Search

Coral reef ecosystems are at the forefront of biodiversity loss and climate change-mediated transformations. This is expected to have profound consequences for the functioning of these ecosystems. However, assessments of ecosystem function on reefs are often spatially limited, within biogeographic realms, or rely on presumed proxies such as traits. To address these shortcomings and assess the effects of biogeography and fish presence on the critical ecosystem function of macroalgal removal, we used assays of six algal genera across three reef habitats in two biogeographically distinct locations: Little Cayman in the Caribbean and Lizard Island on the Great Barrier Reef (GBR). Patterns of fish feeding and realised ecosystem function were strikingly similar between the two geographic locations, despite a threefold difference in the local diversity of nominally herbivorous fishes, a 2.4-fold difference in the diversity of fishes feeding and differences in the biogeographic history of the two locations. In both regions, a single species dominated the function: a surgeonfish, Naso unicornis, at the GBR location and, surprisingly, a triggerfish, Melichthys niger, at the Caribbean location. Both species, especially M. niger, were relatively rare, compared to other nominally herbivorous fishes, in censuses covering more than 14,000 m2 at each location. Our study provides novel insights into the critical function of macroalgal removal in hyperdiverse coral reef ecosystems, highlighting: (a) that function can transcend biogeographic, taxonomic and historical constraints; and (b) shortcomings in our assumptions regarding fish presence and realised ecosystem function on coral reefs.

Western Australia’s coral reefs have largely escaped the chronic pressures affecting other reefs around the world, but are regularly affected by seasonal storms and cyclones, and increasingly by heat stress and coral bleaching. Reef systems north of 18°S have been impacted by heat stress and coral bleaching during strong El Nino phases and those further south during strong La Nina phases. Cumulative heat stress and the extent of bleaching throughout the northern reefs in 2016 were higher than at any other time on record. To assess the changing regime of disturbance to reef systems across Western Australia (WA), we linked their site-specific exposure to damaging waves and heat stress since 1990 with mean changes in coral cover. Since 2010, there has been a noticeable increase in heat stress and coral bleaching across WA. Over half the reef systems have been severely impacted by coral bleaching since 2010, which was further compounded by cyclones at some reefs. For most (75%) reef systems with long-term data (5–26 yrs), mean coral cover is currently at (or near) the lowest on record and a full recovery is unlikely if disturbances continue to intensify with climate change. However, some reefs have not yet experienced severe bleaching and their coral cover has remained relatively stable or increased in recent years. Additionally, within all reef systems the condition of communities and their exposure to disturbances varied spatially. Identifying the communities least susceptible to future disturbances and linking them through networks of protected areas, based on patterns of larval connectivity, are important research and management priorities in coming years while the causes of climate change are addressed.

Canopy-forming macroalgae can construct extensive meadow habitats in tropical seascapes occupied by fishes that span a diversity of taxa, life-history stages and ecological roles. Our synthesis assessed whether these tropical macroalgal habitats have unique fish assemblages, provide fish nurseries and support local fisheries. We also applied a meta-analysis of independent surveys across 23 tropical reef locations in 11 countries to examine how macroalgal canopy condition is related to the abundance of macroalgal-associated fishes. Over 627 fish species were documented in tropical macroalgal meadows, with 218 of these taxa exhibiting higher local abundance within this habitat (cf. nearby coral reef) during at least one life-history stage. Major overlap (40%-43%) in local fish species richness among macroalgal and seagrass or coral reef habitats suggest macroalgal meadows may provide an important habitat refuge. Moreover, the prominence of juvenile fishes suggests macroalgal meadows facilitate the triphasic life cycle of many fishes occupying diverse tropical seascapes. Correlations between macroalgal canopy structure and juvenile abundance suggests macroalgal habitat condition can influence levels of replenishment in tropical fish populations, including the majority of macroalgal-associated fishes that are targeted by commercial, subsistence or recreational fisheries. While many macroalgal-associated fishery species are of minor commercial value, their local importance for food and livelihood security can be substantial (e.g. up to 60% of landings in Kenyan reef fisheries). Given that macroalgal canopy condition can vary substantially with sea temperature, there is a high likelihood that climate change will impact macroalgal-associated fish and fisheries.

