How does shoreline development impact the recruitment patterns of coral reef fish juveniles (Moorea Island,French Polynesia)?

The study examined the effects of coastal embankment building on fish recruitment in three habitat types (beach-rock, white sand and muddy sand) in the near shore and fringing reef habitats of Moorea lagoon (French Polynesia). The results showed a positive relationship between the presence of embankments and the density and species richness of juvenile fish along the shoreline (whatever the habitat types). However, embankments deteriorated adjacent fringing reefs (decrease of live coral), which led to a decrease of fish density on beach-rock and white sand sites, and a decrease of fish species richness on muddy sand sites.     

Influence of depth on sex-specific energy allocation patterns in a tropical reef fish

The effect of depth on the distribution and sex-specific energy allocation patterns of a common coral reef fish, Chrysiptera rollandi (Pomacentridae), was investigated using depth-stratified collections over a broad depth range (5–39 m) and a translocation experiment. C. rollandi consistently selected rubble habitats at each depth, however abundance patterns did not reflect the availability of the preferred microhabitat suggesting a preference for depth as well as microhabitat. Reproductive investment (gonado-somatic index), energy stores (liver cell density and hepatocyte vacuolation), and overall body condition (hepato-somatic index and Fulton’s K) of female fish varied significantly among depths and among the three reefs sampled. Male conspecifics displayed no variation between depth or reef. Depth influenced growth dynamics, with faster initial growth rates and smaller mean asymptotic lengths with decreasing depth. In female fish, relative gonad weight and overall body condition (Fulton’s K and hepato-somatic index) were generally higher in shallower depths (≤10 m). Hepatic lipid storage was highest at the deepest sites sampled on each reef, whereas hepatic glycogen stores tended to decrease with depth. Depth was found to influence energy allocation dynamics in C. rollandi. While it is unclear what processes directly influenced the depth-related patterns in energy allocation, this study shows that individuals across a broad depth gradient are not all in the same physiological state and may contribute differentially to the population reproductive output. Electronic supplementary material The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Growth of reef fishes in response to live coral cover

Although the global decline in coral reef health is likely to have profound effects on reef associated fishes, these effects are poorly understood. While declining coral cover can reduce the abundance of reef fishes through direct effects on recruitment and/or mortality, recent evidence suggests that individuals may survive in disturbed habitats, but may experience sublethal reductions in their condition. This study examined the response of 2 coral associated damselfishes (Pomacentridae), Chrysiptera parasema and Dascyllus melanurus, to varying levels of live coral cover. Growth, persistence, and the condition of individuals were quantified on replicate coral colonies in 3 coral treatments: 100% live coral (control), 50% live coral (partial) and 0% live coral (dead). The growth rates of both species were directly related to the percentage live coral cover, with individuals associated with dead corals exhibiting the slowest growth, and highest growth on control corals. Such differences in individual growth between treatments were apparent after 29 d. There was no significant difference in the numbers of fishes persisting or the physiological condition of individuals between different treatments on this time-scale. Slower growth in disturbed habitats will delay the onset of maturity, reduce lifetime fecundity and increase individual's vulnerability to gape-limited predation. Hence, immediate effects on recruitment and survival may underestimate the longer-term impacts of declining coral on the structure and diversity of coral-associated reef fish communities.     

Growth acceleration,behaviour and otolith check marks associated with sex change in the wrasse <Emphasis Type="Italic">Halichoeres miniatus</Emphasis>

