularly for endemic reef fish, to rebuild stocks to sustainable levels. Given the difficulty of implementing such controls, the maintenance of a number of large, suitably sited marine reserves appears to offer one of the few practicable chances of conserving the endemic reef fish stocks of KwaZulu-Natal.

Catch vs count: Effects of gill-netting on reef fish populations in southern New Zealand

This study investigated the relationships between visual counts of fishes and the catch from gill-nets on rocky reefs in southern New Zealand. Visual censuses were done and then gill-nets of three mesh sizes (2.5″, 3.5″ and 4.5″) were set in the surveyed areas. There were significant differences among habitats in the assemblages of reef fishes. The number and species of fishes caught by the gill-nets had little overlap with those recorded in visual surveys. The standard lengths of fishes caught in gill-nets were significantly larger than those recorded in the visual surveys, primarily because of differences in the species composition of fishes sampled by the two methods. Resident reef fishes, especially labrids, comprised most of the visual surveys, while transient pelagic species and wide-ranging reef fishes made up the largest proportion of the gill-net catch. The three mesh sizes caught different size fractions of fish populations, but all mesh sizes caught fishes larger than those seen in the visual surveys. There was a significant species × mesh size interaction in the number of fishes caught, indicating that some species were more vulnerable to particular mesh sizes. Both the number of fishes and number of species caught declined sequentially with increasing mesh size. This study shows that visual surveys and the more passive gill-netting sample different fractions of fish populations, and that gill-netting is ineffective at targeting individual species in complex reef habitats.

Influence of conspecifics on choice of settlement sites by larvae of two pomacentrid fishes (Dascyllus aruanus and D. reticulatus) on coral reefs

Field experiments investigating interactions between recruits and resident Dascyllus aruanus and D. reticulatus were carried out between 1978 and 1980 at Lizard Island, Great Barrier Reef, Australia. Standard Coral Units (SCUs) were constructed on a sandy substrate at a low tide depth of about 2 m. Each SCU was composed of living colonies of Pocillopora damicornis, measuring 60 cm in diameter by 40 cm in height. A grid of 12 SCUs was laid out such that adjacent SCUs were 20 m apart and at least 40 m from the nearest natural patch reef. Established SCUs were defaunated, and subsequently each was repopulated with a group of either D. aruanus or D. reticulatus. The fishes on each SCU were then censused over a 6 mo period spanning the recruitment season of these species. Over this period the net increase in numbers of D. aruanus was significantly greater on those SCUs predominantly inhabited by established D. aruanus than on SCUs where established D. reticulatus predominated. Conversely, the net increase in numbers of D. reticulatus was significantly greater on those SCUs predominantly inhabited by established D. reticulatus than on SCUs where established D. aruanus predominated. Although the results did not categorically rule out the possibility of differential survival after recruitment, they are most readily interpreted in terms of enhanced settlement of larvae on coral heads where conspecific fish predominate. Thus competition is not the only way in which resident reef fish may influence larval recruitment.

Depth Zonation of Corals in the Gulf of Panama: Control and Facilitation by Resident Reef Fishes

The mechanism that control the distribution and abundance of major benthic organisms in subtidal coral reef communities are poorly understood. Through field experimentation and manipulations, this study investigated the factors that account for the vertical zonation patterns found on a fringing reef in the Gulf of Panama (Pacific). In this community, pocilloporid corals form a near—monotypic stand (80—85% live coral cover) in shallow water (0—6 m depth) while massive corals, particularly Pavona gigantea, predominate in low density (°18% cover) in the deeper areas of the reef (6—10 m depth). The paucity of massive corals in shallow water results mainly from chronic and severe mortality caused by the omnivorous damselfish Eupomacentrus acapulcoensis (Pomacentridae). These highly territorial fish establish algal mats (used as a source of food and nesting site) on hard substrata, including coral. The effect of this activity was found to be differential with respect to (1) species (i.e., Pavona suffered higher mortality compared to Pocillopora) and (2) depth (i.e., the frequency and extent of damaged Pavona colonies were higher in shallow compared to deep areas). Pavona suffered higher mortality due to differences in coral morphology: damselfish are more effective at removing tissue from a massive (open) surface than from a tightly branched pocilloporid coral. Pavona was most severely damaged in shallow water because of depth—related differences in topographic complexity. By removing the topographic complexity surrounding colonies with damselfish and by adding complex substratum around colonies without damselfish, it was found that damselfish abandon territories when substratum complexity is reduced and immigrate to colonies when complexity is increased. Immigration eventually led to coral mortality. Hence, shelter is a necessary requisite for the establishment of a damselfish territory. Censuses revealed that shelter site availability surrounding Pavona colonies and damselfish abundance (both overall and per individual Pavona colony) were greater in shallow compared to deep water. Damselfish were conspicuously absent from small colonies (<1.5 m2 surface area) that lacked holes or crevices within the colony or immediately adjacent to it; such colonies were only present in deep water. Controlled experiments using live coral fragments (simulating 1—1.5 yr old juveniles) attached to hard substratum demonstrated that pocilloporid corals suffer high mortality and substantially reduced growth rates in deep reef areas due to grazing by pufferfish and parrotfish. The effectiveness of this predation/grazing activity was related to the position of the coral fragments relative to defended damselfish territories. Outside the territory, growth and survivorship of fragments were low for Pocillopora but high for Pavona. Within areas defended by damselfish, Pocillopora and Pavona fragments grew equally well both caged and uncaged. On the algal mat itself, however, Pavona fragments suffered high mortality from damselfish: Pocillopora growth under these conditions was only temporarily reduced. The data in this study indicate that damselfish territories provide a refuge against juvenile mortality for Pocillopora within defended areas, thus facilitating the establishment of branching corals in shallow water and concomitantly reducing or eliminating its potential space competitor, Pavona. A proposed successional model suggests that Pocillopora abundance may ultimately control the abundance of damselfish. This study indicates that community structure on some eastern Pacific reefs may be controlled by the direct and indirect behavioral interactions between damselfish and transient fish corallivores, and physical factors (availability of shelters) limiting the distribution of damselfish.