Submerged reef ecosystems can be very diverse and may serve as important refugia for shallow-water conspecifics. This study quantified the benthic and fish communities of two proximate, predominantly mesophotic coral ecosystems (MCEs), Glomar Shoal and Rankin Bank, which are geographically isolated from other similar features in the region. Glomar Shoal is identified as a key ecological feature (KEF) in the North West Marine Region of Australia. Multibeam surveys were performed to characterise the seafloor and to derive secondary environmental variables, used to explain patterns in benthic and fish communities. Towed video surveys quantified benthic cover, and stereo baited remote underwater stations were used to survey fish abundance and diversity. Surveys were completed in depths of ~ 20–115 m. The two MCEs exhibited distinct communities; Rankin Bank consistently had higher cover (up to 30×) of benthic taxa across depths, and fish communities that were twice as abundant and 1.5× more diverse than Glomar Shoal. The location of the MCEs, depth and rugosity were most influential in structuring benthic communities. Phototrophic taxa, specifically macroalgae and hard corals, had up to 22 × higher cover at Rankin Bank than at Glomar Shoal and were dominant to 80 m (compared to 60 m at Glomar Shoal), presumably due to greater light penetration (lower turbidity) and lower sand cover at greater depths. The 20% coral cover at Rankin Bank was comparable to that reported for shallow reefs. The cover of sand, hard corals and sponges influenced fish communities, with higher abundance and diversity of fish associated with shallow hard coral habitats. This study demonstrated that the two MCEs were unique within the local context, and when coupled with their geographical isolation and biodiversity, presents compelling support for the additional recognition of Rankin Bank as a KEF.

Dredging can have significant impacts on aquatic environments, but the direct effects on fish have not been critically evaluated. Here, a meta-analysis following a conservative approach is used to understand how dredging-related stressors, including suspended sediment, contaminated sediment, hydraulic entrainment and underwater noise, directly influence the effect size and the response elicited in fish across all aquatic ecosystems and all life-history stages. This is followed by an in-depth review summarizing the effects of each dredging-related stressor on fish. Across all dredging-related stressors, studies that reported fish mortality had significantly higher effect sizes than those that describe physiological responses, although indicators of dredge impacts should endeavour to detect effects before excessive mortality occurs. Studies examining the effects of contaminated sediment also had significantly higher effect sizes than studies on clean sediment alone or noise, suggesting additive or synergistic impacts from dredging-related stressors. The early life stages such as eggs and larvae were most likely to suffer lethal impacts, while behavioural effects were more likely to occur in adult catadromous fishes. Both suspended sediment concentration and duration of exposure greatly influenced the type of fish response observed, with both higher concentrations and longer exposure durations associated with fish mortality. The review highlights the need for in situ studies on the effects of dredging on fish which consider the interactive effects of multiple dredging-related stressors and their impact on sensitive species of ecological and fisheries value. This information will improve the management of dredging projects and ultimately minimize their impacts on fish.

Identifying and protecting nursery habitats for species is a key conservation strategy for the long-term sustainability of populations. In tropical ecosystems, macroalgal habitats have recently been identified as nurseries for fish of commercial and conservation significance. Here, we explore how local-scale variations in seaweed habitat quality interact with large-scale climatic conditions (Southern Oscillation Index, SOI) to influence the recruitment of three tropical fish species (Lethrinus spp.), often targeted by fishers. New fish recruits and juveniles of all species were almost exclusively found in macroalgal nursery habitats, while adults of two of these species were predominantly found on adjacent coral reefs. Annual supply rates of new recruits were found to be strongly correlated to variations in the SOI, with La Nina conditions associated with higher recruitment. However, local rates of recruitment were generally poor predictors of older juvenile abundance. Instead, local juvenile abundance was more closely related to structural characteristics of macroalgae nursery habitat quality (density, canopy height, canopy cover) and/or predator biomass, at the time of survey, with species-specific habitat associations apparent. Given the dynamic nature of fish recruitment supply to the SOI, coupled with the effects of climatic and oceanic processes on the structure of macroalgal patches, these results suggest protection of macroalgal nursery habitats that maintain high canopy density, height and cover is critical to supporting the conservation of fish populations.