In protogynous sex-changing fishes, females are expected to compete for the opportunity to change sex following the loss of a dominant male and may exhibit growth and behavioural traits that help them maintain their dominant status after sex change. A male removal experiment was used to examine changes in female growth and behaviour associated with sex change in the haremic wrasse Halichoeres miniatus and to test whether any changes in growth associated with sex change were recorded in otolith microstructure. Dominant females began displaying male-characteristic behaviour almost immediately after the harem male was removed. The frequency of interactions between females increased following male removal. In contrast, feeding frequency of females decreased. The largest one to three females in each social group changed sex following male removal and exhibited an increase in growth associated with sex change. Sex changers grew more than twice as fast as non-sex changers during the experimental period. This growth acceleration may enable new sex-changed males to rapidly reach a size where they can defend the remaining harem from other males. An optical discontinuity (check mark) was present in the otoliths of sex-changed fish, and otolith accretion rate increased significantly after the check mark, corresponding with the increased growth rate of sex-changing females. Wild caught males, but not females, exhibited an analogous check mark in their otoliths and similar increases in otolith increment widths after the check. This indicates that an increase in growth rate is a regular feature of sex-change dynamics of H. miniatus. Communicated by Environment Editor Prof. Rob van Woesik     

Small-scale spatial variation in the elemental composition of otoliths of Stegastes nigricans (Pomacentridae) in French Polynesia

Solution-based inductively coupled plasma-mass spectrometry (ICP-MS) was used to determine if Stegastes nigricans collected from 15 sites in French Polynesia could be distinguished by the trace element composition of their otoliths. A total of 293 adults were collected by spearing and their otoliths were analysed. We found that elemental signatures differed significantly among sites within and between the islands of Tahiti and Moorea (p<0.001), primarily due to variation in concentrations of the elements Ba, Ca, Li, Mg, Mn, Na, Sr and Y. The otoliths of fish collected within Papeete Harbour in Tahiti had distinctive elemental signatures characterised by relatively high concentrations of Mn. Otoliths of these fish could be distinguished from others that were collected only a small distance (200 m) from the harbour. This is the first time that differences in chemical composition of otoliths have been reported at such small spatial scales and this trait may prove useful for the studies of connectivity of populations at within reef scales.     

Replenishment of fish populations in the enclosed lagoon of Taiaro Atoll: (Tuamotu Archipelago, French Polynesia) evidence from eggs and larvae

 Taiaro Atoll Lagoon is normally isolated from the ocean, but at least 125 marine fish species of 31 families are present there. We sampled fish larvae in Taiaro Lagoon and the nearby ocean in February 1994 with plankton net, neuston net and light trap to investigate which taxa were completing their life cycles in the lagoon. Concentrations of fish eggs and larvae were very high in the lagoon indicating intense spawning, but larvae of only 18 taxa of 10 families were present. Only six, a callionymid, gobiids, a hemiramphid, a microdesmid, and two pomacentrids, were present across a full range of pelagic sizes, and were clearly completing their pelagic stage in the lagoon. Four other taxa, an apogonid, two labrids and a scarid, were common, but the largest individuals were small (<5 mm) postflexion larvae. These may have been completing their pelagic stage in the lagoon. The remaining lagoonal larvae (eight taxa) were rare and at the preflexion stage, so we could only conclude that they hatched from eggs spawned in the lagoon. Nineteen taxa of 15 families found as adults in the lagoon were present outside the lagoon as larvae, but not inside, suggesting that they may not normally complete their life cycles in the lagoon. Horizontal distributions of larvae in the lagoon are apparently due to the interaction of larval vertical distribution behaviour with a wind-driven countercurrent system. Accepted: 16 October 1996     

Ecological structure of assemblages of coral reef fishes on isolated patch reefs

A 9-year study of the structure of assemblages of fish on 20 coral patch reefs, based on 20 non-manipulative censuses, revealed a total of 141 species from 34 families, although 40 species accounted for over 95% of sightings of fish. The average patch reef was 8.5 m² in surface area, and supported 125 fish of 20 species at a census. All reefs showed at least a two-fold variation among censuses in total numbers of fish present, and 12 showed ten-fold variations. There was also substantial variation in the composition and relative abundances of species present on each patch reef, such that censuses of a single patch reef were on average about 50% different from each other in percent similarity of species composition (Czekanowski's index). Species differed substantially in the degree to which their numbers varied from census to census, and in the degree to which their dispersion among patch reefs was modified from census to census. We characterize the 40 most common species with respect to these attributes. The variations in assemblage structure cannot be attributed to responses of fish to a changing physical structure of patch reefs, nor to the comings and goings of numerous rare species. Our results support and extend earlier reports on this study, which have stressed the lack of persistant structure for assemblages on these patch reefs. While reef fishes clearly have microhabitat preferences which are expressed at settlement, the variations in microhabitat offered by the patch reefs are insufficient to segregate many species of fish by patch reef. Instead, at the scale of single patch reefs, and, to a degree, at the larger scale of the 20 patch reefs, most of the 141 species of fish are distributed without regard to differences in habitat structure among reefs, and patterns of distribution change over time. Implications for general understanding of assemblage dynamics for fish over more extensive patches of reef habitat are considered.     

Variation in the effects of larval history on juvenile performance of a temperate reef fish

For organisms with complex life cycles, the transition between life stages can act as a significant demographic and selective bottleneck. Variation in developmental and growth rates among individuals present in one stage (e.g. larvae), due to initial differences in parental input and/or environmental conditions experienced, can propagate to future stages (e.g. juveniles), and such ‘carry‐over effects’ can shape fitness and phenotypic distributions within a population. However, variation in the strength of carry‐over effects between life stages and the intensity of selective mortality acting on intrinsic variation, and how these might be mediated by environmental variability in natural systems, is poorly known. Here, we evaluate variation in the strength to which larval growth histories can mediate juvenile performance (growth and survival), for a reef fish (Forsterygion lapillum) common to rocky reefs of New Zealand. We used otoliths to reconstruct demographic histories of recently settled fish that were sampled across cohorts, sites and microhabitats. We quantified sources of variation in the strength of carry‐over effects and selective mortality that operate on larval growth histories. We found overall that individuals that grew fast as larvae tended to experience proportional growth advantages as juveniles. However, the strength of growth advantages being maintained into the juvenile period varied among cohorts, sites and microhabitats. Specifically, a stronger growth advantage was found on some microhabitats (e.g. mixed stands of macroalgae) relative to others (e.g. monocultures of Carpophyllum maschalocarpum) for some cohorts and sites only. For other cohorts and sites, the degree of coupling between larval and juvenile growth rates was either indistinguishable between microhabitats or else not evident. Similarly, the intensity of growth‐based selective mortality varied among cohorts, sites and microhabitats: for the cohort and site where carry‐over effects differed between microhabitats, we also observed difference in the intensity to which fish with rapid larval growth rates were favoured. Overall, our results highlight how this spatial and temporal patchiness in extrinsic factors can interact with intrinsic variation of recruiting individuals to have a major influence on the resulting distribution of juveniles and their phenotypic traits.     

High mobility of the protandrous anemonefish Amphiprion frenatus: nonrandom pair formation in limited shelter space

Protandry in anemonefishes has been attributed to random pair formation, which results from their limited mobility after random recruitment to isolated host anemones. The recruitment and movement of the anemonefish Amphiprion frenatus were investigated in relation to its group structure and the spatial distribution of its host anemone on a coral reef, where it inhabits isolated single hosts and interhost movement is rare. A juvenile tended to be recruited to a host from which a former resident(s) had disappeared, indicating that larval recruitment is not random. After mass bleaching of corals during which many hosts died, a quarter of adults moved between hosts on average 42m in 3 weeks, indicating that their mobility is potentially high. The few migrations under normal conditions were probably due to the low benefit of movements. Even after the movements, a female was much larger than her mate in a host, and a large female tended to pair with a large male. Although body size of females was positively correlated with their host size, that of males was not. It is suggested that the size of a single host does not allow the coexistence of two or more large fish, and the size composition of each pair is affected by the host size. Protandry in the monogamous fish may be attributed to the nonrandom pair formation in the limited space of a host.     