Coral reefs have been heavily affected by elevated sea-surface temperature (SST) and coral bleaching since the late 1980s; however, until recently coastal reefs of north-western Australia have been relatively unaffected compared to Timor Sea and eastern Australian reefs. We compare SST time series with changes in coral cover spanning a period of up to 36 years to describe temporal and spatial variability in bleaching and associated coral mortality throughout the Pilbara–Ningaloo region. Declines in coral cover ranged from 12.5 to 51.3%, with relative declines ranging from 38 to 92%. Since 2013, coral cover throughout the region has declined to historically low levels at four of five subregions, with impaired recovery occurring at two subregions. Observations are consistent with global trends of repeated severe heat waves, coral bleaching and acute declines in coral cover. Locations within this study region have already experienced multiple coral-bleaching events within a period of less than 5 years. There is a high likelihood that reefs in the western Pilbara and northern Ningaloo regions will experience more frequent marine heatwaves, coral bleaching and mortality events in the future. Action, therefore, needs to be taken now to support the resilience of coral reef ecosystems in the region, which is arguably the most important coral-reef province on Australia’s western coast.

Understanding the dynamics of juvenile fish populations is a key component to understanding variation in adult populations, however, juvenile fish are frequently found in a variety of structurally complex habitats which are often not easily sampled. The use of baited (BRUV) and unbaited (RUV) remote underwater stereo-video to study juvenile fish populations is uncommon and thus far, no study has directly compared the efficacy of these two methods solely for juvenile fish assemblages. The macro-tidal Kimberley region in the north-west of Australia is challenging to sample using diver-based methods due to the strong currents and prevalence of saltwater crocodiles (Crocodylus porosus). We sampled 60 Stereo-RUVs and 60 Stereo-BRUVs across four shallow-water (1-6 m) benthic habitats: coral, mangrove, macroalgae, and seagrass, to contrast the effect of the presence or absence of bait, deployment period, in-water visibility and tidally driven water speed on estimates of relative total abundance, species richness, and composition of the juvenile fish assemblage. No difference was detected in the ability of stereo-BRUV or stereo-RUV to quantify the relative total abundance, species richness, or assemblage composition of juvenile fish. We also found little effect of in-water visibility or tidally driven water speed, likely as a result of stratifying our sampling to slack water periods. We found that a deployment period of 10 min for Stereo-BRUVs and 15 min for Stereo-RUVs was optimum for sampling the juvenile fish assemblage across all four contrasting habitats. Since no statistical significance was observed between 10 and 15 min, we recommend that Stereo-RUVs deployed for 15 min during tidal slack water conditions are an optimum way to provide consistent results for comparisons of juvenile fish assemblage metrics across the benthic habitats studied within this region. The controlled video methodologies described here provided an opportunity to access un-sampled juvenile fish habitats and assemblages where traditional diver-based methodologies are impossible. Applying these methods broadly can help to improve our understanding of the status of juvenile fish and infer the reproductive success and viability of their populations across a range of habitat and environments.

Within the complex food webs that occur on coral reefs, mesopredatory fish consume small-bodied prey and transfer accumulated biomass to other trophic levels. We estimated biomass, growth and mortality rates of three common mesopredators from Ningaloo Reef in Western Australia to calculate their annual turnover rates and potential contribution to local trophic dynamics. Biomass estimates of the serranid Epinephelus rivulatus (4.46 ± 0.76 g m−2) were an order of magnitude greater than two smaller-bodied mesopredatory fishes, Pseudochromis fuscus (0.10 ± 0.03 g m−2) and Parapercis clathrata (0.23 ± 0.31 g m−2). Growth parameters generated from a von Bertalanffy growth function fitted to size-at-age data, however, indicated that mortality rates for the three mesopredators were similar and that 32–55 % of fish survived each year. Consequently, interspecific differences in annual turnover rates among E. rivulatus (1.9 g m−2 yr−1), Pa. clathrata (0.10 g m−2 yr−1) and Ps. fuscus (0.07 g m−2 yr−1) were an artefact of differences in local biomass estimates. The rapid turnover estimates for E. rivulatus suggest this species is an important conduit of energy within the isolated patch reef habitat where it is typically found, while Ps. fuscus and Pa. clathrata channel smaller amounts of energy from specific habitats in the Ningaloo lagoon. Apparent differences in habitat, diet and turnover rates of the three species examined provide an insight into the different roles these species play in coral reef food webs and suggest that life-history traits allow for variability in the local and spatial contribution of these species at Ningaloo Reef. Moreover, calculating turnover rates of a broader suite of fish species from a range of trophic groups will help better define the role of fishes in coral reef trophic dynamics.