Predicting reef fish connectivity from biogeographic patterns and larval dispersal modelling to inform the development of marine reserve networks

Developing networks of no-take marine reserves is often hindered by uncertainty in the extent to which local marine populations are connected to one another through larval dispersal and recruitment (connectivity). While patterns of connectivity can be predicted by larval dispersal models and validated by empirical methods, biogeographic approaches have rarely been used to investigate connectivity at spatial scales relevant to reserve networks (10's–100's of km). Here, species assemblage patterns in coral reef fish were used together with an individual-based model of dispersal of reef fish larvae to infer patterns of connectivity in a ∼300 km wide region in the Philippines that included the Bohol Sea and adjacent bodies of water. A dominant current flows through the study region, which may facilitate connectivity among >100 no-take reserves. Connectivity was first investigated by analysing data on the presence/absence of 216 species of reef fish and habitat variables across 61 sites. Hierarchical clustering of sites reflecting species assemblage patterns distinguished a major group of sites in the Bohol Sea (Bray–Curtis similarity >70%) from sites situated in adjacent bodies of water (bays, channels between islands and a local sea). The grouping of sites could be partly explained by a combination of degree of embayment, % cover of sand and % cover of rubble (Spearman rank correlation, ρ_w = 0.42). The individual-based model simulated dispersal of reef fish larvae monthly for three consecutive years in the region. The results of simulations, using a range of pelagic larval durations (15–45 days), were consistent with the species assemblage patterns. Sites in the model that showed strongest potential connectivity corresponded to the majority of sites that comprised the Bohol Sea group suggested by hierarchical clustering. Most sites in the model that exhibited weak connectivity were groups of sites which had fish assemblages that were least similar to those in the Bohol Sea group. Concurrent findings from the two approaches suggest a strong influence of local oceanography and geography on broad spatial patterns of connectivity. The predictions can be used as an initial basis to organise existing reserves to form ecologically meaningful networks. This study showed that species assemblage patterns could be a viable supplementary indicator of connectivity if used together with predictions from a larval dispersal model and if the potential effect of habitat on the structuring of species assemblages is taken into consideration.     

Growth and reproduction in Canthigaster valentini (Pisces,Tetraodontidae): a comparison of a toxic reef fish with other reef fishes

Synopsis The growth and reproduction of Cantigaster valentini were studied in two sites at Lizard Island, Australia. C. valentini was found to be a gonochore, with a sex ratio very close to 1:1; sexes could be distinguished externally. The growth (in length) of known individuals from both sites was measured at least every two months over two years. Growth rates of males and females decrease as their sizes increase. Growth rates differ between sexes and between sites: males generally grow faster than females and individuals at Mermaid Cove generally grow faster than individuals at Palfrey Island. Spawning is demersal, it occurs daily between 0800 and 1600h, and continues year-round. For females the interval between successive spawnings varies from about 4 days in the warm-water season to about 10 days in the cool-water season. From a comparison of local reproductive output and local recruitment survivorship of larvae in the plankton was estimated to be much higher than in another species (Pomacentrus wardi) for which a similar estimate was available. We suggest that some aspects of the reproductive strategy of C. valentini differ from other, non-toxic reef fishes in ways consistent with a reduced threat of predation upon adults, eggs, and larvae: courtship and spawning are unhurried and occur throughout most of the day; spawning is unrelated to lunar cycles; there is no parental care or defense of fertilized eggs; and embryos often hatch on rising tides.Centre for Environmental and Urban StudiesSenior author's present address: School of Biological Sciences F07, University of Sydney, Sydney 2006, N.S.W., Australia     

Lunar rhythms in growth of larval fish

Growth and survival of larval fishes is highly variable and unpredictable. Our limited understanding of this variation constrains our ability to forecast population dynamics and effectively manage fisheries. Here we show that daily growth rates of a coral reef fish (the sixbar wrasse, Thalassoma hardwicke) are strongly lunar-periodic and predicted by the timing of nocturnal brightness: growth was maximized when the first half of the night was dark and the second half of the night was bright. Cloud cover that obscured moonlight facilitated a ‘natural experiment’, and confirmed the effect of moonlight on growth. We suggest that lunar-periodic growth may be attributable to light-mediated suppression of diel vertical migrations of predators and prey. Accounting for such effects will improve our capacity to predict the future dynamics of marine populations, especially in response to climate-driven changes in nocturnal cloud cover and intensification of artificial light, which could lead to population declines by reducing larval survival and growth.     