We thank staff at the Department of Parks and Wildlife Exmouth for field support, AIMS 2013 Appropriation funding 3.3.5 and the Australian Research Council for financial support.

The diets of four common mesopredator fishes were examined in the back-reef habitat of a subtropical fringing reef system during the summer months. Quantitative gut content analyses revealed that crustaceans, represented >60% of ingested prey (% mass) by the latticed sand-perch Parapercis clathrata, brown dottyback Pseudochromis fuscus and half-moon grouper Epinephelus rivulatus. Dietary analyses also provided insights into ontogenetic shifts. Juvenile P. fuscus ingested large numbers of crustaceans (amphipods and isopods); these small prey were rarely found in larger individuals (

We thank P. Tinkler and staff of the Exmouth Office of the Department of Environment and Conservation Western Australia for field assistance; A. Heyward, R. Lowe, and C. Simpson for helpful discussions; the Australian Research Council, Australian Institute of Marine Science, Western Australian Marine Science Institute, and the Department of Environment and Conservation Western Australia for funding; and two anonymous reviewers for helpful comments. Data collection was conducted in the traditional country of the Jinigudira and Bayungu tribes of the Thalanyji people.

Fish predation is often cited as a key process in structuring sea urchin populations on coral reefs, with population outbreaks often being related to the removal of key predators through overfishing. However, moderate–high densities of the sea urchin Echinometra mathaei have been reported on a reef with relatively intact predator assemblages; Ningaloo Reef, Western Australia. This study examined the relationship between the densities of E. mathaei and its potential predators, to gain some insight into the potential role of predation in structuring E. mathaei populations on Ningaloo Reef. To identify potential urchin predators we recorded predation events on tethered E. mathaei using stationary video cameras. Seven fish species preyed on the tethered urchins with two labrid species, Coris aygula and Choerodon rubescens, accounting for 65% of all observed predation events. There was, however, no evidence for the role of predation in determining E. mathaei populations either within or among habitats. Despite the densities of E. mathaei varying from 0.1 to 219.7 ind 100 m− 2 among habitats, the density and biomass of potential urchin predators displayed limited variation among habitats. Furthermore, the density of E. mathaei was positively related to that of their predators on the reef slope and the back reef. While the overall density of potential predators (53.3 ind ha− 1) was comparable to other protected reefs, the suite of predators differed from that of other regions. In particular, large triggerfish species (f. Balistidae), the dominant predators of E. mathaei on other Indo-Pacific reefs, were rare or absent. While the lack of these species may have contributed to the moderate–high densities of E. mathaei on Ningaloo Reef, other factors such as larval supply, food availability and habitat characteristics may be important. Irrespective of the mechanisms, moderate–high densities of E. mathaei should not be universally viewed as an indicator of reef degradation.

The spring and summer of 2010/11 saw an exceptionally strong La Nina push warm waters from Indonesia down the Western Australian coastline, resulting in a host of extraordinary biological oddities including significant bleaching of Western Australian corals. Here, we report a 79–92 % decline in coral cover for a location in the Ningaloo Marine Park where sustained high water temperatures over an 8-month period left just 1–6 % of corals alive. The severity of bleaching provided an opportunity to investigate the resilience of different taxonomic groups and colony size classes to an acute but protracted episode of thermal stress. While the sub-dominant community of massive growth forms fared reasonably well, the dominant Acropora and Montipora assemblages all died, with the exception of the

Lipofuscin, an autofluorescent biomarker of physiological wear-and-tear, was concentrated in those areas of a fish's midbrain responsible for visual performance, suggesting a potentially strong link between physiological specialization, ecological adaptation and senescence.

The number of relatively untouched coral reefs worldwide is rapidly decreasing. Nevertheless, one coral reef ecosystem remains relatively intact: the largest west-continental reef ecosystem in the world, Ningaloo Reef in Western Australia. This study investigated the status of 2 potential bio-indicators for coral reef decline, macroalgae and sea urchin densities, on this reef. Sur- prisingly, both were abundant, with the presence of extensive macroalgal beds in the sandy lagoon and a sea urchin-dominated reef slope. The algal distribution on Ningaloo reflected marked cross- shelf variation in the composition of fish functional groups, with only the back reef and the reef slope exhibiting high grazing rates (completely scraped every 43 and 59 d, respectively). Estimated bioero- sion rates by fishes ranged between 1 and 2.3 kg m -2 yr -1 . Echinoids only played a significant role in bioerosion on the reef slopes owing to their high abundance in that habitat (>12 individuals m -2 ). Here, estimated echinoid erosion equalled that of the most abundant excavating parrotfish, Chloru- rus sordidus. High echinoid and macroalgal abundances on this relatively intact reef system highlight the need for caution when using these metrics for evaluating reef ecosystem condition.