Measuring recruitment of minute larvae in a complex field environment: The corallivorous nudibranch Phestilla sibogae (Bergh)

The nudibranch Phestilla sibogae feeds only on corals of the genus Porites. The nudibranch's minute (∼ 200 μm) larvae are specifically induced to settle and metamorphose by a chemical cue released by the coral, causing the larvae to recruit to reefs composed predominantly of Porites compressa. In this study, we investigated temporal and spatial patterns of recruitment of P. sibogae into coral reefs in Kane?ohe Bay, HI. We collected heads of P. compressa at 3-week intervals for 3 years, brought them to the laboratory and maintained them in aquaria fed with filtered seawater for 2 weeks, and then examined them for the presence of juvenile P. sibogae that had grown large enough to be seen. We found that P. sibogae recruits to the Porites reefs of Kane?ohe Bay sporadically and unpredictably throughout the year. Although most coral samples contained no or very few P. sibogae, three periods of intense recruitment (90-450 juvenile P. sibogae kg^− 1 of coral) were recorded, all in different seasons. Size-frequency analysis of recruits on the coral revealed high rates of post-settlement mortality in the field, most likely due to predation. Given the short pre-competent larval period of P. sibogae, the low rate of flushing of Kane?ohe Bay and the patterns of recruitment observed, we conclude that this population of P. sibogae is essentially a self-recruiting one. Two of the sampled reefs were characterized by unidirectional flow, allowing us to test a model of transport of larvae of P. sibogae responding to dissolved coral cue in turbulent, wavy flow. The model predicts that more larvae will be transported into upstream portions of a reef than into downstream portions, a prediction confirmed by analysis of the field-recruitment data. Furthermore, field releases of larval mimic particles also showed that most mimics landed in the upstream areas of reefs and down among the bases of coral branches, rather than at their tips.     

Strategies and trajectories of coral reef fish larvae optimizing self-recruitment

Like many marine organisms, most coral reef fishes have a dispersive larval phase. The fate of this phase is of great concern for their ecology as it may determine population demography and connectivity. As direct study of the larval phase is difficult, we tackle the question of dispersion from an opposite point of view and study self-recruitment. In this paper, we propose a mathematical model of the pelagic phase, parameterized by a limited number of factors (currents, predator and prey distributions, energy budgets) and which focuses on the behavioral response of the larvae to these factors. We evaluate optimal behavioral strategies of the larvae (i.e. strategies that maximize the probability of return to the natal reef) and examine the trajectories of dispersal that they induce. Mathematically, larval behavior is described by a controlled Markov process. A strategy induces a sequence, indexed by time steps, of "decisions" (e.g. looking for food, swimming in a given direction). Biological, physical and topographic constraints are captured through the transition probabilities and the sets of possible decisions. Optimal strategies are found by means of the so-called stochastic dynamic programming equation. A computer program is developed and optimal decisions and trajectories are numerically derived. We conclude that this technique can be considered as a good tool to represent plausible larval behaviors and that it has great potential in terms of theoretical investigations and also for field applications.     