Ocean acidification is one of the key threats facing coral reef ecosystems, but there are few estimates of spatial and temporal variability in pH among reef habitats. The present study documents levels of spatial variability in pH among coral reef habitats (9 to 10), among locations separated by 100’s km of latitude and between east (Great Barrier Reef, GBR) and west (Ningaloo Reef) coasts of Australia. Differences were found in pH between inshore and offshore waters along Ningaloo Reef (means 8.45, 8.53, respectively). Replicate assessments here ranged from 8.22 to 8.64. On the GBR, the range of values over all habitats and replicates was 0.39 pH units (7.98 to 8.37). There were minor but significant differences of 0.05 pH units between 5 consecutive days for habitats on average. Highest pH was recorded in filamentous algal beds maintained by the damselfish Dischistodus perspicillatus. Lowest pH was found in water extracted from sand-dwelling goby holes. While there were marked changes in pH over a 48-h sampling period among 4 habitats at Lizard Island (GBR), there was little evidence of a diel trend. Understanding how pH varies at scales that are relevant to organisms that live on shallow coral reefs is crucial for the design and interpretation of experiments that test the effects on organisms of the changes in water chemistry predicted to affect oceans in the future.

The distribution and abundance of small cryptic reef fishes were quantified among microhabitats and reef zones at both exposed and sheltered sites at Lizard Island, Great Barrier Reef. A total of 1042 individuals from 44 species in 8 families were sampled. Marked variation in abun- dance, species numbers, size-class distribution and species composition were displayed among reef zones at exposed sites; in contrast, comparatively little variation was found in sheltered sites. The exposed, wave-swept front flat reef zone was clearly the most depauperate and species-poor zone with calmer back reef, lagoonal and reef base zones containing the highest numbers of individuals and species. Larger, heavier individuals predominated in exposed wave-swept reef zones whereas smaller, lighter individuals were commonly found in calmer waters. Sheltered sites exhibited little variation among zones. In these sites, microhabitat appeared to play a much greater role in shaping community composition. Overall, 68.5% of all individuals (70.1% at exposed and 65.4% at sheltered sites) were collected from sand/rubble microhabitats as opposed to 31.5% from open reef microhab- itats. Assemblages at exposed and sheltered sites exhibited little taxonomic overlap, even of highly abundant species. These results indicate that wave energy plays a significant role in structuring small cryptic coral reef fish communities, with microhabitat type playing a key role in the absence of wave energy.

An examination of the trophic status of a cryptobenthic reef fish community from the central Great Barrier Reef was carried out to evaluate the potential role of cryptobenthic fishes in coral reef ecosystems. Using frequency of occurrence data, dietary analyses revealed a diverse range of trophic groups, although detritivory (in 10 out of 16 species and 39.3% of individuals) and carnivory (5 of 16 species and 40.5% of individuals) clearly dominate as trophic modes. Herbivory (1 species; 2.4% of individuals) is only a minor component in the community trophic structure. Of the 18 dietary categories identified, detritus and copepods were the only constituents represented in all 16 species examined. Although the degree of dietary specialization varied among taxa, the 2 most abundant species, Eviota queenslandica and Istigobius goldmanni, utilised the broadest range of dietary items. Morphology reflected the trophic partitioning among fishes: carnivores were invariably 38 mm TL with GLRs generally exceeding >1.0. The trophic composition and numerical strength of the cryptobenthic fish fauna suggests that cryptobenthic reef fishes have the potential to make a significant contribution to reef trophodynamics along a number of trophic pathways. This prompts a re-evaluation of the roles of reef fishes in the functioning of coral reefs, particularly those related to the recycling of primary production through detrital pathways.

Marine ecosystems are facing increasing exposure to a range of stressors and declines in critical ecological functions. The likelihood of further loss of functions and resilience is dependent, in part, on the extent of functional redundancy (i.e. the capacity of one species to functionally compensate for the loss of another species) within critical functional groups. We used multiple metrics; species richness, generic richness, abundance and reserve capacity (i.e. the relative number of individuals available to fulfil the function if the numerically dominant species is lost), as indicators to assess the potential functional redundancy of four functional groups of herbivorous fishes (browsers, excavators, grazers and scrapers) in two of the worlds' most intact coral reef ecosystems: the Great Barrier Reef (GBR) and Ningaloo Reef in Western Australia. We found marked variations in potential redundancy among habitats within each reef system and functional groups. Despite negligible fishing of herbivorous fishes, coastal habitats in both reef systems had lower functional redundancy compared to offshore locations for all herbivorous fishes collectively and the four functional groups independently. This pattern was consistent in all four indicators of redundancy. The potential vulnerability of these coastal habitats is highlighted by recent shifts from coral to macroalgal dominance on several coastal reefs of the GBR. Our approach provides a simple yet revealing evaluation of potential functional redundancy. Moreover, it highlights the spatial variation in potential vulnerability and resilience of reef systems.

SUMMARY Expert opinion was canvassed to identify crucial knowledge gaps in current understanding of climate change impacts on coral reef fishes. Scientists that had published three or more papers on the effects of climate and environmental factors on reef fishes were invited to submit five questions that, if addressed, would improve our understanding of climate change effects on coral reef fishes. Thirty-three scientists provided 155 questions, and 32 scientists scored these questions in terms of: (i) identifying a knowledge gap, (ii) achievability, (iii) applicability to a broad spectrum of species and reef habitats, and (iv) priority. Forty-two per cent of the questions related to habitat associations and community dynamics of fish, reflecting the established effects and immediate concern relating to climate-induced coral loss and habitat degradation. However, there were also questions on fish demographics, physiology, behaviour and management, all of which could be potentially affected by climate change. Irrespective of their individual expertise and background, scientists scored questions from different topics similarly, suggesting limited bias and recognition of a need for greater interdisciplinary and collaborative research. Presented here are the 53 highest-scoring unique questions. These questions should act as a guide for future research, providing a basis for better assessment and management of climate change impacts on coral reefs and associated fish communities.

The cryptobenthic reef fish communities from four microhabitats at Orpheus Island, central Great Barrier Reef are described. Eighty-four 0.4m2 samples yielded a total of 368 individuals from 42 species in eight families, with a mean density of 11 individuals m−2 (±1.7SE) and 2.9 species 0.4 m−2 (±0.2SE). Caves contained the highest number of both individuals (120) and species (26), followed by sand/rubble, soft coral, and open reefs. Microhabitat associations included cave and soft coral specialists. Site fidelity in 71 tagged individuals of 4 species was high, with a mean recapture rate of 53% (±8.4SE) remaining within the ~0.4 m2 sampling area after a 48-h period. Behavioural observations also reflect this limited movement, with the dominant mode of behaviour in 7 species being a motionless state (67.5% ±11.6SE), followed by feeding (21.8% ±8.7SE), hiding (6.3% ±1.6SE), and swimming (4.4% ±1.5SE). Two distinct behavioural groups are identified: (1) sedentary forms, characterised by long periods of immobility (5 species); and (2) winnowers, characterised by long feeding bouts (2 species). The fine-scale partitioning of microhabitats, restricted home ranges, and sedentary behaviour of many cryptobenthic reef fish species suggest that this reef fish community exhibits similar patterns of habitat utilisation to their larger reef-fish counterparts, but at a much finer scale.

Tropical coral reefs are facing increasing levels of disturbance, with predictions of ongoing reductions in live coral cover. The response of herbivorous fishes to loss of coral cover was investigated by analysing the relationship between the abundance of different sized herbivorous fishes and live coral cover, on a high flow reef slope and a sheltered back reef on Ningaloo Reef, Western Australia. We found positive relationships between coral cover and the abundance of small (10–20 cm) excavating and scraping parrotfishes on the slope, but a negative relationship between corals and small (10–15 cm) grazers. These relationships were not evident on back reefs. Generally, the abundance of large individuals were not correlated with live coral in either habitat. We suggest that in physically demanding environments small parrotfishes may utilise the structural complexity of corals to resist strong water flow and/or to reduce predation. Small acanthurid grazers may school to reduce energy costs and minimise predation risk. The loss of coral cover appears to be particularly important for small fishes in high energy habitats, especially non-schooling species.