Effects of coral transplantation and giant clam restocking on the structure of fish communities on degraded patch reefs

Active restoration is being practiced to supplement conservation activities for the purpose of reversing the trend of reef degradation. In the last decade, the feasibility of different restoration approaches such as coral transplantation and restocking of other marine biota has been the focus of research and relatively few have examined experimentally its effects on the resultant communities. In this study, coral transplantation and giant clam restocking were applied on 25 degraded patch reefs (~ 25 m²) inside a marine sanctuary in Pangasinan, northwestern Philippines to examine their effects on the community structure of reef fishes. Five interventions or treatments were employed: 1) “coral” consisted of transplantation of a combination of Acropora spp. and Pocillopora spp. on concrete blocks; 2) “clam” consisted of restocking of Tridacna gigas; 3) “clam+coral” consisted of restocking of T. gigas with Acropora spp. transplanted on their shells; 4) “shell” consisted of deployment of T. gigas shells; and 5) “control” consisted of no intervention. Fish communities on the patch reefs were monitored monthly for 3 months before the intervention and were monitored further for 11 months after the intervention, including 1 recruitment season. After the intervention, the coral cover and the “other biota” category increased in the coral and clam+coral treatments, due to the transplanted corals and deployed giant clams. Consequently, the complexity of the substrate was enhanced. A month after the intervention, a rapid increase in the abundance and species richness of reef fishes on the coral, clam+coral and clam treatments was observed compared to the shell and control treatments. A change in species composition of reef fish assemblage was also apparent in the coral and clam+coral treatments relative to the clam, shell and control, especially 4 months after the intervention. The present experiment demonstrates the feasibility of improving the condition of degraded patch reefs, which can subsequently enhance the fish community. Results also show the importance of the underlying substratum and the abundance of live corals and clams to reef fishes.     

Small-scale demographic variability of the biocolor damselfish, <Emphasis Type="Italic">Stegastes partitus</Emphasis>, in the Florida Keys USA

The demographic responses of reef fish to their environment can be complex and in many cases, quite strong. Growth, mortality, longevity, and even reproductive effort have been demonstrated to vary for the same species of reef fish over scales of 100s to 1,000s of kilometers due to physiological and ecological interactions. Though few studies have explicitly documented it, this sort of habitat-mediated demography can also exist at very local scales. Here we present the results of a 2-year study of the bicolor damselfish, Stegastes partitus, in the Florida Keys, USA. We measured density and distribution, calculated key demographic rates (growth, survival, and fecundity), and characterized the environment (resident fish assemblage, substrate type and complexity, and food availability) of populations living in two adjacent but different habitats, the continuous fore reef and patchy back reef. Fish on the fore reef had an elevated growth rate and asymptotic size, increased mortality, and higher fecundity than fish on the back reef. We identified four potential causative mechanisms for these differences: food availability; competition; intraspecific density-dependent effects; and predation risk. Our data did not support an effect of either food availability or intraspecific density-dependence, but rather suggested that demographic responses are affected by both competition and predation risk.     

Spatial structure of coral reef fish communities at Kudaka Island (Ryukyu Archipelago), Japan

The present study aimed to investigate the spatial structure of fish communities at juvenile and adult stages on coral reefs at Kudaka Island (Ryukyu Archipelago, Japan) and to relate spatial patterns in the structure of the fish communities to gradients in environmental variables. Diurnal visual censuses allowed us to record 2,602 juveniles belonging to 60 species and 1,543 adults belonging to 53 species from October to December 2005. The distribution of species highlighted that the juvenile community was organised into three distinct assemblages, rather than exhibiting gradual change in community structure along the cross-reef gradient. Correlations between spatial patterns of juvenile community and environmental variables revealed that the most significant factors explaining variation in community structure were coral rubble and coral slab. In contrast, the adult community was organised into one assemblage, and the most significant variation factors in community structure were depth, live coral in massive form, live coral in branched form, dead coral and sand. Overall, the present study showed that most juvenile and adult coral reef fish at Kudaka Island exhibited striking patterns in their distribution and depth and some biological factors (e.g., abundance of live coral, dead coral and coral rubble) might exert considerable influence on the distribution of fishes.     

Strong nonlinear selection against fluctuating asymmetry in wild populations of a marine fish

Theoretical links between fluctuating asymmetry (FA) and fitness have led many to use FA as a proxy for average fitness. However, studies examining whether asymmetry actually correlates with individual fitness in wild populations are relatively rare and often use simple measures of association (e.g., correlation coefficients). Consequently, the pattern of selection on asymmetry in the wild is seldom clear. We examined selection on FA of pectoral fin morphology in two wild populations of a marine fish (the kelp perch; Brachyistius frenatus). As expected, variance in signed FA in each initial sample was significantly greater than that found in the surviving population, indicating selection against FA. Our estimate of the fitness surface confirmed perfect symmetry as the phenotypic optimum and indicated strong, nonlinear selection against asymmetry. No difference in the form of selection was detected between populations. However, the level of FA in the initial samples varied among populations, leading to an overall difference in the level of selective mortality. Our results suggest that selection on asymmetry in wild populations may be strongly nonlinear, and indicate that the demographic costs of asymmetry may play a substantial role in the dynamics of populations.     

Habitat complexity modifies the impact of piscivores on a coral reef fish population

Patterns in juvenile mortality rates can have a profound affect on the distribution and abundance of adult individuals, and may be the result of a number of interacting factors. Field observations at Lizard Island (Great Barrier Reef, Australia) showed that for a coral reef damselfish, Pomacentrus moluccensis, juvenile mortality (over 1 year) varied between 20 and almost 100% among sites. Correlative data showed that juvenile mortality increased as a function of initial densities (recruitment), predator densities and the availability of preferred coral substrata. A multiple regression showed that these three variables together did not explain significantly more variation in mortality than the single factor showing the strongest relationship. This appeared to be because recruitment, predator densities and preferred coral substrata were all highly correlated, suggesting that one, two or all of these factors may be influencing juvenile mortality rates. One hypothesis was that density-dependent mortality in juveniles was the result of an interaction between predators (which appear to aggregate at high-recruitment sites) and the availability of preferred substrata (predator refuges). We tested this hypothesis by using both laboratory and field experiments to see whether fish predation could significantly alter survivorship of this damselfish, and whether this impact was dependent upon the coral substratum. The laboratory experiment was designed to test the effects of three common predators (Pseudochromis fuscus, Cephalopholis boenak and Thalassoma lunare) and three different coral substrata that varied in their complexity (Pocillopora damicornis, Acropora nasuta and A. nobilis) on the survival of juvenile Pomacentrus moluccensis. There was a significant interaction between predator species and microhabitat in determining survival. Pseudochromis fuscus and C. boenak were both significantly better at capturing juvenile damselfish than T. lunare. Juvenile survivorship was significantly better when they were given the more complex corals, Pocillopora damicornis and A. nasuta, compared with those given the open-structured species A. nobilis. This pattern reflects habitat selection in the field. Predators differed in their strike rates and the proportion of strikes that were successful, but all exhibited greater success at prey capture where A. nobilis was provided as shelter. The interaction between the effect of predator species and microhabitat structure on damselfish survival was tested in the field for a cohort of juvenile Pomacentrus moluccensis. We examined juvenile survival in the presence and absence of two predators that co-occur on natural patch reefs (C. boenak and Pseudochromis fuscus). The experimental patch reefs we used for this purpose were constructed from both high complexity (Pocillopora damicornis) and low complexity (A. nobilis) coral substrata. Both juveniles and predators were translocated to reefs at natural densities. The effects of predation were clearly dependent upon the microhabitat. Reefs of the high-complexity coral with predators supported the same high numbers of Pomacentrus moluccensis as the reefs with no resident predators. However, damselfish abundance was significantly lower on low-complexity reefs with resident predators, relative to the other treatments. Background rates of loss were high, even on preferred coral in the absence of the manipulated predator, suggesting that transient predators may be even more important than the residents. We suggest that adult abundances in this species were strongly influenced by the densities of different predators and the availability of preferred refuges. Received: 3 April 1997 / Accepted: 26 August 1997